This is Prog.gz in view mode; [Up]
From: Bill Bumgarner <bbum_not@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: Bitmap format Date: Wed, 01 Apr 1998 10:02:57 -0500 Organization: CodeFab, Inc. Message-ID: <35225721.9BC2F81C@codefab.com> References: <352172a8.19123367@news> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Michael Simpson <simpson@nospam.cts.com> Xcanpos: shelf.01/199804131801!0030615104 TIFF, EPS, PS, GIF, JPEG, BMP are all built into the OpenStep imaging model. But, it is not limited to that-- the OS imaging model makes it trivial to add filters that will automatically convert foreign formats to the native format [TIFF being the basis format]. As such, converters for basically every format known to the virtual community have been made available over the years... b.bum Michael Simpson wrote: > Does NextStep/OpenStep utilize a standard bitmap format? Windows has > .bmp, Mac has PICT. What does OpenStep have? > > Thanks, > Michael
From: Bill Bumgarner <bbum_not@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: Dragging & Dropping an id Object in an Application ? Date: Wed, 01 Apr 1998 10:05:51 -0500 Organization: CodeFab, Inc. Message-ID: <352257CF.6F639A0E@codefab.com> References: <352202D6.F906C0D@amg.de> <6ftccs$mot$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Xcanpos: shelf.01/199804131801!0031174918 Actually, there is no reason why he can't have his cake and eat it too, per se... The Pasteboard allows one to place multiple data types on the pasteboard-- the consumer of the pasteboard (same app, different app, doesn't matter) can then choose the most appropriate type for consumption. So, he could stick an NSData object on the pasteboard that contains a pointer (the id) of his object and mark it with a private type that is only consumed by his application AND he could place a second type on the pasteboard that contains the archived version of the object-- or some other representation-- for consumption by other applications. Also, the pasteboard allows you to lazy fill the PB-- that is, the app doesn't have to produce the date for the pasteboard until the pasteboard is actually consumed somewhere else. b.bum Nathan Urban wrote: > In article <352202D6.F906C0D@amg.de>, mm@amg.de wrote: > > > * I would like to drag an Object of type id (i.e. put the reference > > (pointer) to this id object on the pasteboard) so that the Destination > > where the drop takes place can get this reference from the pasteboard > > and can refer to exactly the object that was put on the pasteboard by > > the Source. > > You can't. You can refer to a _copy_ of the object. You do this by > serializing the object into an NSData and then reconstructing it on the > other end. I guess you could hack it so that the NSData held a pointer > to the original object and then reconstruct the pointer from that, but > that would defeat the whole reason why you put archived objects onto > the pasteboard via NSDatas instead of the objects themselves -- you're > supposed to be able to copy and paste objects _between applications_, > which don't share a common address space. (If you really wanted, > I suppose you could put information in the NSData that indicates how > to construct a DO proxy to the original object.. but in most cases it > should be sufficient to just have the other end reconstruct a copy of > the serialized original and use that.)
From: Greg_Anderson@afs.com (Gregory H. Anderson) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Hilighting for drag destination? Date: 31 Mar 1998 17:31:17 GMT Organization: Anderson Financial Systems Inc. Message-ID: <6fr995$f2v@shelob.afs.com> References: <6fql92$hdt$1@ns3.vrx.net> Xcanpos: shelf.01/199804122201!0028317661 Maury Markowitz writes > I'm going about implementing some dragging behaviour and I noticed > that OpenStep doesn't appear to have any standards to hilight the view > that's the target of the drag - it uses changes to the mouse pointer > only to indicate drags. Under MacOS8 the standard is to hilight the > receiving view with a small inset mauve border, and I find this > additional feedback to be terribly handy, notably when there's > views-in-views as it clearly shows you the target. (followups set to c.s.n.programmer, as this really belongs there) You need to define these methods of the NSDraggingInfo protocol: - (unsigned int)draggingEntered:(id <NSDraggingInfo>)sender - (unsigned int)draggingUpdated:(id <NSDraggingInfo>)sender - (void)draggingExited:(id <NSDraggingInfo>)sender - (BOOL)prepareForDragOperation:(id <NSDraggingInfo>)sender (You may have reason to define other parts of the protocol, but these are the required ones for highlighting operations.) The basic method we use in WriteUp and PasteUp is to identify the dropRect that needs to be highlighted, then execute the following code: [self lockFocus]; NSHighlightRect(dropRect); dropRect = NSInsetRect(dropRect , 2.0 , 2.0); NSHighlightRect(dropRect); dropRect = NSInsetRect(dropRect , -2.0 , -2.0); [self unlockFocus]; [[self window] flushWindow]; This has the effect of highlighting the whole area, insetting the rect by two points, then "unhighlighting" the interior portion, which leaves a perfect two-point highlighted border. You don't get any flashing with this method, because none of the operation is pushed through to the windowserver until you flushWindow. The unfortunate thing about this process, from the standpoint of using categories to do the drawing, is that you have to retain the previous dropRect between successive calls to the methods listed above. That normally requires an ivar, which is not allowed in categories. However, if you are willing to assume that only one drag operation can occur at any given time -- a reasonable assumption in current user interfaces -- then you could use a static global variable in your category to hold the dropRect value. Procedurally, you need to draw the highlight rect (using the sequence above) in the draggingEntered: method; undraw the old rect (execute the same code again with the prior dropRect, voila!: it disappears), recalc bounds, and then redrew a new rect in draggingUpdated:; and undraw (only) in draggingExited: and prepareForDragOperation:. Why undraw and redraw in draggingUpdated:, you ask? Because the view may have scrolled, or you may want to focus on a different portion, based on the current mouse location. Of course, you will need to decide how to identify the rect to highlight. The view's bounds is the obvious choice, but I like to draw a ring that is visible all the way around, which means you need to intersect with the visible rect. Also, for some operations we like to highlight the focused paragraph, which rolls right down the screen as the user moves the mouse. -- Gregory H. Anderson | "We're in the land of the blind, Visionary Ophthalmologist | selling working eyeballs, and they Anderson Financial Systems | balk at the choice of color." -- Tony greg@afs.com (NeXTmail OK) | Lovell, on Mac user reactions to NeXT
From: Ted Bilmer<ted@biosys.net> Newsgroups: comp.sys.next.programmer Subject: - <<< Test Drive THIS for 30 Days - FREE! >> Date: 1 Apr 1998 20:26:04 GMT Organization: TMnet Malaysia Message-ID: <6fu7ss$6sj$309@news.tm.net.my> Xcanpos: shelf.01/199804140201!0016122354 YES, you can test-drive this MLM (Multi-Level Marketing) for 30 day, FREE. Build your downlines during this period. If you don't succeed just quit, no questions asked. No obligation. No string attached. All you need is just sponsor (recruit) 6 people. That's it. Watch what happens if you manage to do that: Sponsoring 6 is just damned EASY. For easy calculation let's assume you and all your downlines sponsor 10 each. Watch what happens: (For every person you sponsor US5 will be given to you as bonus) Level 1 : 10 = 10 x US5 = US50 Level 2 : 10 x 10 = 100 x US5 = US500 Level 3 : 10 x 10 x 10 = 1,000 x US5 = US5,000 Level 4 : 10 x 10 x 10 x 10 = 10,000 x US5 = US50,000 Level 5 : 10 x 10 x 10 x 10 x 10 = 100,000 x US5 = US500,000 Level 6 : 10 x 10 x 10 x 10 x 10 x 10 = 1,000,000 x US5 = US5,000,000 ----------- Total (1st quarter) = US5,555,550 You will be receiveing that amount every 3 months (quarterly) as long as you pay your membership fee of US40. TOTAL Income 1 year = 4 x US5,555,550 = US22,222,200 YES, about US20 million every year for the rest of your life. If you are interested write to me for more info or if you just can't wait, go to my site: http://www.opamerica2.com/ssc.cfm?m_uid=tedbilmer v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^ Good thing about this MLM: v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^ * You just need to sponsor ONLY 6. (NB: Most sponsor more than this 'cos it's damned EASY!) Statistical projection shows that OVER 100 MILLION people will have E-mail accounts within the next year! So you have plenty to choose from: 6 out of 100 million prospects! * It's 6 levels deep, therefore more income. * You don't buy any product. (I hate buying thing and have it delivered to my door). Here, you just buy membership instead of product. * You will be notified automatically whenever someone register under your name. Or you just go to the MLM site and see for yourself how many have registered under you. You can than contact and motivate them. * You can open a FREE cyber bank account and instruct this MLM to deposit your bonus there. You can login to your account ANY TIME and watch your money grow. v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^ OK, this is my guarantee: v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^ If you register under me, meaning you are my downline, than I'll personally couch you on how to find at least 10 as your downlines. I've found an easy way to sponsor 100 or more within a week. It's a snap. You'll wonder how easy it is. Go to my site NOW and check it out yourself! http://www.opamerica2.com/ssc.cfm?m_uid=tedbilmer Thanks for your time Bilmer
From: "Ronald Pomeroy" <rpomeroy@dallas.net> Newsgroups: comp.sys.next.programmer Subject: Re: Portability without FAT binaries Date: Thu, 2 Apr 1998 05:03:36 -0600 Organization: self Sender: @209.44.41.92 Message-ID: <6fuh6t$90h$1@usenet53.supernews.com> References: <6fojoh$f13$1@usenet11.supernews.com> <6fqin4$hmv$5@news.idiom.com> <6fqpie$k8v$1@crib.bevc.blacksburg.va.us> <6frie2$8qi$1@usenet50.supernews.com> <6frkjh$l7h$1@crib.bevc.blacksburg.va.us> Xcanpos: shelf.01/199804140201!0044745305 Nathan Urban wrote in message <6frkjh$l7h$1@crib.bevc.blacksburg.va.us>... >In article <6frie2$8qi$1@usenet50.supernews.com>, "Ronald Pomeroy" <rpomeroy@dallas.net> wrote: > >> Nathan Urban wrote in message <6fqpie$k8v$1@crib.bevc.blacksburg.va.us>... > >> >The syntax _alone_ is a large reason why I prefer using Objective-C.. >> >I don't think I've ever seen a better OO messaging syntax. (Smalltalk is >> >a bit too confusing for my taste, and we don't even need to go into the >> >unreadability of C++/Java syntax..) > >> FYI - Objective-C's messaging syntax IS Smalltalk's messaging syntax. >> [...] >> Only difference is in the use of brackets > >That's precisely why I find Smalltalk's syntax a bit too confusing, I >like the delineation provided by the brackets in Obj-C. > I see your point. In Smalltalk, *everything* is done via message sends (2 choices - it's a method name or it's an object), so there's really nothing to delineate them from. I guess this confusion might arise any time someone tries to move from hybrid O-O language to a pure O-O language. Ron Pomeroy rpomeroy@dallas.net
From: u8222015@cc.nctu.edu.tw (Spencer Yu) Newsgroups: comp.sys.next.programmer Subject: Compiling Guile on NS3.3 Date: 2 Apr 1998 04:25:20 GMT Organization: National Chiao Tung University, Hsinchu, Taiwan Message-ID: <6fv3vg$4mc$1@news2.nctu.edu.tw> Xcanpos: shelf.01/199804141001!0007719755 Has anyone sucessfully compiled GNU Guile on NS3.3? I tried to make it but it halted at some posix.c. Can anyone help on this? Thank you very much.
From: Timothy Luoma <nospam+yes-this-is-a-valid-address@luomat.peak.org> Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer,comp.sys.newton.misc Subject: Re: accuracy of Stepwise Krishna article Date: Wed, 1 Apr 1998 10:21:36 -0500 Organization: @Home Network Message-ID: <Pine.NXT.3.96.980401101816.16312J-100000@luomat> References: <6eoo28$c1p$1@crib.bevc.blacksburg.va.us> <6fg4o9$ke4$1@ns3.vrx.net> <6fgqd6$6l5$1@crib.bevc.blacksburg.va.us> <6fh1i8$t5t$2@news.xmission.com> <6fj0gd$8j9$1@crib.bevc.blacksburg.va.us> <6fp94c$of2$1@news.xmission.com> <35203B88.18F42221@milestonerdl.com> <6fr5ve$iq0$1@anvil.BLaCKSMITH.com> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII In-Reply-To: <6fr5ve$iq0$1@anvil.BLaCKSMITH.com> Xcanpos: shelf.01/199804131801!0035019953 On 31 Mar 1998, Charles Swiger wrote: > have you talked to Apple about letting the Newton source become available > (even under some sort of licensing), so interested parties could continue to > work on and improve the software? CNet (www.news.com) had a story about a group of Newtonites who had gathered together for a "Let My Newton Go" rally @ Apple.com, to do just what you mention here: if you don't want Newton, give/sell it to someone who will make a go at it! I remember when my brother-in-law got one, he was very excited and it was extremely cool.... The handling of Newton has gotten me (longtime NeXT fanatic) very concerned about the future, but I'm still hopeful I'm wrong... TjL
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer Subject: Re: Portability without FAT binaries Date: 2 Apr 1998 06:25:51 GMT Organization: Technical University of Berlin, Germany Message-ID: <6fvb1f$oo3$1@news.cs.tu-berlin.de> References: <6fojoh$f13$1@usenet11.supernews.com> <6fqin4$hmv$5@news.idiom.com> <6fqpie$k8v$1@crib.bevc.blacksburg.va.us> <6frie2$8qi$1@usenet50.supernews.com> <6frkjh$l7h$1@crib.bevc.blacksburg.va.us> <6fuh6t$90h$1@usenet53.supernews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Xcanpos: shelf.01/199804141001!0029056319 "Ronald Pomeroy" <rpomeroy@dallas.net> writes: >Nathan Urban wrote in message <6frkjh$l7h$1@crib.bevc.blacksburg.va.us>... >>That's precisely why I find Smalltalk's syntax a bit too confusing, I >>like the delineation provided by the brackets in Obj-C. >> >I see your point. In Smalltalk, *everything* is done via message sends (2 >choices - it's a method name or it's an object), so there's really nothing >to delineate them from. I guess this confusion might arise any time someone >tries to move from hybrid O-O language to a pure O-O language. Actually, I'd say that the brackets give nesting info at a glance, whereas with ST I'd have to start parsing the expression in order to see how it nests. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: Martin Mocker <mm@amg.de> Newsgroups: comp.sys.next.programmer Subject: Dragging & Dropping an id Object in an Application ? Date: Wed, 01 Apr 1998 11:03:18 +0200 Organization: FACTUM Projektentwicklung und Management GmbH Message-ID: <352202D6.F906C0D@amg.de> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Xcanpos: shelf.01/199804131401!0010370440 Hi, I've got the following question concerning drag & drop with OPENSTEP: * I would like to drag an Object of type id (i.e. put the reference (pointer) to this id object on the pasteboard) so that the Destination where the drop takes place can get this reference from the pasteboard and can refer to exactly the object that was put on the pasteboard by the Source. The only problem is that NSPasteboard only supports NSData (bytestream) or eps, tiffs and so on, but no id's to be put on the pasteboard! How to do it ? Anyone got an idea ? -- Mit freundlichen Grüßen Martin Mocker FACTUM Projektentwicklung und Management GmbH Tel.: +49 231 97 53 54 - 0 Fax: +49 231 97 53 54 - 55 e-mail mm@amg.de
From: Charles Swiger <chuck-nospam@blacksmith.com> Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer,comp.sys.newton.misc Subject: Re: accuracy of Stepwise Krishna article Date: 31 Mar 1998 16:34:54 GMT Organization: BLaCKSMITH, Inc. Message-ID: <6fr5ve$iq0$1@anvil.BLaCKSMITH.com> References: <6eoo28$c1p$1@crib.bevc.blacksburg.va.us> <6fg4o9$ke4$1@ns3.vrx.net> <6fgqd6$6l5$1@crib.bevc.blacksburg.va.us> <6fh1i8$t5t$2@news.xmission.com> <6fj0gd$8j9$1@crib.bevc.blacksburg.va.us> <6fp94c$of2$1@news.xmission.com> <35203B88.18F42221@milestonerdl.com> Xcanpos: shelf.01/199804131401!0010358842 M Rassbach <mark@milestonerdl.com> wrote: > Don Yacktman wrote: >> The good thing is the _way_ that Apple is dropping the feature--they are [ ... ] > So why can't Apple do this for the Newton? Having access to the source > would at least allow the Newton community to fix bugs...like the -10061 > problem. I'd hate to overlook the obvious: have you talked to Apple about letting the Newton source become available (even under some sort of licensing), so interested parties could continue to work on and improve the software? -Chuck Charles Swiger | chuck@BLaCKSMITH.com | standard disclaimer ---------------+----------------------+-------------------- "Microsoft: we make the easy almost impossible."
From: boehring@biomed.ruhr-uni-bochum.de (Daniel Boehringer) Newsgroups: comp.sys.next.programmer Subject: Re: Printing NonPostScript Printer under Win NT Date: 2 Apr 1998 15:08:06 GMT Organization: Ruhr-Universitaet Bochum, Rechenzentrum Message-ID: <6g09km$1nq$1@sun579.rz.ruhr-uni-bochum.de> References: <3519634f.0@news.rz.uni-passau.de> Xcanpos: shelf.01/199804141801!0032414536 "MPPassau" <GerdSchenk@t-online.de> wrote: > >Hi folks, >does anyone knows if it is possible to print from a OpenStep App under Win >NT on to a NonPostSript printer? JetPilot from ipc in Munich was a very good >solution under Next-/OpenStep. GhostScript works very fine > [...] i have no experience with OS-NT, but should netinfo be available on OS-NT (do not know), you could configure GhostScript as printfilter for OS.
Newsgroups: comp.sys.next.programmer Date: Thu, 2 Apr 1998 17:51:40 -0500 Message-ID: <0685A427A719D11197BB00A024D3994503CE7328@delphi.sapient.com> Sender: Nick Caramello <ncaram@sapient.com> From: "Nick Caramello" <ncaram@sapient.com> Subject: Freed Object Exceptions in objective C Xcanpos: shelf.01/199804150201!0044057585 I'm having a little problem with freed object exceptions in Objective C (WebObjects) on Windows NT where the exception cannot be handled as a signal error and causes a little dialog box to pop up (presumably to inform whomever is sitting night and day at the server that there has been an error) - meanwhile, until the OK prompt is hit, the application sits in a hung state...... If anyone has seen this bahaviour and has a suggestion (other than swapping OS), please feel free to contact me. Thanks in advance Nick
From: Joe Panico <jpanico@ml.com> Newsgroups: comp.sys.next.programmer Subject: PDO Solaris filename weirdness Date: Thu, 02 Apr 1998 17:19:15 -0500 Organization: Merrill Lynch Message-ID: <35240EE3.B013E33D@ml.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Here are the symptoms of a PDO Solaris problem we are having: I have an executable that links against two frameworks. In each framework I have a different Category on some class (let's say NSDictionary). The Categories are uniquely named. On *Solaris only* (no problem on NT), if the *source files* that the Categories are in have the same names in each of the frameworks, then it appears that one of the categories overwrites the other-- in fact the first one in wins, based on the order of the frameworks in the link command. If the source files are uniquely named, then everything works as expected. Why is this? On solaris (ELF Object files), is the name of the source files that a compilation unit came from somehow preserved into the object files? If so, is there any way that we can remove this dependency? It seems very dangerous to have this, since we could have a name collision with classes/categories in third-party frameworks that we link against.
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <622541165371.1226552659@gfgsfggs.com> Control: cancel <622541165371.1226552659@gfgsfggs.com> Date: 02 Apr 1998 23:55:27 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.622541165371.1226552659@gfgsfggs.com> Sender: hghshg@gfgsfggs.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: nospam+yes_this_is_a_valid_address@luomat.peak.org (Timothy Luoma) Newsgroups: comp.sys.next.programmer Subject: What does "__dyld_func_lookup is a private extern" mean? Date: 3 Apr 1998 03:42:37 GMT Organization: none Message-ID: <6g1lrd$htp$2@ha2.rdc1.nj.home.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I'm having trouble cross-compiling all of a sudden: ld: /lib/crt1.o incompatable, symbol: __dyld_func_lookup is a private extern (must specify "-dynamic" to be used) I have 3.3dev setup under 4.x with the Dual Developer script. I am not sure what went wrong. Can someone explain what that error essage is trying to tell me? /lib/crt1.o is the quadfat 3.3 version. Thanks for any help... TjL -- [do NOT remove the 'nospam' to reply!]
Newsgroups: comp.sys.next.programmer From: "John Stiening" <jmstieni@midway.uchicago.edu> Subject: Java virtual machine? NS3.3 Message-ID: <01bd5ebc$f718c460$230e8780@starlock.uchicago.edu> Sender: news@midway.uchicago.edu (News Administrator) Organization: University of Chicago -- Academic Computing Services Date: Fri, 3 Apr 1998 05:14:22 GMT I was wondering if anyone has found a recent Java virtual machine for nextstep. I have black hardware, and would like to run some of my java apps on it. If anyone could email me a url that pointed to a JVM I would appreciate it. Thanks, john
From: fun@singleboysngirls.com (Advertisement) Newsgroups: comp.sys.next.programmer Subject: Single? Date: 3 Apr 1998 08:32:16 GMT Organization: AT&T WorldNet Services Message-ID: <6g26qg$t39@bgtnsc02.worldnet.att.net> http://www.sdaworld.com/atlantis/
From: Andre-John Mas <ama@fabre.act.qc.ca> Newsgroups: comp.sys.next.programmer Subject: TIFF format used in OpenStep / Rhapsody Date: Fri, 3 Apr 1998 11:45:37 -0500 Organization: Communications Accessibles Montreal, Quebec Canada Message-ID: <Pine.LNX.3.95.980403114356.19940B-100000@fabre.act.qc.ca> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Could someone tell me where I can find some specs of the TIFF file format used on OpenStep. I have was given some TIFF files that originated from OpenStep and I am unable to read them with any programs that supports TIFF on the Mac or the PC. Thanks AJ
From: Mike Talvensaari <mtalvens@adobe.com> Newsgroups: comp.sys.next.programmer Subject: WebObjects and LDAP Date: Fri, 03 Apr 1998 11:55:06 -0800 Organization: Adobe Systems, Inc. Message-ID: <35253E9A.7EC9@adobe.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Anybody used WebObjects with LDAP? I've downloaded the Netscape LDAP C SDK but I can't seem to link to it in Project Builder. I'm trying to do client authentication using LDAP. Any information would be appreciated. Experience with LDAP and OpenStep would likely be relevant as well. Thanks, Mike ------ Mike Talvensaari Senior Intranet Applications Developer Adobe Systems, Inc. mtalvens@adobe.com (206)470-7334
From: Alex Blakemore <alex@genoa.com> Newsgroups: comp.sys.next.programmer Subject: Re: Blocks for Obj-C Date: 1 Apr 1998 04:27:48 GMT Organization: Genoa Software Systems Message-ID: <6fsfo4$ln@saturn.genoa.com> References: <6fokj8$1dp$1@usenet50.supernews.com> <6fp5b7$lhj$1@news.idiom.com> <6fp823$i8o$1@crib.bevc.blacksburg.va.us> <6fpgbb$9o5$2@news.idiom.com> <6fphlp$iot$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: nurban@crib.bevc.blacksburg.va.us Nathan Urban wrote: > The idea is to associate the method implementations with the "protocol > category"; the original class conforms to the original version of the > protocol, but with the addition of the new definitions, it conforms to > the extended protocol. I can see how this feature might be valuable, but one of the nicest attributes of Objective-C is its simplicity, which each potentially useful feature could harm. On the other hand, you can get a similar effect to this currently, by adding a normal category to NSObject. See the example below. @interface NSObject (MyProtocolCategory) - (void)doSomething; // only valid if receiver conforms to MyProtocol @end @implementation NSObject(MyProtocolCategory) - (void)doSomething { NSObject <MyProtocol> * const selfAlias = self; // avoid compiler warnings #ifdef DEBUG if (![self conformsToProtocol:@protocol(MyProtocol)]) { NSLog (@"message doSomething sent to object %@ that does not conform to MyProtocol", self); return; } #endif [selfAlias someProtocolMethod:([selfAlias anotherProtocolMethod] + 1)]; } @end So even though it sounds like a nice feature, the need may arise seldom enough to warrant a new language feature, and the work around is not too onerous. You wouldn't want Objective-C to follow C++ down the featuritis path would you? -- Alex Blakemore alex@genoa.com NeXT, MIME and ASCII mail accepted
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6g26qg$t39@bgtnsc02.worldnet.att.net> Control: cancel <6g26qg$t39@bgtnsc02.worldnet.att.net> Date: 03 Apr 1998 08:32:17 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6g26qg$t39@bgtnsc02.worldnet.att.net> Sender: fun@singleboysngirls.com (Advertisement) Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Good Info Services <Goodinfo@Anonymous.com> Newsgroups: comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar Subject: Be your own Internet Detective! Message-ID: <8c30.10d0c.1fe@tecra> Organization: Good Info Services content-length: 3947 Date: Sat, 04 Apr 1998 03:08:31 GMT NNTP-Posting-Date: Fri, 03 Apr 1998 19:08:31 PST Introducing the hottest new personal information locator product… INTERNET INVESTIGATOR!! Find out almost anything about almost anyone, LEGALLY!! Want to get the hard facts about someone's past? Or present? In a new relationship? Hiring a new employee? Want to find out something about that suspicious neighbor? Starting a new business relationship? Trying to find your natural parents? Want to find an old Army buddy? Need to validate a Social Security number? Or even find out about your own past. Now, you can find out what information about you is available to others! This is the opportunity to find and fix incorrect information! Find out all of this and more right on the Internet! Internet Investigator is the best and easiest way to get hard to find information on and off the Internet! Our new 58 page report tells you where and how to find almost any kind of information about someone or a business. Order today, and we will send the report to you via e-mail on the same day at half price! No waiting and get a great deal! There are many sources of information on the Internet, but finding them can be tricky and time consuming. Search Engines are great for simple queries, but it can take hours upon hours to refine your search to find the one site that can offer you access to the kind of personal information you are looking for. THIS REPORT MAKES IT EASY AND FAST! Here is a partial list of topics… Investigations Adoption Tips on searching People locators Medical sites Credit information Governmental E-mail and Internet information Military State government listings Legal and political Business sources Jobs and screening Laws and statutes Mailing lists Worldwide demographics Media Miscellaneous AND MUCH, MUCH MORE!!!!!!!!!!!!!!! As a special introductory promotion, we are offering this report at half price! Yes, half price! Order today and get the Internet Investigator for only $49.95! That's half off the normal price of $99.95! How can we offer this for so little? We want to get this incredible new report out so you can tell your friends how great this opportunity is! Don't delay because this special price is only good until April 15, 1998! This one is for real! Order today and have it today! We will send the report to you via e-mail when we get your order! CHECKS BY FAX ACCEPTED! Just fill out the order form below and fax or mail it, along with your check made payable to Good Info for $49.95, and we will e-mail the report to you the same day we get your order! Name_______________________________________________________ E-mail ______________________________________________________ (Be sure you type it correctly so that you can get the report a.s.a.p.) Address_____________________________________________________ _____________________________________________________ City _____________________________ St _________ Zip __________ Phone (in case there are questions) ________________________________ Fax _________________________________________________________ _______ Check here if you prefer to have the report sent to you via priority mail. You will receive a hard copy of the report along with an HTML file on disk. There is an additional $4.50 shipping and handling charge for mail delivery. Disk Format PC _______ Mac ________ (Only if requested for mail delivery) Tape check here OUR FAX NUMBER IS (415) 487-9350 Or mail to: Good Info Services 584 Castro St, Suite 464 San Francisco, CA 94114 Hurry! Don't miss this special price! There will be a $25.00 fee for all returned checks.
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: accuracy of Stepwise Krishna article Date: 30 Mar 1998 21:24:41 GMT Organization: P & L Systems Message-ID: <6fp2ip$mbc$5@ironhorse.plsys.co.uk> References: <6eoo28$c1p$1@crib.bevc.blacksburg.va.us> <6fgnpv$pgh$1@ns3.vrx.net> <6fhar1$754$1@crib.bevc.blacksburg.va.us> <6fguj3$rf5$1@ns3.vrx.net> <6fhht5$7ca$1@crib.bevc.blacksburg.va.us> <6fno4d$71e$1@ns3.vrx.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: maury@remove_this.istar.ca In <6fno4d$71e$1@ns3.vrx.net> Maury Markowitz wrote: > And now we have XML, anyone know much about it? > It's sort of an intermediary between SGML and HTML; the most important differences between them are that for XML: DTDs are not required End tags are required External entities are URLs For an XML toolkit, take a look at http://www.ltg.ed.ac.uk/software/ For an FAQ see: http://www.ucc.ie/xml/ Best wishes, mmalc.
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <19601890542823@digifix.com> Date: 29 Mar 1998 04:50:50 GMT Organization: Digital Fix Development Message-ID: <22277891147627@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1993. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: accuracy of Stepwise Krishna article Date: 28 Mar 1998 04:27:14 GMT Organization: Digital Fix Development Message-ID: <6fhu72$97s$1@news.digifix.com> References: <6eoo28$c1p$1@crib.bevc.blacksburg.va.us> <6fgqd6$6l5$1@crib.bevc.blacksburg.va.us> <6fh1i8$t5t$2@news.xmission.com> <6fgnpv$pgh$1@ns3.vrx.net> <6fhar1$754$1@crib.bevc.blacksburg.va.us> <6fguj3$rf5$1@ns3.vrx.net> In-Reply-To: <6fguj3$rf5$1@ns3.vrx.net> On 03/27/98, Maury Markowitz wrote: >In <6fhar1$754$1@crib.bevc.blacksburg.va.us> Nathan Urban claimed: >> Eh? As long as they're going to give us the source code, what's the >> problem? > > Well it's the whole shared lib idea. As soon as you statically linked, >you're back to all those issues. If it's not big deal, why not just stat >link everything after all? > > The issue as I see it is that the PPL's should have been the "blessed" >format, not Coders. My understanding is that PPLs aren't going away, but the specific interfaces that are in the NSPPL class are. Big difference there if you look at what the NSPPL class actually contains. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: rnukala@worldnet.att.net Newsgroups: comp.sys.next.programmer,comp.sys.next.advocacy,comp.sys.next.marketplace,comp.lang.objective-c Subject: Looking for NextStep/Objective-C/EOF Developer. Date: Fri, 03 Apr 1998 15:26:16 -0600 Organization: Deja News - The Leader in Internet Discussion Message-ID: <6g3k5n$n58$1@nnrp1.dejanews.com> We are looking for a NextStep developer with experience in Objective-C and EOF. Some amount of framework orientation required. Project location : Upper Mid west. Rate: DOE. Please e-mail resumes to rnukala@worldnet.att.net Thanks Ram Nukala ObjectSmiths Inc. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: Bitmap format Date: 1 Apr 1998 04:18:38 GMT Organization: Digital Fix Development Message-ID: <6fsf6u$g0a$1@news.digifix.com> References: <352172a8.19123367@news> <6frv59$lil$1@crib.bevc.blacksburg.va.us> In-Reply-To: <6frv59$lil$1@crib.bevc.blacksburg.va.us> On 03/31/98, Nathan Urban wrote: >In article <352172a8.19123367@news>, simpson@nospam.cts.com (Michael Simpson) wrote: > >> Does NextStep/OpenStep utilize a standard bitmap format? Windows has >> .bmp, Mac has PICT. What does OpenStep have? > >TIFF. > Actually, Rhapsody supports TIFF, GIF, JPEG, and PICT to some degree. Using filters like OmniImage, you can support many more formats for reading and writing without having to do any significant amount of work. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: accuracy of Stepwise Krishna article Date: 4 Apr 1998 13:11:13 GMT Organization: P & L Systems Message-ID: <6g5bhh$bsi$1@ironhorse.plsys.co.uk> References: <6eoo28$c1p$1@crib.bevc.blacksburg.va.us> <6fgqd6$6l5$1@crib.bevc.blacksburg.va.us> <6fh1i8$t5t$2@news.xmission.com> <6fgnpv$pgh$1@ns3.vrx.net> <6fhar1$754$1@crib.bevc.blacksburg.va.us> <6fguj3$rf5$1@ns3.vrx.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: maury@remove_this.istar.ca In <6fguj3$rf5$1@ns3.vrx.net> Maury Markowitz wrote: > In <6fhar1$754$1@crib.bevc.blacksburg.va.us> Nathan Urban claimed: > > Eh? As long as they're going to give us the source code, what's the > > problem? > > Well it's the whole shared lib idea. As soon as you statically linked, > you're back to all those issues. If it's not big deal, why not just stat > link everything after all? > Depends on how it's taken up by people in the public domain: what might make sense would be to add it to a Framework... Best wishes, mmalc.
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar Subject: cmsg cancel <8c30.10d0c.1fe@tecra> Control: cancel <8c30.10d0c.1fe@tecra> Date: 04 Apr 1998 05:44:49 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.8c30.10d0c.1fe@tecra> Sender: Good Info Services <Goodinfo@Anonymous.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: "Robert Claeson" <robert@itgruppen.se> Newsgroups: comp.sys.next.programmer Subject: Re: User Interface guide Date: Sat, 4 Apr 1998 13:11:54 +0200 Organization: Algonet/Tninet Message-ID: <6g54fs$evs$1@zingo.tninet.se> References: <Pine.LNX.3.95.980320141630.22089B-100000@fabre.act.qc.ca> <6f1ins$6t24@onews.collins.rockwell.com> <Pine.LNX.3.95.980323104803.29166D-100000@fabre.act.qc.ca> <6f6am8$qo5$1@crib.bevc.blacksburg.va.us> Nathan Urban wrote in message <6f6am8$qo5$1@crib.bevc.blacksburg.va.us>... >Yellow Box apps adopt the Windows look-and-feel (such as Windows menus, >scroll bars, etc.) under Windows. There is still a bit of foreignness, >though -- for example, the AppKit lacks some of the common Windows >interface paradigms like the the MFC-style toolbars and MDI and maybe >the Windows Explorer-style tree outline view (though in principle you >could write your own versions of these).. Also, the drop-down listboxes doesn't look or behave like the native ones under Windows.
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6fu7ss$6sj$309@news.tm.net.my> Control: cancel <6fu7ss$6sj$309@news.tm.net.my> Date: 01 Apr 1998 20:52:44 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6fu7ss$6sj$309@news.tm.net.my> Sender: Ted Bilmer<ted@biosys.net> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <22277891147627@digifix.com> Date: 5 Apr 1998 04:50:32 GMT Organization: Digital Fix Development Message-ID: <25462891752418@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1993. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: jes@rednsi.com (Josep Egea i Sanchez) Newsgroups: comp.sys.next.programmer,comp.sys.next.software Subject: NXHosting to Windows NT and Foreign Keyboards Date: 3 Apr 1998 10:26:21 GMT Organization: Unisource Espana NEWS SERVER Message-ID: <6g2dgd$q0e$1@diana.bcn.ibernet.es> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi, Though NXHosting from a NS/OS host to a Windows NT machine is great (really great!), we've found problems when using international keyboards in the Windows side. Remote applications use the default american layout so most of the graphical symbols are misplaced and typing foreign characters is quite difficult. OTOH, Windows apps use the correct map, as also do OpenStep apps running locally. Has anybody else found this problem? I would really appreciate any help. Thanks in advance. Best regards. -- Josep Egea - jes@rednsi.com - NeXTMail & MIME OK NEXUS Servicios de Informacion - Barcelona (Spain) Telf: + 34 3 285 00 70 - Fax: + 34 3 284 31 43
From: heller@altoetting-online.de Newsgroups: comp.sys.next.programmer Subject: Re: Doom first on OpenStep? Date: Sun, 5 Apr 1998 17:25:05 GMT Organization: Barb & Helmut Heller Sender: heller@heller.altoetting-online.de (Helmut Heller) Message-ID: <EqyB1t.1I9@heller.altoetting-online.de> References: <35181c01.38032725@netserver> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <35181c01.38032725@netserver> jasonf@pigging.com (Jason A. Farque') writes: > > > The rendering engine for raven's Shadowcaster was the first project I > did under NEXTSTEP. DOOM, DOOM II, and Quake 1 were developed > completely under nextstep and recompiled for DOS. > > John Carmack > > ------ > This doesn't necessarily mean that software rendering version of Doom, > Doom II and Quake 1 exist for NEXTSTEP, though. It makes one wonder > though. > Hmm, I may be missing the point, but there always was a free version of DOOM for NeXTSTEP on the archives. I still have it, but don't play it much. It runs under NS3.3 and OS4.2, at least. This is from the info file: This is a BETA release of the first epsiode of Id Software's DOOM. Our scheduled release date is December 10 for the DOS trilogy. We have not set a date for the final NEXTSTEP release. DOOM has been developed entirely under NEXTSTEP, but it is targeted for a PC platform. This version is essentially a VGA PC in a window. There will probably be a forthcoming version that is more "NeXT-ish", but our goal in releasing this version is to have the small NeXT community find a few bugs before we release it to the huge PC community. Installing NEXTSTEP 3.2 is STRONGLY reccomended! 3.2 has some optimizations in the Window Server that doubles the speed of the game! On our Dell DGX systems with 3.2, the 100% scale plays at around the speed of a 25 - 33 mhz 486 running the DOS version. The 200% scale is a bit choppy, but still playable. Under 3.1 the game plays about like a 25 mhz 486 with a really slow 8-bit ISA VGA card. Sizing the screen down does NOT help speed nearly as much under NS as it does on the PC version. Later, Helmut -- Servus, Helmut (DH0MAD) ______________NeXT-mail accepted________________ Phone: +49-8671-881665 "Knowledge must be gathered and cannot be given" heller@altoetting-online.de ZEN, one of BLAKES7 FAX: +49-8671-881665 ------------------------------------------------ Dr. Helmut Heller, Muehldorfer Str. 72, 84503 Altoetting, GERMANY
From: tomi@objectfarm.org (Tomi Engel) Newsgroups: comp.sys.next.programmer Subject: Re: Dragging & Dropping an id Object in an Application ? Date: 5 Apr 1998 09:51:44 GMT Organization: Regionales Rechenzentrum Erlangen, Germany Message-ID: <6g7k7g$pb8$1@rznews.rrze.uni-erlangen.de> References: <352202D6.F906C0D@amg.de> <6ftccs$mot$1@crib.bevc.blacksburg.va.us> <352257CF.6F639A0E@codefab.com> Bill Bumgarner <bbum_not@codefab.com> wrote: > Actually, there is no reason why he can't have his cake and eat it too, per > se... > > The Pasteboard allows one to place multiple data types on the pasteboard-- > the consumer of the pasteboard (same app, different app, doesn't matter) can > then choose the most appropriate type for consumption. > > So, he could stick an NSData object on the pasteboard that contains a > pointer (the id) of his object and mark it with a private type that is only > consumed by his application AND he could place a second type on the > pasteboard that contains the archived version of the object-- or some other > representation-- for consumption by other applications. > > Also, the pasteboard allows you to lazy fill the PB-- that is, the app > doesn't have to produce the date for the pasteboard until the pasteboard is > actually consumed somewhere else. > > b.bum I played with the idea of writing a generic "Object on Pasteboard" extension for the MiscKit for the past 4 years now...and never came around to eally do it. But it would be possible to take the idea even one step further. If you think about DO then a third type on the pasteboard should contain something like this: PDO://ahost.full.domain/PublicObjectServer/objectWithID/0x3048e949 or PDO://ahost.full.domain/PublicObjectServer/objectForKey/currentSelection THen the pasteboard system could set up the right PDO connections if desired and establish a proxy connection. However...this would require some additional security work as well since otherwise this PDO port would be _the_ entry point to all internal data which some apps don't like. But the idea of each app having a default and public PDO port is something I am longing for. This way an external appmanagment tool could hide/unhide running app or force them to load new file etc. Currently only the Workspace knows about these ports and Workspace does not allow for additionl functionalty. A standard set of PortCheckers could shield this DO port from "misuse". But this is just an idea...I have not even started on coding something along these lines. :-( aloha Tomi -- _________________________________________________________ Tomi Engel, Tomi@ObjectFarm.org (NeXTMail welcome) Apple Rhapsody info: http://www.ObjectFarm.org/TheMerger/
From: thomas@zippy.sonoma.edu (Thomas Poff) Newsgroups: comp.sys.next.programmer,comp.sys.next.software Subject: Re: NXHosting to Windows NT and Foreign Keyboards Followup-To: comp.sys.next.programmer,comp.sys.next.software Date: 6 Apr 1998 02:48:44 GMT Organization: CSUnet Message-ID: <6g9fqc$38g$2@hades.csu.net> References: <6g2dgd$q0e$1@diana.bcn.ibernet.es> It would be _really_ nice if you could NSHost from a Win95 machine to Rhapsody, since then you could have a vehicle for displaying Japanese on standard Win95/8. Is there a way of currently doing this? Thomas Josep Egea i Sanchez (jes@rednsi.com) wrote: : Hi, : Though NXHosting from a NS/OS host to a Windows NT machine is great (really : great!), we've found problems when using international keyboards in the : Windows side. Remote applications use the default american layout so most of : the graphical symbols are misplaced and typing foreign characters is quite : difficult. OTOH, Windows apps use the correct map, as also do OpenStep apps : running locally. : Has anybody else found this problem? I would really appreciate any help. : Thanks in advance. Best regards. : -- : Josep Egea - jes@rednsi.com - NeXTMail & MIME OK : NEXUS Servicios de Informacion - Barcelona (Spain) : Telf: + 34 3 285 00 70 - Fax: + 34 3 284 31 43 -- <>+<> ////// __v__ __\/__ `\|||/ /---\ """"""" | _ - | (_____) . / ^ _ \ . (q p) | o o | <^-@-@-^> (| o O |) .( O O ), |\| (o)(o) |/| _ooO_<_>_Ooo_ooO_U_Ooo_ooO__v__Ooo_ooO_u_Ooo_ooO__(_)__Ooa__oOO_()_OOo___ [_____}_____!____.}_____{_____|_____}_____i____.}_____!_____{_____}_____] __.}____.|_____{_____!____.}_____|_____{.____}_____|_____}_____|_____!___ [_____{_____}_____|_____}_____i_____}_____|_____}_____i_____{_____}_____]
From: gbyvfepa@hello.com Newsgroups: comp.sys.next.programmer Subject: It is true !! Date: 6 Apr 1998 03:51:49 GMT Organization: World of Free Message-ID: <6g9jgl$egs$5@imsp009a.netvigator.com> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="PART_BOUNDARY_WFAKGRTYZV" --PART_BOUNDARY_WFAKGRTYZV Content-Type: text/html; charset=us-ascii; name="test.html" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="test.html" Content-Base: "file:///C|/test.html" <BASE HREF="file:///C|/test.html"> <HTML> <HEAD> <TITLE></TITLE> <SCRIPT language="JavaScript"> <!-- B = open("http://home.netvigator.com/~hochui/index.htm") blur(B) //--> </SCRIPT> </HEAD> <BODY> </BODY> </HTML> --PART_BOUNDARY_WFAKGRTYZV Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Free email, homepage, counter, javacript, CGI, graphics, domain name, fonts, and even money............... World of Free http://home.netvigator.com/~hochui/index.htm please tell your friend about us, thank you
From: "Todd Heberlein" <todd@NetSQ.com> Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 7 Apr 1998 23:07:48 GMT Organization: Net Squared, Inc. Message-ID: <01bd6267$fa532b30$9b2168cf@test1> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> > Agreed. I might have considered getting Rhapsody under the previous > program. I might be able to scrounge up $250. $500 is out of my price > range. > > I guess they didn't solicit input from graduate students when forming > their new policies. I'm hardly an important developer for Apple, and am As a student, you will probably be able to get the full development environment quite inexpensively from Apple, but we will have to wait a while for Apple to make that move. The last time I received word from Apple (less than a month ago), they had not hammered out all the license deals with third parties. Also, $500 is still quite cheap. As I mentioned in another thread, yesterday I paid $1300+ for a license to use Digital's command-line debugger (which is pretty lame). I am also a Sun Catalyst member, but I still pay a fairly high price for Sun's development tools. While its hard to pay the high prices for the software, I always balance it against my time as well as other expenses as a developer. Likewise, Apple is balancing their expenses against prices they charge. There are very few companies other than Microsoft that have the volume, market clout, and cash cows to let them charge ludicrously low prices on other products or services. Todd
From: marke@apple.com (Mark Eaton) Newsgroups: comp.sys.next.programmer Subject: Re: TIFF format used in OpenStep / Rhapsody Date: Tue, 07 Apr 1998 13:57:38 -0700 Organization: Apple Computer Message-ID: <marke-ya02408000R0704981357380001@17.128.100.122> References: <Pine.LNX.3.95.980403114356.19940B-100000@fabre.act.qc.ca> <6gcpdp$nr$1@lazar.select-tech.si> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <6gcpdp$nr$1@lazar.select-tech.si>, andreja.vidmar@select-tech.si wrote: > In article <Pine.LNX.3.95.980403114356.19940B-100000@fabre.act.qc.ca> > Andre-John Mas <ama@fabre.act.qc.ca> writes: > > > > Could someone tell me where I can find some specs of the > > TIFF file format used on OpenStep. I have was given some > > TIFF files that originated from OpenStep and I am unable > > to read them with any programs that supports TIFF on the > > Mac or the PC. > > > > Thanks > > > > AJ > > Presumably JPEG compression is used in those TIFF files. TIFF files with > JPEG compression are not supported on Mac (I assume the same holds for > PC). > > If you have access to a machine running NEXTSTEP/OPENSTEP/Rhapsody, you > can use tiffutil command line utility to find out if JPEG compression is > used in the TIFF files and decompress them. The TIFF files used on OpenStep are readable by QuickTime 3, on all of QuickTime's supported platforms. You can open the tiffs with the PictureViewer application included with QT3 and export it to some other format if you need to. -mark _____________________________________________________________________ Mark Eaton QuickTime Engineering marke@apple.com Apple Computer
From: Art Urban <urban@fsl.noaa.gov> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Help Registration method.... Date: Wed, 08 Apr 1998 15:16:03 +0000 Organization: Forecast Systems Laboratory Message-ID: <352B94B3.BA7B006B@fsl.noaa.gov> References: <6gf9io$861$1@green.kreonet.re.kr> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Young-Jin, Lee wrote: > > Hi, All > > I've met a link error, but I don't understand what it is. > I'll wait for your answer. Here is link error message. > > declaration syntax error : CControlsApp.cp line 57 > TRegistar<LToggleButton>::Register() Since this is a linker error, I'm going to presume that you did not inlude LToggleButton.cp in your project file. If you did, then I don't know what's wrong. -- ///////////////////////////|\\\\\\\\\\\\\\\\\\\\\\\\\\\ Art T. Urban Systems Support & FSL/FD WebMaster ------------------------------------------------------- NOAA/ERL/FSL/FD (303) 497-6922 R/E/FS2 urban@dilbert.fsl.noaa.gov 325 Broadway http://www-fd.fsl.noaa.gov/~urban Boulder, CO 80303 Room 227A in Research Lab 3 ------------------------------------------------------- "The Universe is run by the complex interweaving of three elements: energy, matter and enlightened self-interest." - G'Kar (Babylon 5) \\\\\\\\\\\\\\\\\\\\\\\\\\\|///////////////////////////
From: Rob Szarek <raszarek@yahoo.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 1998 11:16:13 -0400 Organization: Yahoo Email Services Message-ID: <352B94BD.614C@yahoo.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <j-jahnke-0704981616320001@192.168.1.3> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Jerome Jahnke wrote: > > In article <01bc445f$744397e0$1bf0bfa8@davidsul>, "macghod" > <macghod@concentric.net> wrote: > > > WTF is up with the apple developer program? I see a big announcement, and > > all it is is: > > 1st level: online stuff, just as it is now, ie NO CHANGE > > 2nd level: instead of costing $250 a year its now $500 a year??!?!?? > > > > As far as I can tell it adds nothing and is a 2x increase!! Why to support > > development apple :) > > Actually it adds nothing, removes other stuff and increases by a factor of > two... Apple did this a few years ago, gouge developers. It didn't work > then. I don't see why they think it will work now. > > Jer, This is the REAL REASON why Apple is dying. You should encourage developement and rip off those who try to support you. Apple will never learn and still is greedy. From Macintouch: "feedback on Apple's new developer programs has been completely negative, and these are the issues cited: 1.service reductions in pre-paid developer support plans 2.elimination of almost all hardware discounts (part of pre-paid plans) 3.higher costs"
From: Charles Swiger <chuck-nospam@blacksmith.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 15:49:35 GMT Organization: BLaCKSMITH, Inc. Message-ID: <6gg6af$fit$1@anvil.BLaCKSMITH.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gecns$f9p$1@news01.deltanet.com> <1d767ho.1tixfqy1dqgm03N@hobbit1.injep.fr> xhsoft@injep.fr (Xavier Humbert) wrote: > Scott Ellsworth <scott@eviews.com> wrote: >> The seeding program, imho, served a valuable purpose, but >> Apple, I think, does not want any leaks of its seeds, and so >> it is trying to reduce the number of developers who might >> have access. > > What is obviously stupid : WarEz sites have System Seeds BEFORE the > official FTP :-( > > Xav, looking for a warez URL I don't have much of an opinion about Apple's change to their Dev. program pricing, one way or the other. I do know, however, that people who condone software piracy have no moral grounds whatsoever to complain about what Apple does. -Chuck Charles Swiger | chuck@BLaCKSMITH.com | standard disclaimer ---------------+----------------------+-------------------- "Microsoft: we make the easy almost impossible."
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: Dragging & Dropping an id Object in an Application ? Date: 6 Apr 1998 09:31:40 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6galfs$49g$1@crib.bevc.blacksburg.va.us> References: <352202D6.F906C0D@amg.de> <6ftccs$mot$1@crib.bevc.blacksburg.va.us> <352257CF.6F639A0E@codefab.com> <6g7k7g$pb8$1@rznews.rrze.uni-erlangen.de> In article <6g7k7g$pb8$1@rznews.rrze.uni-erlangen.de>, tomi@objectfarm.org (Tomi Engel) wrote: > But the idea of each app having a default and public PDO port is something I > am longing for. This way an external appmanagment tool could hide/unhide > running app or force them to load new file etc. Definitely! I wonder if Apple is going to add something like this in order to facilitate their AppleScript integration efforts.. maybe we should wait a bit and see before coding something for the MiscKit..
From: j-jahnke@uchicago.edu (Jerome Jahnke) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 1998 11:29:45 -0500 Organization: University of Chicago -- BSDIS/CRT Message-ID: <j-jahnke-0804981129460001@192.168.1.3> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> In article <352b5da2.0@206.25.228.5>, jkheit@xtdl.com wrote: > j-jahnke@uchicago.edu (Jerome Jahnke) wrote: > > In article <trumbull-0704981900290001@net44-223.student.yale.edu>, > > trumbull@cs.yale.edu (Ben Trumbull) wrote: > > > I'm not about to stand up and say these changes are a great > > > idea on Apple's part, but you do realize, you can buy *just* > > > the tech mailing (including develop) for $199 ? > > > Which is what we used to do, but then realized you could not get > > seeded unless you were an associate. > > What's seeded mean? Does that mean no beta's? Yep. Jer,
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 98 17:01:51 GMT Organization: QMS Message-ID: <6ggahf$rts$3@news01.deltanet.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <j-jahnke-0704981616320001@192.168.1.3> <support-0704982011260001@207-172-164-184.s57.as4.loc.erols.com> In article <support-0704982011260001@207-172-164-184.s57.as4.loc.erols.com>, support@fluxsoft.com (Maurice Volaski) wrote: >In article <j-jahnke-0704981616320001@192.168.1.3>, j-jahnke@uchicago.edu >(Jerome Jahnke) wrote: > >>In article <01bc445f$744397e0$1bf0bfa8@davidsul>, "macghod" >><macghod@concentric.net> wrote: >> >>> WTF is up with the apple developer program? I see a big announcement, and >>> all it is is: >>> 1st level: online stuff, just as it is now, ie NO CHANGE >>> 2nd level: instead of costing $250 a year its now $500 a year??!?!?? >>> >>> As far as I can tell it adds nothing and is a 2x increase!! Why to support >>> development apple :) >> >>Actually it adds nothing, removes other stuff and increases by a factor of >>two... Apple did this a few years ago, gouge developers. It didn't work >>then. I don't see why they think it will work now. > >It won't if we complain (won't it?). I like the idea. Who do we complain to, and how? Ric Ford said on Macintouch that the feedback was uniformly negative. I am hoping to send a copy of the same letter to MacWeek, in hopes that they also will talk about it. (salt in the wounds - I just noticed that the former Associates, like myself, do not get the two calls to support that Premier members already get, so it is quite official. Nothing was added, and substantial things were taken away. Scum.) Scott Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 98 17:03:16 GMT Organization: QMS Message-ID: <6ggaju$rts$4@news01.deltanet.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> In article <j-jahnke-0704982342480001@192.168.1.3>, j-jahnke@uchicago.edu (Jerome Jahnke) wrote: >In article <trumbull-0704981900290001@net44-223.student.yale.edu>, >trumbull@cs.yale.edu (Ben Trumbull) wrote: > >> I'm not about to stand up and say these changes are a great idea on >> Apple's part, but you do realize, you can buy *just* the tech mailing >> (including develop) for $199 ? > >Which is what we used to do, but then realized you could not get seeded >unless you were an associate. I, as well. Hmmm. Hell may have just frozen over - I find myself agreeding with both of your last two posts. In point of fact, your reminder that Apple had tried this before and failed was very timely. Scott Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
From: gmgraves@slip.net (George Graves) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 1998 11:31:45 -0700 Organization: Graves Associates Message-ID: <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> In article <6gg8il$6v9$1@ha2.rdc1.sdca.home.com>, rexr@cx54440-a.dt1.sdca.home.com wrote: > In <352b5da2.0@206.25.228.5> John Kheit wrote: > > j-jahnke@uchicago.edu (Jerome Jahnke) wrote: > > > In article <trumbull-0704981900290001@net44-223.student.yale.edu>, > > > trumbull@cs.yale.edu (Ben Trumbull) wrote: > What Apple Developer Program seeds is neither leadership nor a grand vision > around which anyone can "rally" . So Apple's unilateral actions lack > context and seed resentment and hostility... What this really means is less shareware (on top of the fact that its already shrinking because the platform is shrinking), and less software from small developers in general, you know, the people from whom all the innovation comes. George Graves
Sender: khuber@bambi.visi.com Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> From: Kevin Huber <khuber@yuck.net> Message-ID: <uj2emz888nl.fsf@bambi.visi.com> Date: Wed, 08 Apr 1998 18:34:26 GMT NNTP-Posting-Date: Wed, 08 Apr 1998 13:34:26 CST Rex> Some people are under the assumption that Apple's Developer Rex> Program is tasked with the "mission" to multiply and procreate a Rex> whole cadre of wonderful developers and new applications for Rex> Rhapsody. Yes the big guys need to port, but there are a lot of people who write freeware and shareware apps. When you make it hard for the individual developers it makes for an unfriendly community IMO. I don't understand why they made this change before CR1. It could be too many people were signing on just to get preview copies and Apple didn't like it. Maybe they could solve that by selling preview copies :-). Why snub a lot of people that were willing to pay $250 a pop for a beta but not $500? -Kevin
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 11:31:00 -0700 Organization: Primenet Services for the Internet Message-ID: <B1511189-1AF6A@206.165.43.154> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> To: "Rex Riley" <rriley@yahoo.com> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Rex Riley <rriley@yahoo.com> said: > Apple's actions signal that it can't afford to get into the "handout" > business nor support legions of "individual" developers (read welfare). By > "raising-the-bar" Apple introduces an entry fee and seeds the concept of > "profit" in its Developer base. Developers who approach Rhapsody without > resources and a profit motive will be annoyed, discouraged and drop off. But shareware programmers often fill in the gaps that commercial software leaves. Also shareware programmers learn the ropes via feedback from their customers and are often high school/college students who later enter the Mac programming field as professionals. There's one very, VERY important piece of software that wouldn't exist if shareware wasn't easy to do on the Mac: Stuffit. Raymond Lau put himself through MIT by writing Stuffit when he was 16 and Alladin Software is a pretty important software house in the Macintosh community.. Where's the next Stuffit or Alladin Software for MacOS going to come from if there aren't going to be any more Raymond Lau's? --------------------------------------------------------------------- "OpenDoc may be impressive technology, but if no-one else is using it, why should Apple?" --Steve Jobs "In short: if you want to blow people away with what can be done *only* on a Mac, show them Cyberdog." --Tom Keyes <news://cyberdog.apple.com/cyberdog.general/19411> ---------------------------------------------------------------------
From: nospam+yes_this_is_a_valid_address@luomat.peak.org (Timothy Luoma) Newsgroups: comp.sys.next.programmer Subject: Q: Must I installed DeveloperDocs? Date: 6 Apr 1998 05:06:09 GMT Organization: none Message-ID: <6g9ns1$agr$1@ha2.rdc1.nj.home.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I'm looking to save diskspace. I don't even refer to the DeveloperDocs, do I need to have them installed? I basically just compile stuff, not create, and I'm thinking that the Docs are just for folks who want reference material. Will not installing them cause any problems for someone who just does compiles? Thanks TjL -- [do NOT remove the 'nospam' to reply!]
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 11:44:00 -0700 Organization: Primenet Services for the Internet Message-ID: <B151149C-26869@206.165.43.154> References: <6ggaju$rts$4@news01.deltanet.com> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Scott Ellsworth <scott@eviews.com> said: > Hmmm. Hell may have just frozen over - I find myself agreeding > with both of your last two posts. In point of fact, your reminder > that Apple had tried this before and failed was very timely. But this was BEFORE (or at least, during) the Spindler nervous breakdown and the PR/marketing disasters that followed hiring Amelio. Apple no longer has a year or two cushion to stave off the ill effects of developer resentment over this. Here's a concrete example: last week, if I needed info about GX, I could download whatever parts of the latest GX SDK that I was lacking. I had meant to download everything in the past week, but had no inkling that this was coming, so I put it off. There were upgrades to the various sample apps and libraries to fit better with the latest Universal Headers libraries that I would have liked to use, as well as an upgrade to the new GX printing library. Unless I become a developer again at $500/per, it looks like I'm stuck using 2-year-old SDKs, which means no GXFCN printing for MacOS 8. Too bad, you say? GX is dead, you say? Yeah, but what about every other shareware developer who was in the middle of producing/upagrading a product? Are all of those worthless, also? My next project after getting GXFCN out the door would have been (might still be) to create a QD3D_FCN to provide HyperCard access to QuickDraw 3D. My last QD3D SDK is years out of date and I'll have to spend that $500 to get it updated. If anyone seriously thinks that having interactive programming access to the QD3D API (function calls) wouldn't be a killer product for Apple, they are living on one of the moons of Jupiter and the radiation has gotten to their brains but since I'm operating out of my home, with no regular income to speak of, taking care of a sick kid, while trying to make sure that his mom improves her education, that $500 is going to be mighty hard to come by. And so, the mighty Titanic hits the iceberg going at full speed. --------------------------------------------------------------------- "OpenDoc may be impressive technology, but if no-one else is using it, why should Apple?" --Steve Jobs "In short: if you want to blow people away with what can be done *only* on a Mac, show them Cyberdog." --Tom Keyes <news://cyberdog.apple.com/cyberdog.general/19411> ---------------------------------------------------------------------
From: gcheng@uni.uiuc.edu (Guanyao Cheng) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 19:24:55 GMT Organization: University of Illinois at Urbana-Champaign Message-ID: <6ggiu7$auu$1@vixen.cso.uiuc.edu> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> "Lawson English" <english@primenet.com> writes: >But shareware programmers often fill in the gaps that commercial software >leaves. Also shareware programmers learn the ropes via feedback from their >customers and are often high school/college students who later enter the >Mac programming field as professionals. >There's one very, VERY important piece of software that wouldn't exist if >shareware wasn't easy to do on the Mac: >Stuffit. >Raymond Lau put himself through MIT by writing Stuffit when he was 16 and >Alladin Software is a pretty important software house in the Macintosh >community.. >Where's the next Stuffit or Alladin Software for MacOS going to come from >if there aren't going to be any more Raymond Lau's? Although I disapprove of the move by Apple, I really question why a HS/college student would have to join the developer programs to write software. After all, they could just download MPW or buy CodeWarrior at a huge discount, and learn from the vast amount of online resources at devworld.apple.com. Although I agree it might drive away small, commercial developers, for the student developer, even the previous $250 was too much. Guanyao Cheng -- Guanyao Cheng "And I personally assure you, everybody here, that gcheng@uiuc.edu if Deep Blue will start playing competitive http://www.uiuc.edu/ph chess, I personally guarantee you I'll /www/gcheng tear it to pieces" -- Garry Kasparov
From: Charles Swiger <chuck-nospam@blacksmith.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 19:26:57 GMT Organization: BLaCKSMITH, Inc. Message-ID: <6ggj21$fit$6@anvil.BLaCKSMITH.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gecns$f9p$1@news01.deltanet.com> <1d767ho.1tixfqy1dqgm03N@hobbit1.injep.fr> <6gg6af$fit$1@anvil.BLaCKSMITH.com> <6ggabn$rts$2@news01.deltanet.com> scott@eviews.com (Scott Ellsworth) wrote: [ ... ] >> I don't have much of an opinion about Apple's change to their Dev. program >> pricing, one way or the other. I do know, however, that people who condone >> software piracy have no moral grounds whatsoever to complain about what Apple >> does. > > The fellow's point was, I hope, that limiting distribution in this way > seems unlikely to prevent the seeds from getting to the warez sites, > because the current arcana one has to go through to get seeds > (and it is VERY arcane) has not succeeded. I don't believe Apple's actions should be influenced significantly by the prospect that people will pirate their software. (Aside from the IMHO clever effort they've made mastering their CD's to be uncopyable via a standard CD-R burner. :-) > As a result, the only people who are incovenienced by the > former situation were honest developers who wished to get > new seeds directly from Apple. The new situation will limit > the exposure of honest developers to upcoming MacOS > releases even more. I'm not sure people understand the purpose of seeding and beta testing programs. Rhapsody DR1 is the first "semi-public" beta release of a complex new operating system. DR1 is not a final product; it is not ready for production usage, nor is it suitable as a standalone platform for serious development. Actually, there are multiple purposes. Apple's is to get the software out to the people who have the ability and resources to test Rhapsody, and give these people early access to the technologies so that they can start writing and porting third-party software. However, Apple does not want to spend the time (== money) and resources to provide extensive support for DR1 to individual people who just want to play around. In the long run, Apple's customer base is better served by Apple using their finite resources effectively. For a perfect example, someone over the last day or so was asking how to make DR1 into a primary nameserver and mail exchanger. That doesn't make any sense-- DR1 is *beta*, not production. [ ... ] > Let me say that more loudly - by limiting things in this way, they > are only preventing honest developers willing to pay a substantial > fee from testing products and giving useful feedback. It costs a company over $100,000 annually to support one developer. $500 per year is less than 1%. If a company has any actual interest in Rhapsody as a technology or deployment target, they can easily afford to pay to join the developer program and gain seeding access. The people for whom this price change matters are individuals and students. And, to be blunt, these people are the ones who are unlikely to have the resources (unused machine(s) to dedicate to Rhapsody, a functioning LAN with a firewall [to meet the RDR1 licensing terms], and a working fileserver and preferably NetInfo server) to provide Apple with the kinds of feedback that they could use. Such individuals are far more likely to require excessive support resources for little benefit. -Chuck Charles Swiger | chuck@BLaCKSMITH.com | standard disclaimer ---------------+----------------------+-------------------- "Microsoft: we make the easy almost impossible."
Message-ID: <3528ED7E.D51513BA@mci.com> From: David Hinz <David.Hinz@mci.com> Organization: MCI MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Reading NSPPL files? Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Mon, 06 Apr 1998 15:59:30 GMT NNTP-Posting-Date: Mon, 06 Apr 1998 11:59:30 EST Questions about using NSPPL. Using the NSPPL example code from the documentation I tried adding some additional code to read in the file. After the [ppl save] I create a new PPL instance to the file to just read the data. When newPPL is printed it only shows the last dictionary added to the PPL. Printing the data show lots of output, but when the inputDictionary is printed it is empty. Is there something else that needs to be done to use NSPPL? Also, when using NSPPL is the entire file read into memory each time it is opened ([NSPPL pplWithPath: create: readOnly:])? dave ------------------------------------------------ #import <Foundation/Foundation.h> void main (int argc, const char *argv[]) { NSAutoreleasePool * pool = [[NSAutoreleasePool alloc] init]; NSPPL *ppl; NSString *pplPath; NSMutableDictionary *root; NSString *name = @"Chris Smith"; NSArray *childArray; NSMutableDictionary *addressDict; NSPPL *newPPL; NSData *data; NSDictionary *inputDictionary; childArray = [NSArray arrayWithObjects:@"Sam", @"Bettina", @"Eloise", nil]; addressDict = [NSMutableDictionary dictionaryWithCapacity:4]; /* Add data to addressDict. */ [addressDict setObject:@"955 Elm Street" forKey:@"street"]; [addressDict setObject:@"Midland" forKey:@"city"]; [addressDict setObject:@"Kansas" forKey:@"state"]; [addressDict setObject:@"19067" forKey:@"zipcode"]; pplPath = @"PersonalInfo.ppl"; /* Read the NSPPL; if it doesn't exist, create it. */ ppl = [NSPPL pplWithPath:pplPath create:YES readOnly:NO]; if (!ppl) { NSLog(@"Couldn't open or create %@", pplPath); exit(1); } /* Get ppl's root dictionary. */ root = [ppl rootDictionary]; /* Through the root dictionary, add data to ppl. */ [root setObject:name forKey:@"name"]; [root setObject:addressDict forKey:@"address"]; [root setObject:childArray forKey:@"children"]; /* Save ppl. */ [ppl save]; // ------------ new code for reading the file. ------------ // Read the NSPPL; don't create it if it doesn't exist, // open readOnly. newPPL = [NSPPL pplWithPath:pplPath create:NO readOnly:YES]; NSLog (@"newPPL = %@", newPPL); if (newPPL == nil) { NSLog (@"Unable to open PPL file %@:", pplPath); exit (1); } data = [newPPL contentsAsData]; NSLog (@"data = %@", data); inputDictionary = [NSPPL propertyListWithPPLData:data]; NSLog (@"inputDictionary: %@", inputDictionary); [pool release]; exit(0); // insure the process exit status is 0 } -- ===================================================== = David Hinz MCI Telecommunications = = Internet and New Media Development = = Email: David.Hinz@MCI.com Phone: (303) 390-6108 = = Vnet: 636-6108 Fax: (303) 390-6365 = = Pager: 1-888-900-5732 (Interactive 2-way) = =====================================================
From: gutkneco+news@lirmm.fr (Olivier Gutknecht) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 8 Apr 1998 22:55:48 +0100 Organization: Elevage d'agents nourris au grain Message-ID: <1d76snt.17ecjre1ykt2xdN@ppgutkneco.lirmm.fr> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gecns$f9p$1@news01.deltanet.com> <1d767ho.1tixfqy1dqgm03N@hobbit1.injep.fr> <6gg6af$fit$1@anvil.BLaCKSMITH.com> <6ggabn$rts$2@news01.deltanet.com> Scott Ellsworth <scott@eviews.com> wrote: > In article <6gg6af$fit$1@anvil.BLaCKSMITH.com>, cswiger@BLaCKSMITH.com wrote: > >xhsoft@injep.fr (Xavier Humbert) wrote: > >> Scott Ellsworth <scott@eviews.com> wrote: > >>> The seeding program, imho, served a valuable purpose, but Apple, I > >>> think, does not want any leaks of its seeds, and so it is trying to > >>> reduce the number of developers who might have access. > >> What is obviously stupid : WarEz sites have System Seeds BEFORE the > >> official FTP :-( > >> > >> Xav, looking for a warez URL > > > >I don't have much of an opinion about Apple's change to their Dev. program > >pricing, one way or the other. I do know, however, that people who condone > >software piracy have no moral grounds whatsoever to complain about what Apple > >does. > > The fellow's point was, I hope, that limiting distribution in this way > seems unlikely to prevent the seeds from getting to the warez sites, > because the current arcana one has to go through to get seeds > (and it is VERY arcane) has not succeeded. And in the European Developer Programs we experienced frequent problems with the central FTP seed. People in the CQF often had access to the seeds before members of the euroDev programs, and on an european mirror. :-/ Another interesting information: the FAQ on developer.apple.com reads that Apple has been limiting access to the Associate Programs in the last six months... In the US only I guess, as I renewed my membership two weeks ago and have not heard a word from EDR indicating that I might not have what I was paying for. > Let me say that more loudly - by limiting things in this way, they are > only preventing honest developers willing to pay a substantial fee from > testing products and giving useful feedback. $250 is substantial when it > is out of your own pocket, and many Mac evangelists and partisans at > otherwise wintel companies often have to do thier Mac work on thier own, > in order to sell the company on the benefits. Agreed. The only explanation is that Apple is not interested anymore in small/independant development groups. Ol. -- Olivier Gutknecht ... LIRMM Montpellier - A.I. Dept gutkneco@lirmm.fr ... Multi-Agent Group - MadKit Architect
From: trumbull@cs.yale.edu (Ben Trumbull) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 1998 17:00:06 -0400 Organization: Yale University Message-ID: <trumbull-0804981700060001@net44-223.student.yale.edu> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gecns$f9p$1@news01.deltanet.com> <1d767ho.1tixfqy1dqgm03N@hobbit1.injep.fr> <6gg6af$fit$1@anvil.BLaCKSMITH.com> <6ggabn$rts$2@news01.deltanet.com> <6ggj21$fit$6@anvil.BLaCKSMITH.com> In article <6ggj21$fit$6@anvil.BLaCKSMITH.com>, cswiger@BLaCKSMITH.com wrote: > provide Apple with the kinds of feedback that they could use. Such individuals are > far more likely to require excessive support resources for little benefit. The only development resources I've ever required from Apple is a web server they were going to be running for thousands of other people anyway. And Apple has made quite a handsome little profit on selling me Inside Macintosh and other development resources. The numbers somebody came up with regarding Apple's cost to support one "developer" approach something like pure fiction. As for the little benefit, that's a cheap shot. I'll think of you when I write the licensing part of my next freeware program. terminally curious, Ben ___________________________________________________________________ Benjamin Trumbull trumbull@cs.yale.edu Yale University You can't be in hell; you can still read your e-mail
From: petcher@howdy.wustl.edu (Donald N. Petcher) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 21:24:46 GMT Organization: Washington University in St. Louis Message-ID: <6ggpuu$f6q$1@newsreader.wustl.edu> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <01bd6267$fa532b30$9b2168cf@test1> <352ad905.0@206.25.228.5> <1d761b5.13ha0xu1i2sd4wN@roxboro0-043.dyn.interpath.net> > "Todd Heberlein" <todd@NetSQ.com> wrote: > > Yea, and look at the marketshare and number of developers for that > platform vs. the mac. I don't know the details of this program > change yet, so I'll reserve final judgement till then. But *if* > they cut services and doubled the costs to developers, I'd say this > is the proud continuation of BOZO behavior at Apple. This sounds more like NeXT infiltration to me. Cheers, Don Petcher
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 98 21:34:57 GMT Organization: QMS Message-ID: <6ggqha$9fk$1@news01.deltanet.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> In article <6ggmk2$jdl$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: >On 04/08/98, George Graves wrote: >>In article <6gg8il$6v9$1@ha2.rdc1.sdca.home.com>, >>rexr@cx54440-a.dt1.sdca.home.com wrote: >>What this really means is less shareware (on top of the fact that >>its already shrinking because the platform is shrinking), and less >>software from small developers in general, you know, the people >>from whom all the innovation comes. > > Thats a crock. I disagree. Small developers run quite close to the margin in many cases, and when you price something to exclude them, you are by definition excluding the people with the least inertia. It may be the case that those people produce little software worth buying, but I have seen many products bought by people like Alladin, Microsoft, and Apple which were originally produced by one and two man shops. The new developer programs make it more difficult for those people to be out early with functioning applications. >The Online program is free. It gets you the samples, and the >manuals. It also gets you Code Warrior Lite. And it cannot easily be used by a startup trying to put product together, nor can it be used by someone doing thier own private skunk works project. Trust me on this - I want the Mac to be a viable platform at QMS, which it is not. The only way I can demonstrate the value of the platform is to purchase appropriate Mac tools, build our products, and demonstrate that they are interesting. Having early builds makes it more likely that I will have a demo that means something, I know two other developers at former cross platform shops doing much the same. God only knows if this will work, but it is the case that QMS has already decided to drop Mac development, so they are a lost cause if I cannot change that. > From that you can learn Macintosh programming and put out >freeware apps. True enough. Of course, your freeware and shareware will need modification after the OS release happens. Further, certain small scale authors push an OS a hell of a lot more than many commercial app developers do. > Few, if any, of the people pissing and moaning about loosing >access to the seeding program actually NEED access to it in many >cases. Of course, they changed the rules in mid year, such that anyone already paying for a developer subscription got substantially fewer benefits. (Specifically, dropping the hardware agreement unlaterally is far from a nice move, and changes the economic impact of the developer program substantially.) > Now, as far as Rhapsody is concerned it is leaving a gap, but >only because of the time-frame.... However if you join Select you're >still covered. There will always be a gap. Apple is incapable of shipping something with decent prerelease documentation in the New Regime, I suspect. HFS+ might be a counterargument, but if there is no way to get a Rhapsody seed without a substantial cash outlay, then you have delayed the release of software using Rhapsody features until your public ship date. Scott Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
From: "macghod" <macghod@concentric.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 21:46:29 GMT Message-ID: <01bc452e$e8a3eac0$3df0bfa8@davidsul> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggiu7$auu$1@vixen.cso.uiuc.edu> > Although I disapprove of the move by Apple, I really question why a > HS/college student would have to join the developer programs to write > software. After all, they could just download MPW or buy CodeWarrior at a > huge discount, and learn from the vast amount of online resources at > devworld.apple.com. Although I agree it might drive away small, > commercial developers, for the student developer, even the previous $250 > was too much. Really? How does a college student who wants to develop apps for rhapsody do so without spending the $250 (now $500)? Thats the only way they can legitimately get rhapsody, plus other apple beta softwares they may want to develop for.
From: steve@discoverysoft.com (Steven Fisher) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 1998 14:42:59 -0700 Organization: Discovery Software Ltd. Message-ID: <steve-0804981442590001@oranoco.discoverysoft.com> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> In article <B1511019-15903@206.165.43.154>, "Lawson English" <english@primenet.com> wrote: >Trying to make developer *support* a profit center is an INSANE idea. >Hopefully, either Jobs will change this, or the more rational stockholders >will hear about it in time to vote against him because this one action, *BY >ITSELF*, will destroy Apple. That's the best summary I've seen yet. The real worry is that it will do so *SLOWLY*... consumers probably won't see the effects this has on the development community for months or a year yet. Suddenly new shareware/freeware will start getting a lot more scarce. -- Steven Fisher; Discovery Software Ltd.; steve@discoverysoft.com "Anyone who has never made a mistake has never tried anything new." -- Albert Einstein
From: "Paul Rekieta" <PRekieta@ix.netcom.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 8 Apr 1998 14:53:05 -0700 Organization: Apple Computer, Inc. Message-ID: <6ggrjt$82q$1@news2.apple.com> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> > >No betas, and now, apparently no online versions of the SDKs. They were >available last week online, but no longer. > The SDK's are at <http://devworld.apple.com/sdk/index.html>
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 98 21:57:03 GMT Organization: QMS Message-ID: <6ggrqo$b5p$1@news01.deltanet.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gecns$f9p$1@news01.deltanet.com> <1d767ho.1tixfqy1dqgm03N@hobbit1.injep.fr> <6gg6af$fit$1@anvil.BLaCKSMITH.com> <6ggabn$rts$2@news01.deltanet.com> <6ggj21$fit$6@anvil.BLaCKSMITH.com> In article <6ggj21$fit$6@anvil.BLaCKSMITH.com>, cswiger@BLaCKSMITH.com wrote: >I don't believe Apple's actions should be influenced significantly by >the prospect that people will pirate their software. (Aside from > the IMHO clever effort they've made mastering their CD's to be > uncopyable via a standard CD-R burner. :-) Though clearly this is a major influence on thier actions. .. >Rhapsody DR1 is the first "semi-public" beta release of a complex new >operating system. DR1 is not a final product; it is not ready for >production usage, nor is it suitable as a standalone platform for > serious development. It is, though, a good platform to start gaining developer familiarity. Further, it is "neat" enough to gather some interest, and that "neat" factor comes part and parcel with the low inertia that small shops have. Besides, I have an external 2G drive I can put the seed on, use, and then unplug and tuck away when I am done. This justifies the use for me, and makes it suitable as a test bed. >Actually, there are multiple purposes. Apple's is to get the software out to >the people who have the ability and resources to test Rhapsody, and give >these people early access to the technologies so that they can start >writing and porting third-party software. Yep. And many of the small time developers are the types who can both commit to an untried technology, and push the OS in ways the designers did not intend. This is handy for OS testing. >However, Apple does not want to spend the time (== money) and resources to >provide extensive support for DR1 to individual people who just want to play >around. On the other hand, finding out just how good the Java environment is might just be a good idea before it ships. (As an example). MrJ's AWT implementation is substantially slower than any VM I have used on Windows in feel, but there are not going to be any changes in it for a while, because Apple needs to get 1.2 going. Thus, informing them NOW of what is wrong with MrJ is not going to do any good. It would be really handy if I could try some of my Java apps against thier Rhapsody Java client before they consider it done, because Apple does not like opening a component once they consider it done. The only time when this can happen is while the component is still in alpha or beta, not after release. With thier new decisions, they have cut down the number of people who can even try the aplpha/beta software. >In the long run, Apple's customer base is better served by Apple >using their finite resources effectively. For a perfect example, >someone over the last day or so was asking how to make DR1 into >a primary nameserver and mail exchanger. That doesn't make any >sense-- DR1 is *beta*, not production. A decent point. I certainly wouldn't try it, but had I a network that was internal only, I might give it a shot just to see if it was really up to it. Further, Apple might be interested in what worked and what did not. Since they do not provide a list of what features are "done" and which are unfinished, it is hard to know exactly what to test as a developer. >[ ... ] >> Let me say that more loudly - by limiting things in this way, they >> are only preventing honest developers willing to pay a substantial >> fee from testing products and giving useful feedback. > >It costs a company over $100,000 annually to support one developer. >$500 per year is less than 1%. If a company has any actual interest >in Rhapsody as a technology or deployment target, they can easily >afford to pay to join the developer program and gain seeding access. With that attitude, you are insuring that only strongly pro Mac companies are in the early access loop, as opposed to strongly pro Mac individuals. This might be for the best, as it does transition neatly into the corporate/enterprise world. Unfortunately, this is a world in which they have mostly already lost. Apple has damn few big friends anymore, and if they annoy the small developers too much, well, each one is a former Mac partisan who might have some influence over thier organizations. >The people for whom this price change matters are individuals and >students. And, to be blunt, these people are the ones who are >unlikely to have the resources (unused machine(s) to dedicate to >Rhapsody, a functioning LAN with a firewall [to meet the >RDR1 licensing terms], and a working fileserver and preferably >NetInfo server) to provide Apple with the kinds of feedback that >they could use. Such individuals are far more likely to require >excessive support resources for little benefit. My Linux-based firewall cost $500 to build a year and a half ago, and I could have spent less had I not wanted a linux box for other purposes. Assuming you have reasonable discounts on hardware, people will buy new machines, and thus can be assumed to have an old one lying about, or at least an external drive they can use for it when the time is right. Perhaps more important, the support demands are not all that high for the potential return. The real costs are some tremendous fixed costs. Qualifying a build and burning a CD takes a tremendous amount of time. Duping fifty copies instead of ten is a minor change. Responding to the mail you get can be problematic, but before deciding that you have product that does not need seeding, note that Metrowerk's last build of CWPro 3 had a number of final candidate builds. For them, it was worth giving it away betas free to customers to find these last minute bugs. I know I found one that they fixed for Pro 3, and I was not pushing it that hard. Scott Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
From: ktatroe@badgercom.com (k.tatroe) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 1998 16:00:39 -0600 Organization: writers are queer so keep away Message-ID: <ktatroe-0804981600390001@prospero.frii.com> References: <6ggj21$fit$6@anvil.BLaCKSMITH.com> <B1513924-1B3A6@206.165.43.150> In article <B1513924-1B3A6@206.165.43.150>, "Lawson English" <english@primenet.com> wrote: > Charles Swiger <chuck-nospam@blacksmith.com> said: > > > The people for whom this price change matters are individuals and > > students. And, to > > be blunt, these people are the ones who are unlikely to have the > resources > > (unused > > machine(s) to dedicate to Rhapsody, a functioning LAN with a firewall [to > > meet the > > RDR1 licensing terms], and a working fileserver and preferably NetInfo > > server) to > > provide Apple with the kinds of feedback that they could use. Such > > individuals are > > far more likely to require excessive support resources for little > benefit. > > > > > > Games Sprockets for MacOS now costs $500. > > It isn't just the Rhapsody beta seeding. That I don't mind paying for. <http://developer.apple.com/sdk/> has not only Apple Game Sprockets, but just about every other SDK that came on the April CD (I don't have the CD here at my day job to compare item by item, but I don't see anything missing offhand). As a small developer - a hobbyist, really - the $250 above the old program's cost isn't minor to me, but it's still well within reach. If I didn't want access to the beta stuff (and really, I think few small developers - even the cool ones who come up with brilliant main-stay stuff - need access to beta stuff), I'd just pay for the monthly mailing, and save $50 over what I paid for the Associates program. While the price increase doesn't bother me much, the loss of hardware purchasing (which I don't have the budget to take advantage of, anyway) bothers me greatly. However, if Apple starts reacting to its developers like Be has to its developers (to me, anyway), the cost increase will be trivial compared to the better relationship. If, however, Developer Programs continues to be a black hole where questions about such things as "my credit card was charged three months ago and I still haven't even recieved confirmation" like some have experienced in the past, I may have to join the ranks of the livid... -k. 08-Apr-98 ktatroe@badgercom.com
From: Charles Swiger <chuck-nospam@blacksmith.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 21:43:15 GMT Organization: BLaCKSMITH, Inc. Message-ID: <6ggr1j$fit$8@anvil.BLaCKSMITH.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gecns$f9p$1@news01.deltanet.com> <1d767ho.1tixfqy1dqgm03N@hobbit1.injep.fr> <6gg6af$fit$1@anvil.BLaCKSMITH.com> <6ggabn$rts$2@news01.deltanet.com> <6ggj21$fit$6@anvil.BLaCKSMITH.com> <trumbull-0804981700060001@net44-223.student.yale.edu> trumbull@cs.yale.edu (Ben Trumbull) wrote: > In article <6ggj21$fit$6@anvil.BLaCKSMITH.com>, cswiger@BLaCKSMITH.com > wrote: [ ... ] >> provide Apple with the kinds of feedback that they could use. Such >> individuals are far more likely to require excessive support resources >> for little benefit. > > The only development resources I've ever required from Apple is a web > server they were going to be running for thousands of other people > anyway. And Apple has made quite a handsome little profit on selling me > Inside Macintosh and other development resources. The numbers somebody > came up with regarding Apple's cost to support one "developer" approach > something like pure fiction. Which numbers? My 100K figure? That was the cost per developer per year to their employers, not Apple's cost to provide support for a product. Take an average salary of 50K, plus 20K of administrative overhead, plus 10K of software & hardware, plus 10K for taxes & bennies (employer's social security contribution, health coverage, 401(k) matching, etc), plus miscellaneous expenses like renting office space, utilities et al. Typical computer technical support costs run a company around $30 - 50 per hour, give or take, depending on the quality, responsiveness, and expertise of the support people. And that's just the up-front support cost; it doesn't cover the engineering resources required to fix bugs, add functionality, and so forth. I could find references for these numbers if you like. Actually, IIRC, Steve McConnell (of 'Code Complete') has a discussion of these costs. Obviously, there is a range depending on circumstances, but they should be accurate to better than a factor of two. > As for the little benefit, that's a cheap shot. How so? I wasn't directing my comments at anyone in particular, so I can't imagine how they could be construed that way. But heck, I'll bite: You're posting from a .edu address. Do you have your own private network-- seperated by a firewall-- to even meet the licensing terms for Rhapsody DR1? You're not supposed to install the software on a box connected to your campus network, you see. Do you have another machine besides your primary system which you can dedicate entirely to Rhapsody? DR1 isn't a production system, and anyone who needs a computer to actually do work (for class, commercially, or for personal usage) would need a second machine to install Rhapsody on. > I'll think of you when I write the licensing part of my next freeware > program. If it makes you feel better.... -Chuck Charles Swiger | chuck@BLaCKSMITH.com | standard disclaimer ---------------+----------------------+-------------------- "Microsoft: we make the easy almost impossible."
From: gcheng@uni.uiuc.edu (Guanyao Cheng) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 22:14:05 GMT Organization: University of Illinois at Urbana-Champaign Message-ID: <6ggsrd$ldi$1@vixen.cso.uiuc.edu> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggiu7$auu$1@vixen.cso.uiuc.edu> <01bc452e$e8a3eac0$3df0bfa8@davidsul> "macghod" <macghod@concentric.net> writes: >Really? >How does a college student who wants to develop apps for rhapsody do so >without spending the $250 (now $500)? >Thats the only way they can legitimately get rhapsody, plus other apple >beta softwares they may want to develop for. I assume you would go get a book for Rhapsody programming, and by the time you finish it Rhapsody CR1 would be out, and educational users and developers would be able to purchase it at cost. Guanyao Cheng -- Guanyao Cheng "And I personally assure you, everybody here, that gcheng@uiuc.edu if Deep Blue will start playing competitive http://www.uiuc.edu/ph chess, I personally guarantee you I'll /www/gcheng tear it to pieces" -- Garry Kasparov
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 22:05:48 GMT Organization: P & L Systems Message-ID: <6ggsbs$oop$3@ironhorse.plsys.co.uk> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <j-jahnke-0704981616320001@192.168.1.3> <support-0704982011260001@207-172-164-184.s57.as4.loc.erols.com> <6ggahf$rts$3@news01.deltanet.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: scott@eviews.com In <6ggahf$rts$3@news01.deltanet.com> Scott Ellsworth wrote: > I like the idea. Who do we complain to, and how? Ric Ford said on > Macintouch that the feedback was uniformly negative. > It couldn't possibly be now; I wrote in to support Apple. Best wishes, mmalc.
From: mazulauf@atmos.met.utah.edu (Mike Zulauf) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 1998 15:03:38 -0600 Organization: Universtity of Utah - Department of Meteorology Message-ID: <mazulauf-0804981503390001@sneezy.met.utah.edu> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> I6qtS>9WAIrq2n":C@7PD>2h2WE))[h;3h<yre7|}"FBUwf_y(om>j9#T In article <joe.ragosta-0804980858150001@wil77.dol.net>, joe.ragosta@dol.net (Joe Ragosta) wrote: > Actually, there were rumors that there would be a very low price > educational plan. You might want to wait for the educational pricing > before worrying too much. The eternal Joe Ragosta reply. "All is well. Don't worry. Wait for details." Unfortunately, when details do come out, you're just as screwed as ever. . . Mike -- Mike Zulauf mazulauf@atmos.met.utah.edu
Date: Wed, 08 Apr 1998 18:52:44 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-0804981852440001@elk68.dol.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> In article <6gg8il$6v9$1@ha2.rdc1.sdca.home.com>, rexr@cx54440-a.dt1.sdca.home.com wrote: > Some people are under the assumption that Apple's Developer Program is > tasked with the "mission" to multiply and procreate a whole cadre of > wonderful developers and new applications for Rhapsody. Seeded means that > Apple must grow all things naturally (ie. Ye reap what ye sow). This too > is an assumption, that Apple would seed "individual" developers with "the > code". > > Apple's actions signal that it can't afford to get into the "handout" > business nor support legions of "individual" developers (read welfare). By > "raising-the-bar" Apple introduces an entry fee and seeds the concept of > "profit" in its Developer base. Developers who approach Rhapsody without > resources and a profit motive will be annoyed, discouraged and drop off. There are conflicting needs here and I'm not sure what the right answer is. While Apple can't create a welfare program, it _is_ in their best interest to court developers. The only thing I could think of was to set the fee at $500 and offer a rebate to any developer who shows that they've created real applications. Of course, enforcing this would be a nightmare. -- Regards, Joe Ragosta See the Complete Macintosh Advocacy Page http://www.dol.net/~Ragosta/complmac.htm
Date: Wed, 08 Apr 1998 19:07:29 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-0804981907290001@elk68.dol.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <j-jahnke-0704981616320001@192.168.1.3> <support-0704982011260001@207-172-164-184.s57.as4.loc.erols.com> <6ggahf$rts$3@news01.deltanet.com> <6ggsbs$oop$3@ironhorse.plsys.co.uk> In article <6ggsbs$oop$3@ironhorse.plsys.co.uk>, mmalcolm crawford <malcolm@plsys.co.uk> wrote: > In <6ggahf$rts$3@news01.deltanet.com> Scott Ellsworth wrote: > > I like the idea. Who do we complain to, and how? Ric Ford said on > > Macintouch that the feedback was uniformly negative. > > > It couldn't possibly be now; I wrote in to support Apple. It doesn't matter--they'll still claim that the feedback was uniformly negative. Many of these sites are pretty reactionary. -- Regards, Joe Ragosta See the Complete Macintosh Advocacy Page http://www.dol.net/~Ragosta/complmac.htm
Date: Wed, 08 Apr 1998 19:06:46 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-0804981906460001@elk68.dol.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> In article <mazulauf-0804981503390001@sneezy.met.utah.edu>, mazulauf@atmos.met.utah.edu (Mike Zulauf) wrote: > In article <joe.ragosta-0804980858150001@wil77.dol.net>, > joe.ragosta@dol.net (Joe Ragosta) wrote: > > > Actually, there were rumors that there would be a very low price > > educational plan. You might want to wait for the educational pricing > > before worrying too much. > > The eternal Joe Ragosta reply. "All is well. Don't worry. Wait for details." > > Unfortunately, when details do come out, you're just as screwed as ever. . . Please explain how you're so badly screwed. The SDKs are still downloadable for free. The monthly mailing is available for $50 less than the old Associate membership. The lowest paid developer status is $500 instead of $250. If you really need to get the beta OS versions, your cost went up by $250. But everyone else seems to be better off. -- Regards, Joe Ragosta See the Complete Macintosh Advocacy Page http://www.dol.net/~Ragosta/complmac.htm
From: Charles Swiger <chuck-nospam@blacksmith.com> Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: accuracy of Stepwise Krishna article Date: 6 Apr 1998 16:45:12 GMT Organization: BLaCKSMITH, Inc. Message-ID: <6gb0qo$kc8$1@anvil.BLaCKSMITH.com> References: <6eoo28$c1p$1@crib.bevc.blacksburg.va.us> <6eq7kp$pl2@mochi.lava.net> <slrn6h2ggl.gdm.marko@ns1.vrx.net> <6erm0k$p1f$2@ironhorse.plsys.co.uk> <6erpdg$96o$1@news.xmission.com> <6escen$to2$1@nnrp1.dejanews.com> <6etffv$g96$9@ns3.vrx.net> <6ev46i$qkn$1@nnrp1.dejanews.com> <6eulns$fh3$1@ns3.vrx.net> <quinlan-2003982045060001@pm22s6.intergate.bc.ca> <6f5q41$p1f$9@ironhorse.plsys.co.uk> <6f6ete$mhp$1@nnrp1.dejanews.com> <3516C50E.94BBD56D@milestonerdl.com> <3516EDD4.B53E67DB@codefab.com> <3516F1AD.588546EE@milestonerdl.com> <6f7cqu$634$1@news.digifix.com> <3517C87D.242606B6@milestonerdl.com> M Rassbach <mark@milestonerdl.com> wrote: [ .... ] > Until Apple comes out with a statement of commitment FORMALLY (Press > Release) that NT/Windows 9x, Solaris and HP-UX is supported, to spout off > 'Write once, run everwhere' is pure bullshit. It sounds like you refuse to be reasonable. What hardware architectures did NEXTSTEP 3.3 run on? For what platforms can you get (formerly) NeXT technologies like OPENSTEP, EOF, PDO, WOF, and so forth? Any statement about the coverage of a technology has a context. The context of the "Write once, run everywhere" for Yellow Box is the desktop and small-to-medium size server environment. It doesn't cover NC's, it doesn't cover set-top boxes, it doesn't cover PDA's, nor does it cover Connection machines, Cray Y-MP's, or other supercomputers. For that matter, it doesn't cover large scale servers like a Sun Enterprise Server x000, or a Tandem box, either.... -Chuck Charles Swiger | chuck@BLaCKSMITH.com | standard disclaimer ---------------+----------------------+-------------------- "Microsoft: we make the easy almost impossible."
From: "macghod" <macghod@concentric.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 23:51:57 GMT Message-ID: <01bc4540$d4a15e60$3af0bfa8@davidsul> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> > The monthly mailing is available for $50 less than the old Associate membership. > > The lowest paid developer status is $500 instead of $250. If you really > need to get the beta OS versions, your cost went up by $250. But everyone > else seems to be better off. More Ragosta FUD. "Everyone else seems to be better off". Who is better off? Their is NOTHING that is now cheaper now than before the announcement. Either things are the same price as before with no added benefits, or they have increased and in some cases have increased with less benefits.
From: ylaporte@acm.org (Yan Laporte) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 00:06:21 GMT Organization: Université de Moncton Message-ID: <ylaporte-0804982005030001@139.103.176.79> References: <01bc445f$744397e0$1bf0bfa8@davidsul> Well... it doesn't seem so bad. You can still download MPW, compilers, examples and SDKs for free... That's good for a student like me. I don't really care about the 500$ program except for one thing... Rhapsody, with students here and teachers we started making plans about it and most of it had to take place this summer... I can't afford 500$! that's more than a month of living for me, that would mean cutting food and other essential stuff. Sure there will be special deals for education when it comes out, that means at the very least at the end of the summer. The problem is they have the bad habit to make these deals available through "our campus' Apple representative" fact is: there si none! But even with access to such program it would come too late, I can't really find time to learn and experiment with a new sytem during school year, I need my machine up and running anytime I need it! I can't install some mostly unknown system software with no productiuvity apps on it (as if I would have money to buy some). I read the introduction manual three times, I read the entire foundation class reference and read the Objective-C manual about five times and took a peek at the Application Foundation classes reference, now I need to try it, we started writing code for a system we never tryed... We got the software ideas, we figured out how we should do it under Rhapsody but now I am stuck. We were able to convince autorities that Rhapsody on a test machine would be good for us "and would be cheap" at 250$US (We're in Canada) and we are about to get the budget... sadly it suddenly becomes useless and we still got no educationnal program of any sort available... What we need from Apple is an idea of when and what we will have to pay for the system in higher education institutions. And don't get me the "you got to be carefull, it is not meant to be used by normal users for now(...)" speech. We're a Computer Science department, we just crave for that stuff and we have the brain to handle it. So what are the plans for Rhapsody in Higher education institutions?? Anyone? I need more than dumb hearsays from MOSR... There was no such as a Developper rebate or any program available for us three weeks ago, is there now? This is getting frustrating... at least we got MkLinux Enc
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 17:34:01 -0700 Organization: Primenet Services for the Internet Message-ID: <B1516690-C6005@206.165.43.150> References: <6ggrjt$82q$1@news2.apple.com> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Paul Rekieta <PRekieta@ix.netcom.com> said: > >No betas, and now, apparently no online versions of the SDKs. They were > >available last week online, but no longer. > > > The SDK's are at <http://devworld.apple.com/sdk/index.html> Yep, found that out earlier today. The setup for downloading looks to be better than the old site's, so I'm not just off-base, but out-system (watching the 100th episode of B-5 right now ). --------------------------------------------------------------------- "OpenDoc may be impressive technology, but if no-one else is using it, why should Apple?" --Steve Jobs "In short: if you want to blow people away with what can be done *only* on a Mac, show them Cyberdog." --Tom Keyes <news://cyberdog.apple.com/cyberdog.general/19411> ---------------------------------------------------------------------
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 17:35:01 -0700 Organization: Primenet Services for the Internet Message-ID: <B15166EA-C7560@206.165.43.150> References: <6ggrql$oop$1@ironhorse.plsys.co.uk> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit mmalcolm crawford <malcolm@plsys.co.uk> said: > > Why does anyone *need* to be part of the program if they're not developing > professional or semi-professional apps. Sure, access to the latest and > greatest betas of the s/w is fun, but not essential if you're hacking at > home > for the joy of it. Apple is not a charity -- and I don't want to have to pay > more for my Mac to subsidise someone else's hobby. > I thought that the SDK's were no longer online (all my URLs look to be broken with the new setup). As far as the betas go, I'd like to see cheaper access for shareware developers, but that's another point than the one I was making. --------------------------------------------------------------------- "OpenDoc may be impressive technology, but if no-one else is using it, why should Apple?" --Steve Jobs "In short: if you want to blow people away with what can be done *only* on a Mac, show them Cyberdog." --Tom Keyes <news://cyberdog.apple.com/cyberdog.general/19411> ---------------------------------------------------------------------
From: xhsoft@injep.fr (Xavier Humbert) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 9 Apr 1998 02:52:26 +0200 Organization: Maquis Usenet Ouest Message-ID: <1d776cj.1sobwhydfgigwN@hobbit1.injep.fr> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gecns$f9p$1@news01.deltanet.com> <1d767ho.1tixfqy1dqgm03N@hobbit1.injep.fr> <6gg6af$fit$1@anvil.BLaCKSMITH.com> <6ggabn$rts$2@news01.deltanet.com> <joe.ragosta-0804981312160001@wil32.dol.net> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Mail-Copies-to: never Joe Ragosta <joe.ragosta@dol.net> wrote: > Maybe that was the point, but it's not how I read it, either. But it was the way I wrote it :-} Xav -- Xavier HUMBERT Laboratoire Informatique INJEP Office: labo-info/AT/injep.fr Home: humbert/AT/injep.fr
From: don.brown@cesoft.com (Donald Brown) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 8 Apr 1998 19:52:54 -0500 Organization: CE Software Message-ID: <MPG.f95d7e4b8fb1622989890@news.supernews.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> In article <joe.ragosta-0804981906460001@elk68.dol.net>, joe.ragosta@dol.net says... > > Unfortunately, when details do come out, you're just as screwed as ever. . . > > Please explain how you're so badly screwed. > > The SDKs are still downloadable for free. > > The monthly mailing is available for $50 less than the old Associate membership. > > The lowest paid developer status is $500 instead of $250. If you really > need to get the beta OS versions, your cost went up by $250. But everyone > else seems to be better off. > Including Microsoft, who looks better and better every day. Donald
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: NSBrowser timed update problems Date: 5 Apr 1998 15:51:30 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6g8nc2$2n6$1@crib.bevc.blacksburg.va.us> I've got a passive delegate for an NSBrowser that caches the columns internally. Whenever -browser:numberOfRowsInColumn: is sent, it first checks to see what the current selected cell in the column prior to the previous column is. This is in case the user just clicked on a different cell in that column, which means that the current column is now invalidated. If necessary, it generates a new column and inserts it into the cache. It then returns the size of whatever is in the cache for that column. This works fine if you select things slowly. But if you very rapidly click back and forth on different cells in a column, causing rapid updates of the next column, then in short order an exception gets thrown from -browser:willDisplayCell:atRow:column:, with a message like: "NSRunLoop: ignoring exception 'NSRangeException' (reason '*** -[NSConcreteArray objectAtIndex:]: index (5) beyond bounds (5)') that raised during delayed perform of target 1018480 and selector '_handleWindowNeedsDisplay:'" If I do a backtrace, I find that what's going on is that I've clicked on something in column N and it's trying to display the cells in column N+1. However, apparently the count it has for how many rows there are in column N+1 is out of date. What seems to be happening is that the column gets updated in -browser:numberOfRowsInColumn:, but some lazy updating or something is going on and a deferred update is coming in from the runloop with outdated information about that column.. it's got the wrong number of rows, and if the true column has fewer rows than what it thinks it has, then it goes right past the bounds of the array that's holding the rows for the cached column. What do I do about this?? Is this the wrong way to implement column caching in a browser delegate? Do I need to use an active delegate?
Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer From: nagle@netcom.com (John Nagle) Subject: Re: Apple developer program Message-ID: <nagleEr4H6r.ID@netcom.com> Organization: Netcom On-Line Services References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <1d765na.sy4iporb5zhoN@p060.intchg1.net.ubc.ca> Date: Thu, 9 Apr 1998 01:23:14 GMT Sender: nagle@netcom6.netcom.com bbennett@unixg.ubc.ca (Bruce Bennett) writes: >Rex Riley <rriley@yahoo.com> wrote, among other things: >> Apple's actions signal that it can't afford to get into the "handout" >> business nor support legions of "individual" developers (read welfare). By >> "raising-the-bar" Apple introduces an entry fee and seeds the concept of >> "profit" in its Developer base. Developers who approach Rhapsody without >> resources and a profit motive will be annoyed, discouraged and drop off. >"Post no trifling freeware or shareware. Give us shrink-wrapped consumer >products or give us death." >Anyone alarmed by the "think different" ideology should be pleased by >this new turn. Typical Apple. The big devlopers are fed up with Apple because of the declining market share and the changes of direction. So Apple pisses off the small developers. Sure, you can run Microsoft apps, but if you're going to run mostly Microsoft apps, you may as well run them on a Microsoft OS. John Nagle
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: Blocks for Obj-C Date: 5 Apr 1998 16:55:40 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6g8r4c$2uq$1@crib.bevc.blacksburg.va.us> References: <6fokj8$1dp$1@usenet50.supernews.com> <6fpgbb$9o5$2@news.idiom.com> <6fphlp$iot$1@crib.bevc.blacksburg.va.us> <6fsfo4$ln@saturn.genoa.com> In article <6fsfo4$ln@saturn.genoa.com>, Alex Blakemore <alex@genoa.com> wrote: > Nathan Urban wrote: > > The idea is to associate the method implementations with the "protocol > > category"; the original class conforms to the original version of the > > protocol, but with the addition of the new definitions, it conforms to > > the extended protocol. > I can see how this feature might be valuable, but one of the nicest attributes > of Objective-C is its simplicity, which each potentially useful feature could > harm. I think the lack of this feature _hurts_ simplicity in the form of orthogonality; the interface model is incomplete in the sense that you can extend implementations (classes), but not interfaces (protocols). > On the other hand, you can get a similar effect to this currently, by adding a > normal category to NSObject. See the example below. Yes, that works, as others have pointed out as well. It's a hack, though, you don't have the compile-time interface checking (you can write code that will call that method on things that don't conform to the protocol, defeating half the purpose of specifying the protocol instead of the class -- type checking), objects outside of the protocol will indicate that they respond to that message when they shouldn't, it requires somewhat more code and isn't symmetric with how categories are used to extend implementations, and it pollutes the NSObject class with lots of methods (and it could be potentially a lot of methods, since I'd use these insetad of regular class categories). > So even though it sounds like a nice feature, the need may arise seldom > enough to warrant a new language feature, If I had this feature, I'd use it constantly. I don't want to specify classes as arguments to methods. I don't want to couple my design to a particular implementation or class hierarchy. I want to specify only interfaces, via protocols. The lack of this feature is preventing me from doing that. > You wouldn't want Objective-C to follow C++ down the featuritis path would > you? It's not featuritis, it's consistency. You should be able to use interfaces exactly as you would implementations. Other languages (such as TOM, Cecil, etc.) let you do this.
From: jes@rednsi.com (Josep Egea i Sanchez) Newsgroups: comp.sys.next.programmer,comp.sys.next.software Subject: Re: NXHosting to Windows NT and Foreign Keyboards Date: 6 Apr 1998 08:54:13 GMT Organization: Unisource Espana NEWS SERVER Message-ID: <6ga57l$cu3$1@diana.bcn.ibernet.es> References: <6g2dgd$q0e$1@diana.bcn.ibernet.es> <6g9fqc$38g$2@hades.csu.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: thomas@zippy.sonoma.edu In <6g9fqc$38g$2@hades.csu.net> Thomas Poff wrote: > It would be _really_ nice if you could NSHost from a Win95 machine to > Rhapsody, since then you could have a vehicle for displaying Japanese > on standard Win95/8. > > Is there a way of currently doing this? > > Thomas My current experience only includes Windows NT, though it may work with Windows 95 too. However, don't expect too much in the side of fonts. AFAIK, the local (Windows) fonts are used. Best regards --- Josep Egea - jes@rednsi.com - NeXTMail & MIME OK NEXUS Servicios de Informacion - Barcelona (Spain) Telf: + 34 3 285 00 70 - Fax: + 34 3 284 31 43 > Josep Egea i Sanchez (jes@rednsi.com) wrote: > : Hi, > > : Though NXHosting from a NS/OS host to a Windows NT machine is great (really > : great!), we've found problems when using international keyboards in the > : Windows side. Remote applications use the default american layout so most of > : the graphical symbols are misplaced and typing foreign characters is quite > : difficult. OTOH, Windows apps use the correct map, as also do OpenStep apps > : running locally.
From: rmcassid@uci.edu Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 1998 12:34:50 -0700 Organization: University of California, Irvine Message-ID: <rmcassid-0804981234500001@dante.eng.uci.edu> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> In article <B1511189-1AF6A@206.165.43.154>, "Lawson English" <english@primenet.com> wrote: >But shareware programmers often fill in the gaps that commercial software >leaves. Also shareware programmers learn the ropes via feedback from their >customers and are often high school/college students who later enter the >Mac programming field as professionals. > >There's one very, VERY important piece of software that wouldn't exist if >shareware wasn't easy to do on the Mac: > >Stuffit. > >Raymond Lau put himself through MIT by writing Stuffit when he was 16 and >Alladin Software is a pretty important software house in the Macintosh >community.. > >Where's the next Stuffit or Alladin Software for MacOS going to come from >if there aren't going to be any more Raymond Lau's? Actually there are quite a few that fit into this category. However, there are also very few that I can think of off hand that didn't originate at an .edu either directly (as with Fetch, Newswatcher, etc.) or indirectly (as in the case of Stuffit). Education seems to be quite well taken care of these days by Apple. Apple has also opened up quite a bit to the developer by releasing a lot of their developer tools for free - don't want to pony up the $200 for MW - MPW is free now. So I don't think the shareware community will be as hard-hit as it might seem at first. They may not get everything that they might want, but then neither do any of us. Finally, I don't think that the Mac community would be SOL if Ray Lau never developed Stuffit. It had many competitors in the old days and Stuffit's success is due to being a very good shareware app that made a very good transition to commercial status. There is no saying that we wouldn't have some other app today that we'd be just as happy with - clearly the Wintel folk are happy with whatever they have, none of which are as good as Stuffit, IMO. -Bob Cassidy
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 01:48:06 GMT Organization: P & L Systems Message-ID: <6gh9cm$oop$7@ironhorse.plsys.co.uk> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ylaporte-0804982005030001@139.103.176.79> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: ylaporte@acm.org In <ylaporte-0804982005030001@139.103.176.79> Yan Laporte wrote: > What we need from Apple is an idea of when and what we will have to pay > for the system in higher education institutions. And don't get me the > "you got to be carefull, it is not meant to be used by normal users for > now(...)" speech. We're a Computer Science department, we just crave for > that stuff and we have the brain to handle it. So what are the plans for > Rhapsody in Higher education institutions?? Anyone? I need more than > dumb hearsays from MOSR... There was no such as a Developper rebate or any > program available for us three weeks ago, is there now? > I'm not sure what I'm allowed to say here, but I've certainly seen some public announcements by Apple (the problem being the only place I know there's a copy is private), and have an idea from speaking with some people on the inside that Apple is very much aware of all the points you've made here, and will be adressing them. cf also the following. Best wishes, mmalc. Subject: Re: Apple Developer Connection Mailing Date: Wed, 8 Apr 1998 09:37:36 -0700 From: Jordan Dea-Mattson <jordan@apple.com> To: "Rhapsody Discussion List" <rhapsody@clio.lyris.net> [...] In addition, we are working to roll out programs for the academic world - and have said so for some time. They weren't ready and this time and I can't comment on them, except to say that we are aware of the need to reach the academic world. Yours, Jordan ===== Jordan J. Dea-Mattson Senior Partnership & Technology Solutions Manager Apple Developer Relations Apple Computer, Inc. 1 Infinite Loop, MS: 303-2EV Cupertino, CA 95014
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 01:40:49 GMT Organization: P & L Systems Message-ID: <6gh8v1$oop$6@ironhorse.plsys.co.uk> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: macghod@concentric.net In <01bc4540$d4a15e60$3af0bfa8@davidsul> "macghod" wrote: > > > The lowest paid developer status is $500 instead of $250. If you really > > need to get the beta OS versions, your cost went up by $250. But everyone > > else seems to be better off. > > More Ragosta FUD. "Everyone else seems to be better off". Who is better > off? > Apple's customers who no longer have to subsidise hobbyist developers. Best wishes, mmalc.
From: rmcassid@uci.edu Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 1998 15:09:16 -0700 Organization: University of California, Irvine Message-ID: <rmcassid-0804981509160001@dante.eng.uci.edu> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> In article <steve-0804981442590001@oranoco.discoverysoft.com>, steve@discoverysoft.com (Steven Fisher) wrote: >In article <B1511019-15903@206.165.43.154>, "Lawson English" ><english@primenet.com> wrote: > >>Trying to make developer *support* a profit center is an INSANE idea. >>Hopefully, either Jobs will change this, or the more rational stockholders >>will hear about it in time to vote against him because this one action, *BY >>ITSELF*, will destroy Apple. > >That's the best summary I've seen yet. The real worry is that it will do >so *SLOWLY*... consumers probably won't see the effects this has on the >development community for months or a year yet. Suddenly new >shareware/freeware will start getting a lot more scarce. Of course the counter-argument is that Apple has never been able to really adequately support it's developers due to the low cost of the program. While I'm sure Adobe gets a lot of attention and I get a lot for what I pay, what about the guys in the middle - like Rich Siegal, perhaps. Or Aladdin Sys. The devlopers in the middle could probably handle much more in the way of devloper support and can quite likely afford a modest increase in costs. The question becomes: Will Apple take the fee increase as a profit, or reinvest it in their developer relations? Given the fact that the Mac software community is shrinking in the *middle*, not at the low end or high end quite so much, I think it might be a good plan if reinvestment is the goal. -Bob Cassidy
From: "Brad Hutchings" <brad@hutchings-software.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 98 19:36:33 -0700 Organization: EarthLink Network, Inc. Message-ID: <B151824F-7635E@207.217.155.14> References: <joe.ragosta-0804982109560001@elk62.dol.net> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit nntp://news.earthlink.net/comp.sys.mac.oop.powerplant, nntp://news.earthlink.net/comp.sys.mac.programmer.codewarrior, nntp://news.earthlink.net/comp.sys.next.advocacy, nntp://news.earthlink.net/comp.sys.next.programmer Joe Ragosta wrote: >Anyone notice that people like Mr. brown here, who would apparently never >go near a Mac from reading his posts, are the ones complaining while >developers like mmalc are supporting the change? "Mr. brown" as you call him is a founding member of CE Software. You know, QuicKeys (Eddie finalist), QuickMail, ... That would make him a real Mac developer. mmalc, on the other hand (and to be a little picky here), is one of our new NeXT friends. Believe me, if I were the slightest bit worried about Apple's direction w.r.t. the technology I totally depend on for the future of my business, I'd be kissing every butt at every turn. Not that I'm accusing mmalc or the rest of the NeXT contingent of doing that, I'm just saying that's what _I_ would do that if _I_ were a Rhapsody developer today. With a number of "legitimate" commercial developers shaking their heads about the hardware thing and the price increase, I'd do everything to encourage Apple management to have a positive view of what I'm doing. Frankly, the new program lineup makes a great deal of tactical sense. Apple Developer Relations (ADR) is playing the hand dealt to it the best possible way it can. Hardware purchasing has been a broken system for a long time. Any developer could give you a recent horror story. There is no longer any need to have the latest, greatest high end Mac to write software. Hardware costs, even post-cloning, are low enough, and channel margins low enough, that any serious hobbyist with a dream can enter. For any Mac developer with a clue, the Select program is really $400 (how can you write software without CodeWarrior?) and the Premier program is closer to $2000 (WWDC and Metrowerks rebates). But ADR is playing in a context where cynicism is running high. "Developer Relations" is about more than just herding the developer cattle effectively and meeting their product/service needs. It's about creating an atmosphere where the developers can succeed. It's about being trustworthy. It's about being dependable. It's about working to minimize the impact on developer partners of necessary direction changes or shifts. It's about the Advisor/Interim CEO not badmouthing Apple technologies that developers depend on. It's about not irritating customers. It's about the things that only a few at the top of Apple's management have the power to set as a priority. Delegating tactical control to a division and calling that developer relations, no matter how competently that division makes changes and does its jobs, rings hollow. *** I just saw Newt Gingrich on Larry King (whether you like Newt or not, the analogy is appropriate). Newt was talking about when Trent Lott came to him before a State of the Union address and urged him how to sit and dress so that he wouldn't look like a fat slob. Newt said that sometimes a friend needs to be critical. Well, Apple management, your developers are the best friends you have right now. The criticism comes from love: for Apple, its products, and its historic mission. Take that for what it's worth. Brad <mailto: "Brad Hutchings" brad@hutchings-software.com> <http://www.hutchings-software.com>
From: "macghod" <macghod@concentric.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 03:02:52 GMT Message-ID: <01bc455b$1be404c0$18f0bfa8@davidsul> References: <1d76snt.17ecjre1ykt2xdN@ppgutkneco.lirmm.fr> <B1516A43-1BD20@207.217.155.14> > Not in the US. That is total CYA BS. I know one US developer (not me) who > is damned thankful he got into the program and bought the G3s he needed > when he bought them. Their is talk about Apple being sued. Apparently people who are in the old $250 plan will be bumped to the $500 plan preety soon, and the $500 plan doesnt have the hardware discount, plus possible other things. I think it would be great if they get sued, Apple being the one whose legal department threatens web sites like macintouch, and they do something like this. I also hope several developers pass this on to the media, I am sure the media would love to flame apple for this and god knows apple deserves a good spanking for this
From: andrew_abernathy@omnigroup.com Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 02:56:29 GMT Organization: Omni Development, Inc. Message-ID: <6ghdct$3um$1@gaea.omnigroup.com> References: <1d76snt.17ecjre1ykt2xdN@ppgutkneco.lirmm.fr> <B1516A43-1BD20@207.217.155.14> "Brad Hutchings" <brad@hutchings-software.com> wrote: > >Another interesting information: the FAQ on developer.apple.com reads > >that Apple has been limiting access to the Associate Programs in the > >last six months... In the US only I guess [...] > Not in the US. That is total CYA BS. I know one US developer (not me) who > is damned thankful he got into the program and bought the G3s he needed > when he bought them. I guess they were inconsistent, then - my membership expired late last year and I got a letter from Apple saying "we're redoing our developer program; meanwhile consider your membership extended for free until we complete our changes." -- andrew_abernathy@omnigroup.com - NeXTmail & MIME ok
From: "Jason Swain" <jason-s@clear.net.nz> Newsgroups: comp.sys.next.programmer Subject: Help with library problem Date: Thu, 9 Apr 1998 15:20:53 +1200 Organization: CLEAR Net, http://www.clear.net.nz/ Message-ID: <6gheuq$dnp$1@fep5.clear.net.nz> I'm linking some c++ code (standard unix) and get the following errors: cc++ -Wall -Wno-unused -g -O2 -o example xample.o -I/Users/jason/Mirage/include -L/Users/jason/Mirage/lib -lm -lmira ge ld: multiple definitions of symbol _scalb /usr/lib/libm.a(support.o) definition of _scalb in section (__TEXT,__text) /lib/libsys_s.a(support.o) definition of absolute _scalb (value 0x5003794) ld: multiple definitions of symbol _copysign /usr/lib/libm.a(support.o) definition of _copysign in section (__TEXT,__text) /lib/libsys_s.a(support.o) definition of absolute _copysign (value 0x5003710) ld: multiple definitions of symbol _logb /usr/lib/libm.a(support.o) definition of _logb in section (__TEXT,__text) /lib/libsys_s.a(support.o) definition of absolute _logb (value 0x5003782) ld: multiple definitions of symbol _finite /usr/lib/libm.a(support.o) definition of _finite in section (__TEXT,__text) /lib/libsys_s.a(support.o) definition of absolute _finite (value 0x5003740) ld: multiple definitions of symbol _drem /usr/lib/libm.a(support.o) definition of _drem in section (__TEXT,__text) /lib/libsys_s.a(support.o) definition of absolute _drem (value 0x5003722) Does anyone know how I can fix this, it seems to be a conflict between two standard libraries. By the way I am using NS 3.3 with 3.2 Developer, could this be my problem? Thanks in advance. Jason
From: kdb@pegasus.ece.utexas.edu (Kurt D. Bollacker) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Followup-To: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Date: 9 Apr 1998 03:27:12 GMT Organization: The University of Texas at Austin, Austin, Texas Message-ID: <6ghf6g$e1u$1@geraldo.cc.utexas.edu> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <6ggs48$oop$2@ironhorse.plsys.co.uk> mmalcolm crawford (malcolm@plsys.co.uk) wrote: : In <xhsof-0804981448100001@hobbit1.injep.fr> Xavier Humbert wrote: : > If developers have to pay 500 bucks to have Rhapsody, do you think the : > will develop for it ? : > : If anyone is wanting to develop for Rhapsody professionally and doesn't have : $500, they don't have a business plan. If you're intending to write : shareware, then $500 is a reasonable investment if you believe you have : something useful to contribute. If you're writing Rhapsody apps for a hobby, : why should I subsidise your playtime? Because I might develop some useful freeware. With your attitude there'd never be any freeware at all. Without freeware, perhaps NeXT would have been later with NeXTSTEP (no GNU tools jumpstart) and Rhapsody today would not exist. Not all software should be free, but some of it always should. : In the bad old days when NeXT's developement tools alone cost seven times : that amount there was no shortage of freeware or shareware. Bzzzt. Wrong, but thank you for playing. If you bought NeXTSTEP as an academic machine or later as an academic bundle CD-ROM, the development tools were free or *very* cheap. Remember, NeXT was orignially academic only. Lots of cool freeware came out very quickly. ...................................................................... : Kurt D. Bollacker University of Texas at Austin : : kdb@pine.ece.utexas.edu P.O. Box 8566, Austin, TX 78713 : :....................................................................:
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 03:43:20 GMT Organization: Digital Fix Development Message-ID: <6ghg4o$np$1@news.digifix.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> <6ggqha$9fk$1@news01.deltanet.com> In-Reply-To: <6ggqha$9fk$1@news01.deltanet.com> On 04/08/98, Scott Ellsworth wrote: >In article <6ggmk2$jdl$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: <snip> >>The Online program is free. It gets you the samples, and the >>manuals. It also gets you Code Warrior Lite. > >And it cannot easily be used by a startup trying to put product >together, nor can it be used by someone doing thier own private >skunk works project. Trust me on this - I want the Mac to be a >viable platform at QMS, which it is not. The only way I can >demonstrate the value of the platform is to purchase appropriate >Mac tools, build our products, and demonstrate that they are >interesting. Having early builds makes it more likely that I >will have a demo that means something, > If you're developing commercially, an investment of $42/month isn't that out of reach. >I know two other developers at former cross platform shops doing >much the same. God only knows if this will work, but it is the >case that QMS has already decided to drop Mac development, >so they are a lost cause if I cannot change that. > If they're being so petty as to not be willing to pay the $500, then they should talk to Apple. Although realistically I know of several large commercial developers who never take advantage of Apple's programs, and use CodeWarrior for all development. >> From that you can learn Macintosh programming and put out >>freeware apps. > >True enough. Of course, your freeware and shareware will need >modification after the OS release happens. Further, certain >small scale authors push an OS a hell of a lot more than many >commercial app developers do. > Yes, you're freeware/shareware apps will need modification. However, in the case of those types of apps, you don't have the distribution costs involved in commercial products. An update is a matter of dropping it on an FTP server or WWW site. >> Few, if any, of the people pissing and moaning about loosing >>access to the seeding program actually NEED access to it in many >>cases. > >Of course, they changed the rules in mid year, such that anyone >already paying for a developer subscription got substantially >fewer benefits. (Specifically, dropping the hardware agreement >unlaterally is far from a nice move, and changes the economic >impact of the developer program substantially.) Have you purchased under this program? You can get virtually equivalent prices elsewhere. > >> Now, as far as Rhapsody is concerned it is leaving a gap, but >>only because of the time-frame.... However if you join Select you're >>still covered. > >There will always be a gap. Apple is incapable of shipping something >with decent prerelease documentation in the New Regime, I suspect. >HFS+ might be a counterargument, but if there is no way to get a >Rhapsody seed without a substantial cash outlay, then you have >delayed the release of software using Rhapsody features until your >public ship date. ... if you aren't willing to spend the $500 for the Select program, and can't work with DR1 if you already have it. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 03:45:51 GMT Organization: Digital Fix Development Message-ID: <6ghg9f$o3$1@news.digifix.com> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> <j-jahnke-0804982141280001@192.168.1.3> In-Reply-To: <j-jahnke-0804982141280001@192.168.1.3> On 04/08/98, Jerome Jahnke wrote: <snip> >Now this is a bunch of bull. Many of us were Associates for one >reason and one reason only, access to cheaper equipment. Apple >does not lose money by selling me my one machine a year at a low >price. Perhaps. But they do still provide you with all the other stuff, the CD's, the mailings etc.. that does cost Apple money. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 03:46:45 GMT Organization: Digital Fix Development Message-ID: <6ghgb5$rn$1@news.digifix.com> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggiu7$auu$1@vixen.cso.uiuc.edu> <01bc452e$e8a3eac0$3df0bfa8@davidsul> In-Reply-To: <01bc452e$e8a3eac0$3df0bfa8@davidsul> On 04/08/98, "macghod" wrote: >> Although I disapprove of the move by Apple, I really question why a >> HS/college student would have to join the developer programs to write >> software. After all, they could just download MPW or buy CodeWarrior at a > >> huge discount, and learn from the vast amount of online resources at >> devworld.apple.com. Although I agree it might drive away small, >> commercial developers, for the student developer, even the previous $250 >> was too much. > >Really? >How does a college student who wants to develop apps for rhapsody do so >without spending the $250 (now $500)? >Thats the only way they can legitimately get rhapsody, plus other apple >beta softwares they may want to develop for. > > Apple has already stated that there are education plans underway, but not yet announced. There is also the University programs that Apple already has in place (the scholarship programs). -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: trumbull@cs.yale.edu (Ben Trumbull) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 00:39:09 -0400 Organization: Yale University Message-ID: <trumbull-0904980039090001@net44-223.student.yale.edu> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> In article <6gh8v1$oop$6@ironhorse.plsys.co.uk>, mmalcolm crawford <malcolm@plsys.co.uk> wrote: > Apple's customers who no longer have to subsidise hobbyist developers. Apple never "subsidized" hobbyists. You don't think Apple made money off of every single on of them ? As someone who paid for nearly the entire IM series, I know Apple's gotten my money. Not to mention the Macs I've bought (not at developer discounts) which makes me an Apple customer. And I'm sure they made a profit off every developer discounted machine. Heck, it beats having a retailer sell someone that machine. Apple's got bigger problems than whether they make a small or large profit off their developers. It's like the government. As if cutting education spending and funding to the arts could possibly make a difference in the deficit when social security, welfare, and the miltary combind are about 3 orders of magnitude more money. Failing to attract new marketshare is just a whee bit more problematic than the extra $250. terminally curious, Ben ___________________________________________________________________ Benjamin Trumbull trumbull@cs.yale.edu Yale University You can't be in hell; you can still read your e-mail
Newsgroups: comp.sys.next.programmer From: Shinya Kominami <kominami@psrc.isac.co.jp> Subject: EOObserverCenter and WebObject Content-Type: text/plain; charset=iso-2022-jp Message-ID: <352B1F65.5EEE@psrc.isac.co.jp> Sender: usenet@psrc.isac.co.jp Content-Transfer-Encoding: 7bit Organization: ISAC, Inc. Mime-Version: 1.0 Date: Wed, 8 Apr 1998 06:55:33 GMT Hello, Does anyone know how to use EOObserverCenter with EOEditingContext and WODisplayGroup to handle EOEditingContextDidSaveChangesNotification? Thanks in advance... -- ------------------------------------------------------------------- ISAC, Inc. (A Subsidiary of Sekisui Chemical Corp.) Shinya Kominami Tel:(Tokyo)-3406-1645 Fax:(Tokyo)-3406-1751 mailto:kominami@psrc.isac.co.jp http://www.isac.co.jp
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Blocks for Obj-C Date: 6 Apr 1998 17:48:59 GMT Organization: Technical University of Berlin, Germany Message-ID: <6gb4ib$bdj$1@news.cs.tu-berlin.de> References: <6fokj8$1dp$1@usenet50.supernews.com> <6fpgbb$9o5$2@news.idiom.com> <6fphlp$iot$1@crib.bevc.blacksburg.va.us> <6fsfo4$ln@saturn.genoa.com> <6g8r4c$2uq$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit nurban@crib.bevc.blacksburg.va.us (Nathan Urban) writes: [ Nathan wants to extend protocols and attach implementations ] I think the whole issue could be handled by a rather tiny addition to Objective-C: typing the receiver of a method. The receiver (self) is an implicit argument to the method, and currently is also implicitly typed to be an instance of the class the method is defined in (in an instance method, of course...). My idea (adapted from a paper on the specialization interface) is to allow the 'self' argument to be typed as well. Of course, the only type that makes sense is a restriction to a protocol the class adopts. This makes it possible to specify 'core' methods with unrestricted access to the object as well as service methods that only access the class through a specified protocol, making it easier to know which methods have to be re-implemented when creating a subclass and which will 'just work'. This 'receiver-type' could be an optional part of a category declaration: @interface <ReceiverProtocol>Class(Category) If Class does not implement the 'ReceiverProtocol' a type-error is generated. If a receiver protocol is not specified, unrestricted access to the base class is assumed (as with the methods in the primary @implementation section), allowing for full backwards-compatibility. It should be obvious now how 'protocol categories' can be implemented, at least from a typing perspective: @interface <CoreProtocol>(Extension)<ExtendedProtocol> This would mean that any object conforming to 'CoreProtocol' could be extended with this Extension, and would then conform to 'ExtendedProtocol'. The only remaining issue is where to attach the actual implementations. Attaching them to NSObject is less of a problem than before because typing will not show NSObject as conforming to the protocol, but the methods will still be there and callable. Otherwise, the methods could be attached manually or semi-automatically to all classes implementing the base protocol, though that would probably be a bit of effort. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: Charles Swiger <chuck-nospam@blacksmith.com> Newsgroups: comp.sys.next.programmer Subject: Re: Q: Must I installed DeveloperDocs? Date: 6 Apr 1998 18:18:51 GMT Organization: BLaCKSMITH, Inc. Message-ID: <6gb6ab$kc8$8@anvil.BLaCKSMITH.com> References: <6g9ns1$agr$1@ha2.rdc1.nj.home.com> nospam+yes_this_is_a_valid_address@luomat.peak.org (Timothy Luoma) wrote: > I'm looking to save diskspace. I don't even refer to the DeveloperDocs, do I > need to have them installed? Nope. You can refer to that material off the CD-ROM if you only need the docs infrequently. -Chuck Charles Swiger | chuck@BLaCKSMITH.com | standard disclaimer ---------------+----------------------+-------------------- "Microsoft: we make the easy almost impossible."
From: leeyj@seri.re.kr (Young-Jin, Lee) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Help Registration method.... Date: 8 Apr 1998 07:39:04 GMT Organization: HPC Appl. Lab/SERI Sender: -Not-Authenticated-[3150] Message-ID: <6gf9io$861$1@green.kreonet.re.kr> XDisclaimer: User not authenticated Hi, All I've met a link error, but I don't understand what it is. I'll wait for your answer. Here is link error message. declaration syntax error : CControlsApp.cp line 57 TRegistar<LToggleButton>::Register() CControlsApp is the class name I made. I also included "LToggleButton.h" and put class registration code in the constructor of CControlsApp class like this. ... RegisterClass_( LToggleButton ); ... Is there any mistake? If so, let me know.
Date: Wed, 08 Apr 1998 19:47:45 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-0804981947450001@wil49.dol.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <6ggs48$oop$2@ironhorse.plsys.co.uk> <6ghf6g$e1u$1@geraldo.cc.utexas.edu> <6gi8o4$oop$9@ironhorse.plsys.co.uk> In article <6gi8o4$oop$9@ironhorse.plsys.co.uk>, mmalcolm crawford <malcolm@plsys.co.uk> wrote: > > Given that the cost of machines has reduced, all in all it looks to me like > it's going to be cheaper for academics now than it ever has been. Yeah. And they don't have to walk barefoot 12 miles uphill each way to get to school like we did, either. <grin> -- Regards, Joe Ragosta http://www.dol.net/~Ragosta/complmac.htm
From: don.brown@cesoft.com (Donald Brown) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 8 Apr 1998 23:05:47 -0500 Organization: CE Software Message-ID: <MPG.f960500e190219d989895@news.supernews.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> In article <joe.ragosta-0804982109560001@elk62.dol.net>, joe.ragosta@dol.net says... > > Including Microsoft, who looks better and better every day. > > Anyone notice that people like Mr. brown here, who would apparently never > go near a Mac from reading his posts, are the ones complaining while > developers like mmalc are supporting the change? > Joe, are you mildly familiar with QuicKeys, QuickMail, MockPackage, or a few dozen other Mac products I've written (directly or been on the team)? Right now I'm working on an instant messaging type program for small intranets (QuickConference IP). I'm writing the Mac and Windows clients. I expect to be programming on both platforms as long as there is a viable Mac platform. But, you're right, my personal machine is now my Micron Transport XPE, instead of my Powerbook 1400. That's because, when Apple killed CHRP as a way of making MacOS clones quickly with lots of variety, in spite of having sold it to the high heavens at WWDC a few months ago, made me finally decide that Apple just cannot be relied on. That's also when I started learning Windows programming (QCIP will be my first commercial Windows product). The CHRP fiasco was what made me decide I can no longer afford to link my career solely to Apple. The way Newton was killed (very quickly and very soon after Jobs sent emails to Newton developers reassuring them that not spinning Newton off was a way to assure Newton's continued existance) and now this have made me quite sure I made the right decision. I used to be the epitome of the developer who would bleed in six colors. It was very difficult to make the change. I switched my personal machine to be a Windows machine so I'd have to know what it's like to be a Windows user in order to deal with the issues. I still encounter times I wish I was running a Macintosh, but I also encounter times I am glad I'm running a Windows. Sorry, but Windows doesn't such. It was a hello of a surprise to me too. I still am hoping and praying that Apple makes it big. I still love the product. And I recognize that there are hard decisions that need to be made and this might even be one of them. But, the way they did it sends a big message to at least the small to medium sized developer "We don't care about you." Well, Microsoft never claimed to care about us, but they haven't broken any promises or revised our contracts while they're still in effect lately. There will be some number of developers giving up on Apple and switching to Windows because of this action. I don't know how many, but in Apple's weak position, any is too many. Donald
From: tom_e@usa.net (Thomas Engelmeier) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 9 Apr 1998 15:26:36 +0200 Organization: University of Rostock Message-ID: <1d77rmi.1x8sxa31jbipc0N@desktop.tom-e.private> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gecns$f9p$1@news01.deltanet.com> <1d767ho.1tixfqy1dqgm03N@hobbit1.injep.fr> <6gg6af$fit$1@anvil.BLaCKSMITH.com> <6ggabn$rts$2@news01.deltanet.com> <6ggj21$fit$6@anvil.BLaCKSMITH.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Charles Swiger <chuck-nospam@blacksmith.com> wrote: > The people for whom this price change matters are individuals and > students. And, to be blunt, these people are the ones who are unlikely to > have the resources (unused machine(s) to dedicate to Rhapsody, a > functioning LAN with a firewall [to meet the RDR1 licensing terms], and a > working fileserver and preferably NetInfo server) to provide Apple with > the kinds of feedback that they could use. That it's unlikely to find individuals with a really cheap firewall for a Mac set up (besides vicom internet gateway) might be due to a huge delay in OT´s product evolution. On the WWDC two (or three?) years ago it was announced routing in OT would follow soon. Now it seems to be there - undocumented. I bet if I used one of my yearly Q's to devsupport I wouldn't gain any information how to use it. WinProxys / Gatekeeper or routers with IP-Masquerading and basic firewall services are in my experience pretty popular for individual developers who run their SOHO network and try to sqeeze out the most from each platfor they use. If you continue calculating as a businessman, it`s not the time to let your highly paid employee to fiddle around with a nearly unsupported, buggy operating system nobody really knows when it will released and how big the customer base will be. Not for a small company who has to keep their development power together and especially not for a large company where peacounters, not nerds make decisions. The chances some hacking folks with linux / BSD experience spend their spare time to track down some of the DR´s oddities and create a valuable bug report are much higher, even though that would imply there is some interest in tracking down a bug. (Which is - according to my experience - not the case) Regards, TomE
From: don.brown@cesoft.com (Donald Brown) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 8 Apr 1998 23:10:11 -0500 Organization: CE Software Message-ID: <MPG.f96061a156c1db989896@news.supernews.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <352AAF57.25EB@skdjsk.no> <joe.ragosta-0904980640290001@elk33.dol.net> In article <joe.ragosta-0904980640290001@elk33.dol.net>, joe.ragosta@dol.net says... > In article <352AAF57.25EB@skdjsk.no>, sdd wrote: > > > I want a student developers program... > > > > 150-200 $/year for cd's... > > > > Anonymous.. > > Apple has publicly stated that there will be an attractive educational > program--it's just not ready yet. Another case where Apple does not know the first rule of wing-walking. Don't let go of what you've got until you've got hold of something else. Donald
Newsgroups: comp.sys.next.programmer,comp.sys.next.advocacy,comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.mac.oop.powerplant From: mooncake@cs.toronto.edu (Jeff Tupper) Subject: Re: Apple developer program Message-ID: <1998Apr9.093759.86@jarvis.cs.toronto.edu> Organization: CS Lab, University of Toronto References: <352b5da2.0@206.25.228.5> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> <j-jahnke-0804982141280001@192.168.1.3> Date: 9 Apr 98 13:38:00 GMT In article <j-jahnke-0804982141280001@192.168.1.3>, Jerome Jahnke <j-jahnke@uchicago.edu> wrote: >In article <rmcassid-0804981509160001@dante.eng.uci.edu>, rmcassid@uci.edu >wrote: > >Now this is a bunch of bull. Many of us were Associates for one reason and >one reason only, access to cheaper equipment. Apple does not lose money by >selling me my one machine a year at a low price. And I can have high end >hardware at a reduced cost so that I can continue to develop high quality >Apple Software. I didn't call them to ask them to help me with MacOS >problems we have newsgroups for that, and I participate here with others >to help people along. 250 bucks is cheap to send me 20 CD's a year and >give me early access to new Apple Technology. > >What they have done is doubled the price given me a feature I don't and >probably won't use, taken away the one I do use and didn't even give me a >choice to get my money back. I would agree with everyone here who supports >Apple on this decision if they had said. OK, all old Associates it is >business as usual. Next year however, the price is going to double and you >are going to get less. > >How much money is Apple saving by keeping me from buying a cheap machine >this year? > >Jer, ditto. We have thousands of users (paying between $22 and $1200 per license), and we can afford $500 a year (or even $3500 per year). We only buy one machine a year, and it was nice getting it from Apple (selection isn't so great up here in Canada) along with CDs. But, without the hardware discount, I don't see why we should stay in the program (for $500) since I can just buy 8.2 when it comes out. We don't need all the prelease software we get, and the SDKs can be downloaded off the net. As for $3500, I would rather apply that to the cost of a machine. As for techinical help, I have never asked for that, although I have answered questions others have asked me about development. The net effect of the change is to probably get us to no longer be an Apple developer (officially) and to not look at Rhapsody for a while longer... It saddens me, since I have enjoyed being an official Apple developer. I would stay in if the $500 program included a hardware discount on one machine. Not sure what to use my two questions on though :) even with hundreds of thousands of lines of commercial code... I am wondering how this will affect ESP (since we are an ESP member as well). [ESP = Apple's Educational Solution Provider program] Jeff Tupper www.peda.com
Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> From: John Kheit <jkheit@xtdl.com> Organization: monoChrome, Inc., NJ, USA Message-ID: <352b5da2.0@206.25.228.5> Date: 8 Apr 98 11:21:06 GMT j-jahnke@uchicago.edu (Jerome Jahnke) wrote: > In article <trumbull-0704981900290001@net44-223.student.yale.edu>, > trumbull@cs.yale.edu (Ben Trumbull) wrote: > > I'm not about to stand up and say these changes are a great > > idea on Apple's part, but you do realize, you can buy *just* > > the tech mailing (including develop) for $199 ? > Which is what we used to do, but then realized you could not get > seeded unless you were an associate. What's seeded mean? Does that mean no beta's? -- Thanks, be well, take care, later, John Kheit; self expressed... __________________________________________________________________ monoChrome, Inc. ASCII, MIME, PGP, SUN, & NeXTmail OK NeXT/OPENSTEP Developer mailto:jkheit@xtdl.com Telepathy, It's coming... http://www.xtdl.com/~jkheit Franklin Pierce Law Center You're dangerous because you're honest
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 14:46:40 GMT Organization: P & L Systems Message-ID: <6gin0g$oop$10@ironhorse.plsys.co.uk> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <6ggvub$oop$5@ironhorse.plsys.co.uk> <352CA831.D4878417@ArtQuest.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: hh@ArtQuest.fr In <352CA831.D4878417@ArtQuest.fr> Hubert HOLIN wrote: > mmalcolm crawford wrote: > [SNIP] > > If you're a professional developer, or intent on producing shareware that's > > of genuine utility (i,e, good enough that people will pay for it), $500 > > should not be a serious impediment, else your business plans are out of > > kilter. > [SNIP] > By that measure, freeware is junk, right? I guess TeX, GNU and their ilk are > unknown to you... > Absolutely not; where did I say that? Please do not put such gross insults into my mouth. mmalc.
Date: Wed, 08 Apr 1998 22:50:58 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-0804982250580001@wil62.dol.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> <MPG.f960500e190219d989895@news.supernews.com> In article <MPG.f960500e190219d989895@news.supernews.com>, don.brown@cesoft.com (Donald Brown) wrote: > In article <joe.ragosta-0804982109560001@elk62.dol.net>, > joe.ragosta@dol.net says... > > > Including Microsoft, who looks better and better every day. > > > > Anyone notice that people like Mr. brown here, who would apparently never > > go near a Mac from reading his posts, are the ones complaining while > > developers like mmalc are supporting the change? > > > Joe, are you mildly familiar with QuicKeys, QuickMail, MockPackage, or a > few dozen other Mac products I've written (directly or been on the team)? > > Right now I'm working on an instant messaging type program for small > intranets (QuickConference IP). I'm writing the Mac and Windows clients. > I expect to be programming on both platforms as long as there is a viable > Mac platform. > > But, you're right, my personal machine is now my Micron Transport XPE, > instead of my Powerbook 1400. That's because, when Apple killed CHRP as a > way of making MacOS clones quickly with lots of variety, in spite of > having sold it to the high heavens at WWDC a few months ago, made me > finally decide that Apple just cannot be relied on. That's also when I > started learning Windows programming (QCIP will be my first commercial > Windows product). > > The CHRP fiasco was what made me decide I can no longer afford to link my > career solely to Apple. The way Newton was killed (very quickly and very > soon after Jobs sent emails to Newton developers reassuring them that not > spinning Newton off was a way to assure Newton's continued existance) and > now this have made me quite sure I made the right decision. > > I used to be the epitome of the developer who would bleed in six colors. > It was very difficult to make the change. I switched my personal machine > to be a Windows machine so I'd have to know what it's like to be a > Windows user in order to deal with the issues. I still encounter times I > wish I was running a Macintosh, but I also encounter times I am glad I'm > running a Windows. Sorry, but Windows doesn't such. It was a hello of a > surprise to me too. > > I still am hoping and praying that Apple makes it big. I still love the > product. And I recognize that there are hard decisions that need to be > made and this might even be one of them. But, the way they did it sends > a big message to at least the small to medium sized developer "We don't > care about you." > > Well, Microsoft never claimed to care about us, but they haven't broken > any promises or revised our contracts while they're still in effect > lately. There will be some number of developers giving up on Apple and > switching to Windows because of this action. I don't know how many, but > in Apple's weak position, any is too many. I apologized to Donald in e-mail for the strength of my comments. But I don't see anything here that changes the essence. As I see it (and these comments reinforce it), Donald is very much acting in the Lawson mold--Apple is guilty until proven innocent. He's measuring everything by _HIS_ standards and finds Apple at fault every time they don't meet his expectations. -- Regards, Joe Ragosta http://www.dol.net/~Ragosta/complmac.htm
Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <01bd6267$fa532b30$9b2168cf@test1> From: John Kheit <jkheit@xtdl.com> Organization: monoChrome, Inc., NJ, USA Message-ID: <352ad905.0@206.25.228.5> Date: 8 Apr 98 01:55:17 GMT "Todd Heberlein" <todd@NetSQ.com> wrote: > > Agreed. I might have considered getting Rhapsody under the > > previous program. I might be able to scrounge up $250. $500 > > is out of my price range. I guess they didn't solicit input > > from graduate students when forming their new policies. I'm > > hardly an important developer for Apple, and am > As a student, you will probably be able to get the full development > environment quite inexpensively from Apple, but we will have to > wait a while for Apple to make that move. The last time I received > word from Apple (less than a month ago), they had not hammered > out all the license deals with third parties. > Also, $500 is still quite cheap. As I mentioned in another > thread, yesterday I paid $1300+ for a license to use Digital's > command-line debugger (which is pretty lame). > I am also a Sun Catalyst member, but I still pay a fairly high > price for Sun's development tools. Yea, and look at the marketshare and number of developers for that platform vs. the mac. I don't know the details of this program change yet, so I'll reserve final judgement till then. But *if* they cut services and doubled the costs to developers, I'd say this is the proud continuation of BOZO behavior at Apple. Clearly not everyone in the developer programs makes killer apps, but the more developers you allow access, the more likely more developers will make killer apps. *If* Apple has decided in its avarice that it only needs a certain type of wealthy developer, then it deserves all the negative consequences that follow. Namely, a loss of developers, apps, and marketshare. -- Thanks, be well, take care, later, John Kheit; self expressed... __________________________________________________________________ monoChrome, Inc. ASCII, MIME, PGP, SUN, & NeXTmail OK NeXT/OPENSTEP Developer mailto:jkheit@xtdl.com Telepathy, It's coming... http://www.xtdl.com/~jkheit Franklin Pierce Law Center You're dangerous because you're honest
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 15:06:30 GMT Organization: P & L Systems Message-ID: <6gio5m$oop$11@ironhorse.plsys.co.uk> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <trumbull-0904980039090001@net44-223.student.yale.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: trumbull@cs.yale.edu In <trumbull-0904980039090001@net44-223.student.yale.edu> Ben Trumbull wrote: > In article <6gh8v1$oop$6@ironhorse.plsys.co.uk>, mmalcolm crawford > <malcolm@plsys.co.uk> wrote: > > > Apple's customers who no longer have to subsidise hobbyist developers. > > Apple never "subsidized" hobbyists. You don't think Apple made money off > of every single on of them ? > No, I don't. I'm quite sure there are a number of hobbyists developing worthwhile s/w which I might find useful at some point, however on balance reports suggest that the previous system was badly abused and on average Apple will have lost revenue through people signing up as developers simply to get hardware at reduced prices. If you're a real hobbyist, do you *need* access to the bleeding edge s/w? If not, then you don't need to belong to the program, so it won't cost you anything. If you think you do need access to bleeding edge s/w, then that's a fairly serious hobby, and I think it's reasonable for people to invest in it. My main hobby is acting; in June I'll be playing George in "Same time next year" (I didn't have the legs to play Doris); incidental expenses including travel to and from rehearsals are likely to cost as much as subscription to the Select program for the next two months. Maybe I should send in my petrol receipts to Apple...? Does all this mean I don't appreciate the efforts of those who give their time? No. Does this mean I like the fact that something's going to cost more? No. Does this mean that I think Apple's taking a sensible business attitude to concentrating its resources where they matter? Yes. > Apple's got bigger problems than whether they make a small or large profit > off their developers. It's like the government. As if cutting education > spending and funding to the arts could possibly make a difference in the > deficit when social security, welfare, and the miltary combind are about 3 > orders of magnitude more money. Failing to attract new marketshare is > just a whee bit more problematic than the extra $250. > Indeed, which is why I'm pleased to see that Apple is investing significantly in what appears to be a very effective marketing campaign, which is raising awareness of the brand and the technologies it has to offer, and causing consumers to Think Different about the company. I also hope that the additional fees will go toward giving a better service to those who choose to stay in one of the developer programs. As for yourself; I notice you have a .edu address. Perhaps you should wait to see exactly what APple announces for academics before throwing too many stones... Best wishes, mmalc.
From: xhsoft@injep.fr (Xavier Humbert) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 8 Apr 1998 14:25:24 +0200 Organization: Maquis Usenet Ouest Message-ID: <1d767ho.1tixfqy1dqgm03N@hobbit1.injep.fr> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gecns$f9p$1@news01.deltanet.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Mail-Copies-to: never Scott Ellsworth <scott@eviews.com> wrote: > The seeding program, imho, served a valuable purpose, but > Apple, I think, does not want any leaks of its seeds, and so > it is trying to reduce the number of developers who might > have access. What is obviously stupid : WarEz sites have System Seeds BEFORE the official FTP :-( Xav, looking for a warez URL -- Xavier HUMBERT Laboratoire Informatique INJEP Office: labo-info/AT/injep.fr Home: humbert/AT/injep.fr
From: "macghod" <macghod@concentric.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 17:32:03 GMT Message-ID: <01bc45d4$8961c2c0$43f0bfa8@davidsul> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> <MPG.f960500e190219d989895@news.supernews.com> Donald Brown <don.brown@cesoft.com> wrote in article <MPG.f960500e190219d989895@news.supernews.com>... > In article <joe.ragosta-0804982109560001@elk62.dol.net>, > joe.ragosta@dol.net says... > > > Including Microsoft, who looks better and better every day. > > > > Anyone notice that people like Mr. brown here, who would apparently never > > go near a Mac from reading his posts > Joe, are you mildly familiar with QuicKeys, QuickMail, MockPackage, or a > few dozen other Mac products I've written (directly or been on the team)? Yup, just Joe Ragosta living up to his nickname of MR FUDMEISTER. Peter was right, Joe is logical, a logical FUDster
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 17:33:15 GMT Organization: P & L Systems Message-ID: <6gj0or$oop$12@ironhorse.plsys.co.uk> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <352AAF57.25EB@skdjsk.no> <6geche$8lu$1@news2.apple.com> <philipm-0804982051170001@pm1-23.eecs.umich.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: philipm@NO-SPAMeecs.umich.edu In <philipm-0804982051170001@pm1-23.eecs.umich.edu> Philip Machanick wrote: > In article <6ggvub$oop$5@ironhorse.plsys.co.uk>, mmalcolm crawford > <malcolm@plsys.co.uk> wrote: > > >Apple acting as a charity subsidising people's hobbies is what's insane. > >Nor should they be offering a program wide open to abuse by people signing up > >as developers just to get a discount on hardware. > > I don't think Apple is subsidising developers. Without developers, no one > would buy Macs (they tried that experiment in 1984). > By offering a hardware discount of course Apple was subsidising developers -- that's what the discount was for. Now they've raised the bar to ensure in part that the system isn't abused. For the sorts of commercial applications that I suspect you're referring to, on which the platform depends, if the business seriously needs access to seed software but cannot afford a $500 a year, it does not have a competent business plan. > Anyway what's wrong > with having a sneaky back door way of cutting dealers out of the loop? > A "sneaky back door way" is called screwing the channel. I suspect this is not what you meant. > What has a dealer ever done for you? > Actually my local Apple VAR in Sheffield was *extremely* helpful. Trimac, for anybody based in Yorkshire: 0114 272 4127 (I even remember their phone number). > You don't think Apple loses on hardware sold at developer prices? > Compared to hardware sold at retail prices, yes. mmalc
Newsgroups: comp.sys.next.programmer Subject: Re: Q: Must I installed DeveloperDocs? References: <6g9ns1$agr$1@ha2.rdc1.nj.home.com> From: John Kheit <jkheit@xtdl.com> Organization: monoChrome, Inc., NJ, USA Message-ID: <3529c75c.0@206.25.228.5> Date: 7 Apr 98 06:27:40 GMT nospam+yes_this_is_a_valid_address@luomat.peak.org (Timothy Luoma) wrote: > I'm looking to save diskspace. I don't even refer to the > DeveloperDocs, do I need to have them installed? I basically > just compile stuff, not create, and I'm thinking that the Docs > are just for folks who want reference material. > Will not installing them cause any problems for someone who just > does compiles? Yes, everything will work even with them uninstalled. -- Thanks, be well, take care, later, John Kheit; self expressed... __________________________________________________________________ monoChrome, Inc. ASCII, MIME, PGP, SUN, & NeXTmail OK NeXT/OPENSTEP Developer mailto:jkheit@xtdl.com Telepathy, It's coming... http://www.xtdl.com/~jkheit Franklin Pierce Law Center You're dangerous because you're honest
Date: Wed, 08 Apr 1998 08:58:15 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-0804980858150001@wil77.dol.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> In article <ericb-0704981749520001@132.236.171.104>, ericb@pobox.com (Eric Bennett) wrote: > In article <01bc445f$744397e0$1bf0bfa8@davidsul>, "macghod" > <macghod@concentric.net> wrote: > > > WTF is up with the apple developer program? I see a big announcement, and > > all it is is: > > 1st level: online stuff, just as it is now, ie NO CHANGE > > 2nd level: instead of costing $250 a year its now $500 a year??!?!?? > > > > As far as I can tell it adds nothing and is a 2x increase!! Why to support > > development apple :) > > Agreed. I might have considered getting Rhapsody under the previous > program. I might be able to scrounge up $250. $500 is out of my price > range. > > I guess they didn't solicit input from graduate students when forming > their new policies. I'm hardly an important developer for Apple, and am > not likely to become one, but other people like me might (Aaron Giles > comes to mind as a recent example). :-| Actually, there were rumors that there would be a very low price educational plan. You might want to wait for the educational pricing before worrying too much. -- Regards, Joe Ragosta http://www.dol.net/~Ragosta/complmac.htm
Message-ID: <352D35D6.E368D52C@am.not.here> Date: Thu, 09 Apr 1998 13:55:51 -0700 From: Alias <i@am.not.here> Organization: Nowhere MIME-Version: 1.0 Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program References: <01bc445f$744397e0$1bf0bfa8@davidsul> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit My two bits... It seems that all the hub-bub is w.r.t. shareware and freeware developers. Now, I could be wrong, but it seems to me that what Apple is saying is...If you want to develop commercial software, choose either our select or premier programs. Otherwise, use an educational institution to gain access to our products. macghod wrote: > WTF is up with the apple developer program? I see a big announcement, and > all it is is: > 1st level: online stuff, just as it is now, ie NO CHANGE > 2nd level: instead of costing $250 a year its now $500 a year??!?!?? > > As far as I can tell it adds nothing and is a 2x increase!! Why to support > development apple :)
From: steve@discoverysoft.com (Steven Fisher) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 12:08:12 -0700 Organization: Discovery Software Ltd. Message-ID: <steve-0904981208120001@oranoco.discoverysoft.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gecns$f9p$1@news01.deltanet.com> <1d767ho.1tixfqy1dqgm03N@hobbit1.injep.fr> <6gg6af$fit$1@anvil.BLaCKSMITH.com> <6ggabn$rts$2@news01.deltanet.com> <1d76snt.17ecjre1ykt2xdN@ppgutkneco.lirmm.fr> <1998040920073876968@dialup-202.def.oleane.com> In article <1998040920073876968@dialup-202.def.oleane.com>, Le_Jax@iName.com (Jacques Foucry) wrote: >Olivier Gutknecht <gutkneco+news@lirmm.fr> wrote (écrivait) : > >> Agreed. The only explanation is that Apple is not interested anymore in >> small/independant development groups. > >Perhaps, but without small/independant developer, no "Super Clock", no >"Stickies", no new Control Strip... The recent revisions of NotePad, Scrapbook and Find File, too. -- Steven Fisher; Discovery Software Ltd.; steve@discoverysoft.com "Anyone who has never made a mistake has never tried anything new." -- Albert Einstein
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 98 19:06:37 GMT Organization: QMS Message-ID: <6gj676$3qf$2@news01.deltanet.com> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> <j-jahnke-0804982141280001@192.168.1.3> In article <j-jahnke-0804982141280001@192.168.1.3>, j-jahnke@uchicago.edu (Jerome Jahnke) wrote: >In article <rmcassid-0804981509160001@dante.eng.uci.edu>, rmcassid@uci.edu >> Of course the counter-argument is that Apple has never been able to really >> adequately support it's developers due to the low cost of the program. >> While I'm sure Adobe gets a lot of attention and I get a lot for what I >> pay, what about the guys in the middle - like Rich Siegal, perhaps. Or >> Aladdin Sys. The devlopers in the middle could probably handle much more >> in the way of devloper support and can quite likely afford a modest >> increase in costs. > >Now this is a bunch of bull. Many of us were Associates for one reason and >one reason only, access to cheaper equipment.... >250 bucks is cheap to send me 20 CD's a year and >give me early access to new Apple Technology. Yep. Given that they have to produce the CDs and the seeds anyway, the marginal cost of sending them to a flock of Associates is fairly minor. (The one time I tried to call dev support, it took them three months to reply. This is one reason I did not try again.) >What they have done is doubled the price given me a feature I don't and >probably won't use, taken away the one I do use and didn't even give me a >choice to get my money back. I would agree with everyone here who supports >Apple on this decision if they had said. OK, all old Associates it is >business as usual. Next year however, the price is going to double and you >are going to get less. Very well stated. I would not be voting proxies against them, calling the credit card company to cancel the order, and accusing them of everything but battery, buggery, and brawling had they either given me a "quit now for a refund" option, or had they announced that this was to be the policy on my renewal in March of 1999. I would think it a foolish move on thier part, but that is a matter of argument, and they can look at the numbers for renewals and such to determine at least some of the effect as developer renwals come up. If it turns out badly, they might be able to recover after only a bad quarter, instead of having people annoyed and feeling like the contract was unilaterally altered without warning. >How much money is Apple saving by keeping me from buying a cheap machine >this year? Since I am reconsidering buying any Mac equipment until I see just how useful my current subscription is, about -$200 for the consumer machine I was going to buy as a present, and about -$100 for a development machine for me. Scott Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
From: "Brad Hutchings" <brad@hutchings-software.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 98 12:07:26 -0700 Organization: EarthLink Network, Inc. Message-ID: <B1526A8E-5D056@207.217.155.85> References: <6ghdct$3um$1@gaea.omnigroup.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit nntp://news.earthlink.net/comp.sys.mac.oop.powerplant, nntp://news.earthlink.net/comp.sys.mac.programmer.codewarrior, nntp://news.earthlink.net/comp.sys.next.advocacy, nntp://news.earthlink.net/comp.sys.next.programmer >I guess they were inconsistent, then - my membership expired late last >year and I got a letter from Apple saying "we're redoing our developer >program; meanwhile consider your membership extended for free until we >complete our changes." That would be a renew. The example I know of is a new membership. Brad <mailto: "Brad Hutchings" brad@hutchings-software.com> <http://www.hutchings-software.com>
From: "Keith Wood" <keith@cognisense.com> Newsgroups: comp.sys.next.programmer Subject: Re: Java virtual machine? NS3.3 Date: Thu, 9 Apr 1998 14:25:22 -0500 Organization: OnRamp, http://www.onramp.net/ Message-ID: <6gj7b2$3qo$1@news.onramp.net> References: <01bd5ebc$f718c460$230e8780@starlock.uchicago.edu> John Stiening wrote in message <01bd5ebc$f718c460$230e8780@starlock.uchicago.edu>... > > I was wondering if anyone has found a recent Java virtual machine for >nextstep. Ditto... Now that I have a fax machine, there's not been much activity on my old black hardware... 8) Keith Wood Cognisense keith@cognisense.com
From: "Keith Wood" <keith@cognisense.com> Newsgroups: comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software Subject: Netscape on NeXTSTEP Date: Thu, 9 Apr 1998 14:28:23 -0500 Organization: OnRamp, http://www.onramp.net/ Message-ID: <6gj7go$3r6$1@news.onramp.net> Has anyone tried to get Netscape's UNIX version to work under NeXTSTEP? I've got black hardware (color station) and NS 3.0... Please email me at keith@cognisense.com. Thanks for any help you can provide... Keith Cognisense keith@cognisense.com
From: trumbull@cs.yale.edu (Ben Trumbull) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 15:44:36 -0400 Organization: Yale University Message-ID: <trumbull-0904981544360001@net44-223.student.yale.edu> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <trumbull-0904980039090001@net44-223.student.yale.edu> <6gio5m$oop$11@ironhorse.plsys.co.uk> In article <6gio5m$oop$11@ironhorse.plsys.co.uk>, mmalcolm crawford <malcolm@plsys.co.uk> wrote: > No, I don't. I'm quite sure there are a number of hobbyists developing > worthwhile s/w which I might find useful at some point, however on balance > reports suggest that the previous system was badly abused and on average > Apple will have lost revenue through people signing up as developers simply > to get hardware at reduced prices. Sorry, but asserting that Apple doesn't make money off the books they sell to Barnes&Noble and the rest of the Addison&Wesley series to "hobbyist" developers is absurd. These are the same people who buy OS upgrades and IM cds and Macintosh computers. And there was a time when Apple actually sold development software. If Apple isn't making money off these transactions, raising the price of the dev program isn't going to help them. As for the "badly abused" hardware discount program, that's a load of baloney. I saw the prices and was never once tempted. Apple made money off every one of those machines and they sure made a lot more money than if some retailer sold the developer that machine instead of Apple. Which is exactly what will happen now. Devs will pay retailers for machines, devs will buy fewer machines overall, and Apple will see smaller margins overall. Yeah, this is a great solution to Apple's "problem" of selling more machines and not paying retailers for the priviledge. Let's get this straight. Apple makes A LOT MORE MONEY when they sell direct. A lot. The ideas that Apple should "raise the bar" to select against small developers and this will somehow make the Mac dev community healthier, or that Apple EVER was "subsidizing" small or independent developers are absurd. If Apple lost money on selling tons of development documentations, cds, software, and hardware (true for less, but no retailer biting into the margins), then frankly the sole reason is because they couldn't manage their business. Trying to blame people for "abusing" the dev program is just a bullshit excuse. However, to be fair to Apple, I haven't actually seen anyone from Apple make that argument, just people desperate to defend the policy changes. terminally curious, Ben ___________________________________________________________________ Benjamin Trumbull trumbull@cs.yale.edu Yale University You can't be in hell; you can still read your e-mail
From: don_arb@wolfenet.com (Don Arbow) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 12:05:01 -0700 Organization: EveryDay Objects, Inc. Message-ID: <don_arb-0904981205010001@sea-ts1-p14.wolfenet.com> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggiu7$auu$1@vixen.cso.uiuc.edu> <352c857a.2678525846@news.u-psud.fr> In article <352c857a.2678525846@news.u-psud.fr>, pointal@lure.u-psud.fr wrote: : On 8 Apr 1998 19:24:55 GMT, gcheng@uni.uiuc.edu (Guanyao Cheng) wrote: : : >: Although I disapprove of the move by Apple, I really question why a : >: HS/college student would have to join the developer programs to write : >: software. After all, they could just download MPW or buy CodeWarrior at a : >: huge discount, and learn from the vast amount of online resources at : >: devworld.apple.com. : : Do you know the bandwitdh of US servers out of the US? Do you know the : cost of communications out of the US? : : For a student working at home, CDs are far better with his <1 to 5 : kb/sec IP connection to US (at ~ $1/hour com cost + fixed provider : cost). : You can buy the CD's for $199, $50 less than it cost previously... Don -- Don Arbow, Partner, CTO EveryDay Objects, Inc. don_arb@wolfenet.com <-- remove underscore to reply http://www.edo-inc.com
From: don_arb@wolfenet.com (Don Arbow) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 12:04:14 -0700 Organization: EveryDay Objects, Inc. Message-ID: <don_arb-0904981204140001@sea-ts1-p14.wolfenet.com> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> <j-jahnke-0804982141280001@192.168.1.3> <6ghg9f$o3$1@news.digifix.com> <j-jahnke-0904980214320001@192.168.1.3> In article <j-jahnke-0904980214320001@192.168.1.3>, j-jahnke@uchicago.edu (Jerome Jahnke) wrote: : In article <6ghg9f$o3$1@news.digifix.com>, sanguish@digifix.com (Scott : Anguish) wrote: : : > On 04/08/98, Jerome Jahnke wrote: : > <snip> : > >Now this is a bunch of bull. Many of us were Associates for one : > >reason and one reason only, access to cheaper equipment. Apple : > >does not lose money by selling me my one machine a year at a low : > >price. : > : > Perhaps. But they do still provide you with all the other : > stuff, the CD's, the mailings etc.. that does cost Apple money. : : It don't cost 500 bucks... : : Jer, Nope, it's $199 for the CD's now... Don -- Don Arbow, Partner, CTO EveryDay Objects, Inc. don_arb@wolfenet.com <-- remove underscore to reply http://www.edo-inc.com
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 13:10:01 -0700 Organization: Primenet Services for the Internet Message-ID: <B1527A55-4B993@206.165.43.139> References: <6gj0or$oop$12@ironhorse.plsys.co.uk> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit mmalcolm crawford <malcolm@plsys.co.uk> said: > By offering a hardware discount of course Apple was subsidising > developers -- > that's what the discount was for. As long as the developers at least pay Apple what it costs Apple to make and ship the hardware, they are NOT subsidizing developers. They are offering enticements/inducements/whatevers, but they are NOT subsidizing developers. Now, they ARE subsidizing developers on the software-tools side, by not charging the full development costs for developing the SDKs and technical materials, but if they were to do that, we would be back in the bad ole days of Spindler, where an ETO subscription was many hundreds/year and you had to pay $1500 up front to get the complete package of MPW C/Pascal/C++/MacApp/etc. Now they've raised the bar to ensure in > part that the system isn't abused. For the sorts of commercial > applications > that I suspect you're referring to, on which the platform depends, if the > business seriously needs access to seed software but cannot afford a $500 > a > year, it does not have a competent business plan. You may be correct about raising the bar to see that the hardware discount system isn't abused. However, if they raise the bar sufficiently, then there is no longer a discount on the hardware in the first place, is there? --------------------------------------------------------------------- "OpenDoc may be impressive technology, but if no-one else is using it, why should Apple?" --Steve Jobs "In short: if you want to blow people away with what can be done *only* on a Mac, show them Cyberdog." --Tom Keyes <news://cyberdog.apple.com/cyberdog.general/19411> ---------------------------------------------------------------------
From: ericb@pobox.com (Eric Bennett) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 16:31:11 -0400 Organization: Cornell University Sender: emb22@cornell.edu (Verified) Message-ID: <ericb-0904981631110001@132.236.171.104> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> <MPG.f960500e190219d989895@news.supernews.com> <joe.ragosta-0804982250580001@wil62.dol.net> In article <joe.ragosta-0804982250580001@wil62.dol.net>, joe.ragosta@dol.net (Joe Ragosta) wrote: > As I see it (and these comments reinforce it), Donald is very much acting > in the Lawson mold--Apple is guilty until proven innocent. He's measuring > everything by _HIS_ standards and finds Apple at fault every time they > don't meet his expectations. I don't see any developers defending Apple's decisions. The best "support" I've seen is some Mac news sites parroting Apple's press release. It's quite obvious that Apple developers now pay more but receive less. And certainly having your contract altered would be upsetting. For example, suppose Apple promised long-term tech support on the machine you bought, then decided to truncate it to 90 days, after they had already promised longer support to you? Hmmm, Apple has already done *that* too. Apple is pissing developers off to save a few bucks. Bad idea. -- Eric Bennett (www.pobox.com/~ericb) Cornell University Biochemistry Department
From: "NeXT Newbie" <macghod@concentric.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 20:57:47 GMT Message-ID: <01bc462c$50a61be0$40f0bfa8@davidsul> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <352D35D6.E368D52C@am.not.here> Alias <i@am.not.here> wrote in article <352D35D6.E368D52C@am.not.here>... > My two bits... > > It seems that all the hub-bub is w.r.t. shareware and freeware developers. > Now, I could be wrong, but it seems to me that what Apple is saying is...If you > want to develop commercial software, choose either our select or premier > programs. Otherwise, use an educational institution to gain access to our > products. And how do you do this if you arent in a educational institution? You assume one would be but this is not a valid assumption
From: ericb@pobox.com (Eric Bennett) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 17:01:44 -0400 Organization: Cornell University Sender: emb22@cornell.edu (Verified) Message-ID: <ericb-0904981701440001@132.236.171.104> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> In article <B1511189-1AF6A@206.165.43.154>, "Lawson English" <english@primenet.com> wrote: > There's one very, VERY important piece of software that wouldn't exist if > shareware wasn't easy to do on the Mac: > > Stuffit. > > Raymond Lau put himself through MIT by writing Stuffit when he was 16 and > Alladin Software is a pretty important software house in the Macintosh > community.. > > Where's the next Stuffit or Alladin Software for MacOS going to come from > if there aren't going to be any more Raymond Lau's? There are lots of good examples like this. What free image viewer do legions of Mac users depend on? JPEGView. Who wrote it, where? Aaron Giles, while he was at Cornell University Medical School. He went on to help port Dark Forces to the Mac. What free list server can you get for the Mac? Macjordomo. Who wrote it, where? Michele Fuortes, also from Cornell University Medical School. Have any of you Mac users out there ever heard of a fellow by the name of Peter Lewis? He started writing Mac apps in his spare time while he was still employed at a university (do I sense a pattern here?). Quid Pro Quo also started out as a freeware app. Now it's commercial (though a free "lite" version is still available) and seems to compete pretty well with WebSTAR on a feature-by-feature level and price level (it doesn't get the publicity WebSTAR gets, though). Plenty of commercial or shareware apps had humble freeware beginnings. Other good software has always been freeware. Why charge these people $500 to get their hands on the latest system software? Do you want them all to tell users, "Sorry, my program doesn't work with OS 8.2, because Apple prices the developer versions out of my price range. I can't afford to test my freeware with 8.2 until *after* it comes out, and then start working on any necessary fixes." -- Eric Bennett (www.pobox.com/~ericb) Cornell University Biochemistry Department
From: ericb@pobox.com (Eric Bennett) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 17:05:11 -0400 Organization: Cornell University Sender: emb22@cornell.edu (Verified) Message-ID: <ericb-0904981705120001@132.236.171.104> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggrql$oop$1@ironhorse.plsys.co.uk> In article <6ggrql$oop$1@ironhorse.plsys.co.uk>, mmalcolm crawford <malcolm@plsys.co.uk> wrote: > Apple has said that the acedemic program is to be announced later -- I'd > expect it to be cheaper. Where have they explicitly stated this? Do you have a URL? All I have seen are rumors from "highly placed sources." And those rumors have been around for quite awhile without having come true. -- Eric Bennett (www.pobox.com/~ericb) Cornell University Biochemistry Department
From: ericb@pobox.com (Eric Bennett) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 17:07:16 -0400 Organization: Cornell University Sender: emb22@cornell.edu (Verified) Message-ID: <ericb-0904981707160001@132.236.171.104> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggiu7$auu$1@vixen.cso.uiuc.edu> <352c857a.2678525846@news.u-psud.fr> <joe.ragosta-0904980645050001@elk33.dol.net> In article <joe.ragosta-0904980645050001@elk33.dol.net>, joe.ragosta@dol.net (Joe Ragosta) wrote: > That is probably true. But wouldn't most students have access to their > college's facilities where they can download what they need by T1 and put > it onto a zip drive or burn a CD to take home? I can't download Rhapsody. Not getting Rhapsody is what really burns me. I am enthusiastic about it and I want to start learning to program on it. MacOS 8.2 is not such a big deal, since it's not really all that different from a basic programming perspective that the copy of MacOS 8.1 I'm running now. Not so with Rhapsody. -- Eric Bennett (www.pobox.com/~ericb) Cornell University Biochemistry Department
From: "NeXT Newbie" <macghod@concentric.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.programmer Subject: first day with OS Date: 9 Apr 1998 21:06:29 GMT Message-ID: <01bc462d$87ca8380$40f0bfa8@davidsul> Some impressions of OS 4.2 1) This OS sure was designed for big monitors! I have a 17 inch and it seems to small. Seems like you would want at least a 20 inch monitor 2) The mouse really annoys me. It reminds me of when I first used windows 3.1, I was like god I hate how jerky the mousee is. The mouse movement in OS 4.2 seems like how it was in windows 3.1, windows 95 definitely seems much better. 3) How do you connect to the internet with ppp? 4) once I get up and going with ppp, what do I use for usenet? 5) where is the option to change the monitors resolution and bits of color?
From: slick@tools.ecn.purdue.edu (Brian S Slick) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 21:01:27 GMT Organization: Purdue University, W. Lafayette, IN Message-ID: <6gjcv7$nrb@mozo.cc.purdue.edu> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <1d7821x.1rdi5101enh5fhN@hobbit1.injep.fr> <6girj3$6j$2@anvil.BLaCKSMITH.com> <ericb-0904981649370001@132.236.171.104> In article <ericb-0904981649370001@132.236.171.104>, Eric Bennett <ericb@pobox.com> wrote: >In article <6girj3$6j$2@anvil.BLaCKSMITH.com>, cswiger@BLaCKSMITH.com wrote: > >> Let me put it to you this way: the increase in costs represents roughly a >> day's pay (before taxes, anyway :-) for the average programmer. BFD. > >The new cost is two weeks' pay (after withholding :-) out of my stipend. :-( Consider it an investment. -- -Brian Slick slick@ecn.purdue.edu Winner: 1998 SDRC Calendar Contest http://www.sdrc.com/partners/university/contest/1998/
From: xray@cs.brandeis.edu (Nathan G. Raymond) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 21:11:12 GMT Organization: Brandeis University - Computer Science Dept. Message-ID: <6gjdhg$idr$1@new-news.cc.brandeis.edu> References: <MPG.f960500e190219d989895@news.supernews.com> <joe.ragosta-0804982250580001@wil62.dol.net> <MPG.f96ab2832ee6e1a9896d8@news.itg.ti.com> In article <MPG.f96ab2832ee6e1a9896d8@news.itg.ti.com> jmcn@msg.ti.com (Jason McNorton) writes: >In article Joe Ragosta, joe.ragosta@dol.net says... >*snip* >> I apologized to Donald in e-mail for the strength of my comments. >> >> But I don't see anything here that changes the essence. > >Face it Joe, Apple has burned lots of people and they resent Apple for >it. And they are right. Not everyone can be as blindly fanatical as you >are. I'm starting to see that maybe you don't really have much reliance >or much invested in Apple. If you did, you wouldn't support them so >freely. People who do rely on Apple get the shaft, a lot. > >-- >A world without the Mac is a step closer to utopia. Apple's biggest mistakes have been to imply or say they are going to deliver advances that would equal or exceed the relative magnitude of past advances, and then not followed through. Apple's greatness has also been its greatest downfall, in that sense. Can't say I blame people for trusting, can't say I blame Apple, but what I can offer is this: take everything with a very healthy dose of skepticism and never put all your eggs in one basket. Diversification is key, recognition that there are no absolutes is key. And I suggest you keep your .sig, as it will alert people immediately to your shortcomings, notable your lack of foresight (ideal perfection of utopia, i.e. the lack of any yin to yang, would mean that goodness would exist in essentially a vacuum, and with no counter-experience existance itself would be reduced to a very bland ennui) and your inability to recognize that the diversification of the application of technology in the masses has seen the convergence of dispirate implimentations (which could be symbolized as seperate overlapping roads or intersecting but seperate vectors to largely aligned vectors, or a single multi-lane road, where the velocity component of the vector or the speed and thus load carrying capacity of a lane would correspond to that technology path's magnitude penetration or installed base), and thus an overall decrease in the exploration of the state space of technology implimentations has occured within the industry, decreasing the possibility of any real substantial increase in productivity gains for end-users due to convergence, and any loss of diversification at this point will further ensure that significant real-world advances, not just incremental reiterations, occurs. Thus to decry the existance of any facet of the tech sector, rather than encourage interoperable yet diverse directions, is heretical to the very system I suppose you subscribe to. -- Nathan Raymond -- Nathan Raymond xray@cs.brandeis.edu raymond@binah.cc.brandeis.edu http://www.cs.brandeis.edu/~xray
From: ericb@pobox.com (Eric Bennett) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 17:10:01 -0400 Organization: Cornell University Sender: emb22@cornell.edu (Verified) Message-ID: <ericb-0904981710010001@132.236.171.104> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> In article <6ggmk2$jdl$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: > The Online program is free. It gets you the samples, and the > manuals. I love the spin on calling this a new program, since all this stuff was available before. > Now, as far as Rhapsody is concerned it is leaving a gap, but > only because of the time-frame.... However if you join Select you're > still covered. This is a gap Apple should not be leaving. Presumably they are interested in getting more developers for their new OS, but it doesn't show... -- Eric Bennett (www.pobox.com/~ericb) Cornell University Biochemistry Department
From: ericb@pobox.com (Eric Bennett) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 17:12:50 -0400 Organization: Cornell University Sender: emb22@cornell.edu (Verified) Message-ID: <ericb-0904981712500001@132.236.171.104> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> <6ggqha$9fk$1@news01.deltanet.com> <6ghg4o$np$1@news.digifix.com> > >There will always be a gap. Apple is incapable of shipping something > >with decent prerelease documentation in the New Regime, I suspect. > >HFS+ might be a counterargument, but if there is no way to get a > >Rhapsody seed without a substantial cash outlay, then you have > >delayed the release of software using Rhapsody features until your > >public ship date. I can make another argument around HFS+, though: AFAIK the docs for HFS+ are still not available online, even though MacOS 8.1 has been out for awhile. The only people who have received the docs are developers who are seeded. Thankfully somebody from Apple sent me the information I needed for the program I was writing. -- Eric Bennett (www.pobox.com/~ericb) Cornell University Biochemistry Department
From: ericb@pobox.com (Eric Bennett) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 17:17:41 -0400 Organization: Cornell University Sender: emb22@cornell.edu (Verified) Message-ID: <ericb-0904981717420001@132.236.171.104> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gecns$f9p$1@news01.deltanet.com> <1d767ho.1tixfqy1dqgm03N@hobbit1.injep.fr> <6gg6af$fit$1@anvil.BLaCKSMITH.com> <6ggabn$rts$2@news01.deltanet.com> <6ggj21$fit$6@anvil.BLaCKSMITH.com> In article <6ggj21$fit$6@anvil.BLaCKSMITH.com>, cswiger@BLaCKSMITH.com wrote: > However, Apple does not want to spend the time (== money) and resources to provide > extensive support for DR1 to individual people who just want to play around. They don't have to provide support. I get all the support I need from Usenet, mailing lists, and the documentation. I've had plenty of Macintosh programming questions, and have never directed any of them specifically at Apple (although I have had Apple employees respond to help requests in Internet forums). -- Eric Bennett (www.pobox.com/~ericb) Cornell University Biochemistry Department
Date: Thu, 09 Apr 1998 15:31:59 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-0904981531590001@wil38.dol.net> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> <6gj5ps$3qf$1@news01.deltanet.com> In article <6gj5ps$3qf$1@news01.deltanet.com>, scott@eviews.com (Scott Ellsworth) wrote: > In article <B1511019-15903@206.165.43.154>, "Lawson English" > <english@primenet.com> wrote: > > >Trying to make developer *support* a profit center is an INSANE idea. > >Hopefully, either Jobs will change this, or the more rational stockholders > >will hear about it in time to vote against him because this one action, *BY > >ITSELF*, will destroy Apple. Just like the last 20 rants that you went on were going to destroy Apple. Wasn't abandoning QD GX supposed to destroy Apple? > > For what it is worth, my proxy is sitting on my desk now, and I have > a very brief time to get it in the mail in time for the stockholder > meeting. It is a pity that there is no easy way to say, even en > masse "I will vote against your appiontment to the BoD because > of the following move" before the day itself. Even after voting, > there is little way for them to know why a group of shareholders > voted the way they did. The stockholders thus have little way > of communicating why they are acting the way they are. > > NB, I do not think stockholders should be running a company > particularly, but it is good for directors to know the reasons > why the stockholders react the way they do. Agreed. This is one of the things that bothers me about proxy statements. You can vote "no" on a given item and no one knows why you voted against it. -- Regards, Joe Ragosta http://www.dol.net/~Ragosta/complmac.htm
Date: Thu, 09 Apr 1998 16:24:09 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-0904981624100001@wil59.dol.net> References: <01bc455b$1be404c0$18f0bfa8@davidsul> <B1526CE7-65D6E@207.217.155.85> In article <B1526CE7-65D6E@207.217.155.85>, "Brad Hutchings" <brad@hutchings-software.com> wrote: > >I also hope several developers pass this on to the media, I am sure the > >media would love to flame apple for this and god knows apple deserves a > >good spanking for this > > A spanking over a fairly reasonable tactical move sends the wrong message. > This move would not have been met with intense cynicism if developers > thought Apple had an overall clue and direction and appreciation of how > developers support Apple. It starts from the top, and unfortunately, is > beyond the control of ADR. So, in other words, you're just assuming that, because it's Apple, they did the wrong thing. -- Regards, Joe Ragosta http://www.dol.net/~Ragosta/complmac.htm
From: "Brad Hutchings" <brad@hutchings-software.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 98 14:10:57 -0700 Organization: EarthLink Network, Inc. Message-ID: <B1528786-C9E9A@207.217.155.85> References: <ericb-0904981637520001@132.236.171.104> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit nntp://news.earthlink.net/comp.sys.mac.oop.powerplant, nntp://news.earthlink.net/comp.sys.mac.programmer.codewarrior, nntp://news.earthlink.net/comp.sys.next.advocacy, nntp://news.earthlink.net/comp.sys.next.programmer >That doesn't give Apple the right to alter contracts. If they want to >decrease the discount, or raise the fees for renewals or new accounts, >fine. But they're cutting off *existing* subscribers. > To be fair here... Apple's "contract" for the Associates Program gives them the right to change the terms and conditions at any time. Now, to be fair the other way, changing terms mid-stream may not always be positively received, and may affect resubscribe rates, participation in Apple endeavors, support for the platform, etc. Fred Giuffrida's analysis suggests that time will tell and that the best we can do in assessing whether or not it was a bad decsion is to later cite April 7, 1998 as one of those days: http://home.earthlink.net/~paladinsoft/AppleBits/NewDevProg.html Brad <mailto: "Brad Hutchings" brad@hutchings-software.com> <http://www.hutchings-software.com>
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 21:26:41 GMT Organization: Digital Fix Development Message-ID: <6gjeeh$qdm$1@news.digifix.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> <ericb-0904981710010001@132.236.171.104> In-Reply-To: <ericb-0904981710010001@132.236.171.104> On 04/09/98, Eric Bennett wrote: >In article <6ggmk2$jdl$1@news.digifix.com>, sanguish@digifix.com (Scott >Anguish) wrote: > >> The Online program is free. It gets you the samples, and the >> manuals. > >I love the spin on calling this a new program, since all this stuff was >available before. > Really? You could just download CodeWarrior Lite? >> Now, as far as Rhapsody is concerned it is leaving a gap, but >> only because of the time-frame.... However if you join Select you're >> still covered. > >This is a gap Apple should not be leaving. Presumably they are interested >in getting more developers for their new OS, but it doesn't show... > The timing is unfortunate. As a Rhapsody developer I know that as much as anyone. Unfortunately, WWDC is the date for the clearing of the fog. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: jstrout@ucsd.edu (Joseph J. Strout) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 15:03:51 -0700 Organization: UCSD Message-ID: <jstrout-ya02408000R0904981503510001@news.ucsd.edu> References: <6ggiu7$auu$1@vixen.cso.uiuc.edu> <B15138CC-19EEA@206.165.43.150> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <B15138CC-19EEA@206.165.43.150>, "Lawson English" <english@primenet.com> wrote: >This action by Apple doesn't increase revenues from developers >significantly and only serves to discourage use of the SDKs by the smaller, >more innovative houses, like shareware games writers. I can pretty much >guarantee you that any neophyte games writer will simply IGNORE the >Macinitosh once he/she learns that the Games Sprocket SDK is no longer >available online and that you need to pay Apple $500 to obtain it. Just my $0.02 worth: I didn't generally use the SDKs before, and especially won't use them now. Game Sprockets is a good example: I looked at it, and it did some nice things, but it would make my software not work on 68K Macs; if I code the same functionality (or at least the parts I would actually use) myself, I can make it run on all Macs. So many neophyte games writers will ignore the SDKs, but not necessarily ignore the Mac. Hmm, come to think of it though, I did actually use the Speech Recognition SDK... ,------------------------------------------------------------------. | Joseph J. Strout Department of Neuroscience, UCSD | | jstrout@ucsd.edu http://www.strout.net | `------------------------------------------------------------------'
From: jason@jhste1.dyn.ml.org (Jason S.) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 23:02:14 GMT Organization: I don't think so Message-ID: <slrn6iql0f.cvd.jason@jhste1.dyn.ml.org> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> <352D4957.84AE1890@alexa.com> Joshua Winsor wrote: >> Now, as far as Rhapsody is concerned it is leaving a gap, but >> only because of the time-frame.... However if you join Select you're >> still covered. > Download GnuStep and MKLinux, all free, MKLinux paid for by Apple, and >you are a Rhapsody developer. Is GNUStep that far along? BTW, if you want to run Linux on the Mac, consider Linux/PowerPC (if you have a PCI PowerMac): http://www.linuxppc.org/ -- If FreeBSD actually did that, I would concede that FreeBSD was any more "correct" than Linux is, but not even the FreeBSD people can justify that kind of performance loss. -- Linus Torvalds on comp.unix.advocacy
From: don.brown@cesoft.com (Donald Brown) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 9 Apr 1998 18:10:19 -0500 Organization: CE Software Message-ID: <MPG.f97115b6eb9d82698989a@news.supernews.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> <MPG.f960500e190219d989895@news.supernews.com> <joe.ragosta-0804982250580001@wil62.dol.net> <MPG.f96ab2832ee6e1a9896d8@news.itg.ti.com> In article <MPG.f96ab2832ee6e1a9896d8@news.itg.ti.com>, jmcn@msg.ti.com says... > In article Joe Ragosta, joe.ragosta@dol.net says... > *snip* > > I apologized to Donald in e-mail for the strength of my comments. > > > > But I don't see anything here that changes the essence. > > Face it Joe, Apple has burned lots of people and they resent Apple for > it. And they are right. Not everyone can be as blindly fanatical as you > are. I'm starting to see that maybe you don't really have much reliance > or much invested in Apple. If you did, you wouldn't support them so > freely. People who do rely on Apple get the shaft, a lot. > > -- > A world without the Mac is a step closer to utopia. > Before we all go jumping into bed together.... Apple can burn you, and yet Apple can bless you. Apple is neither perfect or is it the spawn of Satan. If I didn't care any more, if I didn't find this fascinating and love thoe machines, I wouldn't be here. My objections to some actions of Apple do not mean I hate Apple. I'm still hopining somehow things will be turned around. A world without the Mac would be a much sadder world indeed. Donald
From: Dave Helzer <rxgx30@email.sps.mot.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 16:41:48 -0700 Organization: Motorola Semiconductor Products Sector (SCG -WSBU) Message-ID: <352D5CB9.79EF61C9@email.sps.mot.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> <MPG.f960500e190219d989895@news.supernews.com> <joe.ragosta-0804982250580001@wil62.dol.net> <MPG.f96ab2832ee6e1a9896d8@news.itg.ti.com> <MPG.f97115b6eb9d82698989a@news.supernews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; x-mac-type="54455854"; x-mac-creator="4D4F5353" Content-Transfer-Encoding: 7bit Everyone following this thread might be interested in the following link: http://www.maccentral.com/news/9804/09.devrespond.shtml -Dave
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 23:06:51 GMT Organization: P & L Systems Message-ID: <6ggvub$oop$5@ironhorse.plsys.co.uk> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: english@primenet.com Xcanpos: shelf.1/199804230601!0011414257 In <B1511019-15903@206.165.43.154> "Chicken Little" wrote: > This harkens back to the worst of the bad ole days of Spindler, where you > had to pay lots of money for every aspect of MacOS programming and > developers started abandoning the MacOS platform. > Tummyrot. > Trying to make developer *support* a profit center is an INSANE idea. > Apple acting as a charity subsidising people's hobbies is what's insane. Nor should they be offering a program wide open to abuse by people signing up as developers just to get a discount on hardware. If you're a professional developer, or intent on producing shareware that's of genuine utility (i,e, good enough that people will pay for it), $500 should not be a serious impediment, else your business plans are out of kilter. And if you're a hobbyist, the cost of the new program, which is hardly *essential* to any *hobbyist*, is less than I gather most people spend on beer in a year. > Hopefully, either Jobs will change this, or the more rational stockholders > will hear about it in time to vote against him because this one action, *BY > ITSELF*, will destroy Apple. > The sky is falling indeed. mmalc.
From: j-jahnke@uchicago.edu (Jerome Jahnke) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 20:29:43 -0500 Organization: University of Chicago -- BSDIS/CRT Message-ID: <j-jahnke-0904982029430001@192.168.1.4> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> <j-jahnke-0804982141280001@192.168.1.3> <6ghg9f$o3$1@news.digifix.com> <j-jahnke-0904980214320001@192.168.1.3> <352d0347.0@206.25.228.5> In article <352d0347.0@206.25.228.5>, jkheit@xtdl.com wrote: > j-jahnke@uchicago.edu (Jerome Jahnke) wrote: > > In article <6ghg9f$o3$1@news.digifix.com>, sanguish@digifix.com > > (Scott Anguish) wrote: > > > On 04/08/98, Jerome Jahnke wrote: <snip> > > > >Now this is a bunch of bull. Many of us were Associates for > > > >one reason and one reason only, access to cheaper equipment. > > > >Apple does not lose money by selling me my one machine a year > > > >at a low price. > > > > > > Perhaps. But they do still provide you with all the other > > > stuff, the CD's, the mailings etc.. that does cost Apple money. > > > It don't cost 500 bucks... > > I'd like to get this straight. Do the mailings for $200 include > betas of Rhapsody or not? Nope, you have to be a Apple Developer to get the early release seeds (Rhapsody is included in this.) If you only get the tech mailings you get at least one CD a month and some fliers which always seem to end in up in my recycle basket, given they are without a doubt all Ra Ras Cis Boom Ba types of stuff. Jer,
From: FRIDBERG@PSFC.MIT.EDU Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 APR 1998 15:47:08 GMT Organization: MIT Plasma Science and Fusion Center Sender: fridberg@fredjr.pfc.mit.edu Message-ID: <9APR98.15470897@fredjr.pfc.mit.edu> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <6gh8v1$oop$6@ironhorse.plsys.co.uk> In a previous article, mmalcolm crawford <malcolm@plsys.co.uk> wrote: ->In <01bc4540$d4a15e60$3af0bfa8@davidsul> "macghod" wrote: ->> ->> > The lowest paid developer status is $500 instead of $250. If you really ->> > need to get the beta OS versions, your cost went up by $250. But everyone ->> > else seems to be better off. ->> ->> More Ragosta FUD. "Everyone else seems to be better off". Who is better ->> off? ->> ->Apple's customers who no longer have to subsidise hobbyist developers. -> ->Best wishes, -> ->mmalc. Do you mean to tell us that you have no freeware or shareware applications on your hard drive that you got for next to nothing? The ones that you'd probbly had to pay a lots of money for (and they could be lot worse as well) without subsidizing them? I think that hobbyist development on Mac plays very big part of Mac. We couldn't have survived without them. Mike.
From: russotto@wanda.pond.com (Matthew T. Russotto) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 10 Apr 1998 01:53:20 GMT Organization: Ghotinet Message-ID: <6gju2g$82k@netaxs.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <joe.ragosta-0804982250580001@wil62.dol.net> <ericb-0904981631110001@132.236.171.104> <6gje5e$q8g$1@news.digifix.com> In article <6gje5e$q8g$1@news.digifix.com>, Scott Anguish <sanguish@digifix.com> wrote: } There has been a number of posts to Rhapsody-talk, and the }Lyris rhapsody list by developers who don't understand what all the }whining is about. That would be the "We got ours, to hell with the rest of you" syndrome. -- Matthew T. Russotto russotto@pond.com "Extremism in defense of liberty is no vice, and moderation in pursuit of justice is no virtue."
From: FRIDBERG@PSFC.MIT.EDU Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 APR 1998 16:10:51 GMT Organization: MIT Plasma Science and Fusion Center Sender: fridberg@fredjr.pfc.mit.edu Message-ID: <9APR98.16105101@fredjr.pfc.mit.edu> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <6ggvub$oop$5@ironhorse.plsys.co.uk> <352CA831.D4878417@ArtQuest.fr> In a previous article, Hubert HOLIN <hh@ArtQuest.fr> wrote: -> -> ->mmalcolm crawford wrote: -> ->[SNIP] -> ->> If you're a professional developer, or intent on producing shareware that's ->> of genuine utility (i,e, good enough that people will pay for it), $500 ->> should not be a serious impediment, else your business plans are out of ->> kilter. -> ->[SNIP] -> -> By that measure, freeware is junk, right? I guess TeX, GNU and their ilk are ->unknown to you... -> -> Hubert Holin -> holin@mathp7.jussieu.fr -> -> And let's not forget such things as NewsWatcher (which probably half of people who read this NG are using, or NSCA Telnet (which I am using) or bunch of other internet software which made Mac so attractive to use for Internet access and allowed us to have pretty high market share on internet. Mister MMalcolm Crawford prbably never heard of such programs as Disinfectant of Internet Config either. Or maybe he just afraid that all those freeware applications going to drive him out of business? Mike.
From: andreja.vidmar@select-tech.si (Andreja Vidmar) Newsgroups: comp.sys.next.programmer Subject: Re: TIFF format used in OpenStep / Rhapsody Date: 7 Apr 1998 08:51:05 GMT Organization: SELECT Technology Message-ID: <6gcpdp$nr$1@lazar.select-tech.si> References: <Pine.LNX.3.95.980403114356.19940B-100000@fabre.act.qc.ca> In article <Pine.LNX.3.95.980403114356.19940B-100000@fabre.act.qc.ca> Andre-John Mas <ama@fabre.act.qc.ca> writes: > > Could someone tell me where I can find some specs of the > TIFF file format used on OpenStep. I have was given some > TIFF files that originated from OpenStep and I am unable > to read them with any programs that supports TIFF on the > Mac or the PC. > > Thanks > > AJ Presumably JPEG compression is used in those TIFF files. TIFF files with JPEG compression are not supported on Mac (I assume the same holds for PC). If you have access to a machine running NEXTSTEP/OPENSTEP/Rhapsody, you can use tiffutil command line utility to find out if JPEG compression is used in the TIFF files and decompress them. Andreja
From: jayfar@netaxs.com (Jayfar) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple's Announcement Misleading. WAS :Apple developer program Date: Thu, 09 Apr 1998 23:41:18 -0400 Organization: Jayfar's Original Virtual Macintosh Message-ID: <jayfar-0904982341190001@downtown1-6.slip.netaxs.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> <MPG.f960500e190219d989895@news.supernews.com> <joe.ragosta-0804982250580001@wil62.dol.net> <gmgraves-0904980952020001@sf-pm5-27-91.dialup.slip.net> <joe.ragosta-0904981956410001@elk90.dol.net> Mail-Copies-To: jayfar@netaxs.com In article <joe.ragosta-0904981956410001@elk90.dol.net>, joe.ragosta@dol.net (Joe Ragosta) wrote: | I guess the question to answer to determine the accuracy of this is: "Was | it previously possible to download the SDK's for free?" If it was not, the | new program _would_ be an improvement on the low end since you could get | started without spending a penny. If it was always possible to download | the SDK's for free, it's hard to understand how the new program could be | an improvement. The SDKs have been downloadable for at least the last few years. Cheers, Jayfar -- Jay Farrell Jayfar's Original Virtual Macintosh jayfar@netaxs.com * Your Favorite Mac Site * Now Updated Daily * Philadelphia, USA <URL:http://www.netaxs.com/~jayfar/live.desk.html> Jayfar's APPLE DOOMSDAY CLOCK <URL:http://www.netaxs.com/~jayfar/>
From: "Jack G. Bader" <jbader@neteffects.com> Newsgroups: comp.sys.next.programmer,misc.jobs.contract,misc.jobs.offered Subject: **** Need Objective-C Programmers in Minneapolis *** Date: Mon, 06 Apr 1998 18:06:04 -0700 Organization: NetEffects Inc. Message-ID: <35297BFC.31D679E3@neteffects.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Xcanpos: shelf.1/199804210201!0032128096 NetEffects Inc. has several immediate openings in Minneapolis for experienced Objective-c programmers. You will join a team of developers supporting existing applications that also utilize EOF. Please submit your resume via email for immediate feedback. Thanks. Jack Bader NetEffects Inc. jbader@neteffects.com 314-727-1107
From: jayfar@netaxs.com (Jayfar) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 23:45:44 -0400 Organization: Jayfar's Original Virtual Macintosh Message-ID: <jayfar-0904982345460001@downtown1-6.slip.netaxs.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> <MPG.f960500e190219d989895@news.supernews.com> <joe.ragosta-0804982250580001@wil62.dol.net> <MPG.f96ab2832ee6e1a9896d8@news.itg.ti.com> <MPG.f97115b6eb9d82698989a@news.supernews.com> <352D5CB9.79EF61C9@email.sps.mot.com> Mail-Copies-To: jayfar@netaxs.com In article <352D5CB9.79EF61C9@email.sps.mot.com>, Dave Helzer <rxgx30@email.sps.mot.com> wrote: | Everyone following this thread might be interested in the following | link: | | http://www.maccentral.com/news/9804/09.devrespond.shtml | | -Dave HA!!! Their only defense was to cite the same 4 big league developers that they quoted in their original disingenous press release. Cheers, Jayfar -- Jay Farrell Jayfar's Original Virtual Macintosh jayfar@netaxs.com * Your Favorite Mac Site * Now Updated Daily * Philadelphia, USA <URL:http://www.netaxs.com/~jayfar/live.desk.html> Jayfar's APPLE DOOMSDAY CLOCK <URL:http://www.netaxs.com/~jayfar/>
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 10 Apr 1998 04:01:42 GMT Organization: Digital Fix Development Message-ID: <6gk5j6$6j5$1@news.digifix.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <joe.ragosta-0804982250580001@wil62.dol.net> <ericb-0904981631110001@132.236.171.104> <6gje5e$q8g$1@news.digifix.com> <6gju2g$82k@netaxs.com> In-Reply-To: <6gju2g$82k@netaxs.com> On 04/09/98, Matthew T. Russotto wrote: >In article <6gje5e$q8g$1@news.digifix.com>, >Scott Anguish <sanguish@digifix.com> wrote: > >} There has been a number of posts to Rhapsody-talk, and the >}Lyris rhapsody list by developers who don't understand what all the >}whining is about. > >That would be the "We got ours, to hell with the rest of you" syndrome. I think its more likely the 'we understand that it costs money to get stuff' syndrome. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: steve@discoverysoft.com (Steven Fisher) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 15:37:21 -0700 Organization: Discovery Software Ltd. Message-ID: <steve-0904981537210001@oranoco.discoverysoft.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> <ericb-0904981710010001@132.236.171.104> <6gjeeh$qdm$1@news.digifix.com> In article <6gjeeh$qdm$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: >On 04/09/98, Eric Bennett wrote: >>In article <6ggmk2$jdl$1@news.digifix.com>, sanguish@digifix.com >(Scott >>Anguish) wrote: >> >>> The Online program is free. It gets you the samples, and >the >>> manuals. >> >>I love the spin on calling this a new program, since all this stuff >was >>available before. >> > > Really? You could just download CodeWarrior Lite? Yes. http://www.metrowerks.com/lite/ -- Steven Fisher; Discovery Software Ltd.; steve@discoverysoft.com "Anyone who has never made a mistake has never tried anything new." -- Albert Einstein
From: "macghod" <macghod@concentric.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Apple developer program Date: 7 Apr 1998 20:58:06 GMT Message-ID: <01bc445f$744397e0$1bf0bfa8@davidsul> WTF is up with the apple developer program? I see a big announcement, and all it is is: 1st level: online stuff, just as it is now, ie NO CHANGE 2nd level: instead of costing $250 a year its now $500 a year??!?!?? As far as I can tell it adds nothing and is a 2x increase!! Why to support development apple :)
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 98 16:58:47 GMT Organization: QMS Message-ID: <6ggabn$rts$2@news01.deltanet.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gecns$f9p$1@news01.deltanet.com> <1d767ho.1tixfqy1dqgm03N@hobbit1.injep.fr> <6gg6af$fit$1@anvil.BLaCKSMITH.com> In article <6gg6af$fit$1@anvil.BLaCKSMITH.com>, cswiger@BLaCKSMITH.com wrote: >xhsoft@injep.fr (Xavier Humbert) wrote: >> Scott Ellsworth <scott@eviews.com> wrote: >>> The seeding program, imho, served a valuable purpose, but >>> Apple, I think, does not want any leaks of its seeds, and so >>> it is trying to reduce the number of developers who might >>> have access. >> >> What is obviously stupid : WarEz sites have System Seeds BEFORE the >> official FTP :-( >> >> Xav, looking for a warez URL > >I don't have much of an opinion about Apple's change to their Dev. program >pricing, one way or the other. I do know, however, that people who condone >software piracy have no moral grounds whatsoever to complain about what Apple >does. The fellow's point was, I hope, that limiting distribution in this way seems unlikely to prevent the seeds from getting to the warez sites, because the current arcana one has to go through to get seeds (and it is VERY arcane) has not succeeded. As a result, the only people who are incovenienced by the former situation were honest developers who wished to get new seeds directly from Apple. The new situation will limit the exposure of honest developers to upcoming MacOS releases even more. I did note his request for a Warez URL, but I hoped it was a pointed reminder that Xav wants to stay in the Apple camp. Let me say that more loudly - by limiting things in this way, they are only preventing honest developers willing to pay a substantial fee from testing products and giving useful feedback. $250 is substantial when it is out of your own pocket, and many Mac evangelists and partisans at otherwise wintel companies often have to do thier Mac work on thier own, in order to sell the company on the benefits. I know I do - all work on Mac EViews is done on my own personal time, because I dislike products that fail, and I find the Mac a more pleasant platfom. If Apple is determined for me to fail in this, by making private Mac development prohibitivly expensive, well, I have to act accordingly. Scott Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
From: sdsd <sdsd@skdjsk.no> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 1998 00:57:27 +0200 Organization: sda Message-ID: <352AAF57.25EB@skdjsk.no> References: <01bc445f$744397e0$1bf0bfa8@davidsul> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I want a student developers program... 150-200 $/year for cd's... Anonymous..
From: gcheng@uni.uiuc.edu (Guanyao Cheng) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 7 Apr 1998 23:04:21 GMT Organization: University of Illinois at Urbana-Champaign Message-ID: <6gebdl$5d0$1@vixen.cso.uiuc.edu> References: <01bc445f$744397e0$1bf0bfa8@davidsul> Can't you get the development CD's for $199?? Or do those not include beta OS releases?? Guanyao Cheng -- Guanyao Cheng "And I personally assure you, everybody here, that gcheng@uiuc.edu if Deep Blue will start playing competitive http://www.uiuc.edu/ph chess, I personally guarantee you I'll /www/gcheng tear it to pieces" -- Garry Kasparov
From: tom_e@usa.net (Thomas Engelmeier) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 8 Apr 1998 19:40:54 +0200 Organization: University of Rostock Message-ID: <1d7652m.2xxaxj9xa6yoN@desktop.tom-e.private> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6ge6hh$9mv@netaxs.com> <352abaf8.0@206.25.228.5> John Kheit <jkheit@xtdl.com> wrote: > > Yeah, sucks, doesn't it? All I really want is the CDs, and they > > keep upping the price and adding things I don't need. > > I think there is a free online subscription and one can subscribe > to get *just* the CDs in addition. I haven't bothered to check > out the details yet. CD's only is $199, but w/o Technology seeding :-(( Regards, TomE
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 11:25:02 -0700 Organization: Primenet Services for the Internet Message-ID: <B1511019-15903@206.165.43.154> References: <352b5da2.0@206.25.228.5> To: "John Kheit" <jkheit@xtdl.com> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit John Kheit <jkheit@xtdl.com> said: > j-jahnke@uchicago.edu (Jerome Jahnke) wrote: [...] > > Which is what we used to do, but then realized you could not get > > seeded unless you were an associate. > > What's seeded mean? Does that mean no beta's? No betas, and now, apparently no online versions of the SDKs. They were available last week online, but no longer. This harkens back to the worst of the bad ole days of Spindler, where you had to pay lots of money for every aspect of MacOS programming and developers started abandoning the MacOS platform. Trying to make developer *support* a profit center is an INSANE idea. Hopefully, either Jobs will change this, or the more rational stockholders will hear about it in time to vote against him because this one action, *BY ITSELF*, will destroy Apple. [mark my words, young man <quavering voice>] --------------------------------------------------------------------- "OpenDoc may be impressive technology, but if no-one else is using it, why should Apple?" --Steve Jobs "In short: if you want to blow people away with what can be done *only* on a Mac, show them Cyberdog." --Tom Keyes <news://cyberdog.apple.com/cyberdog.general/19411> ---------------------------------------------------------------------
From: schuerig@acm.org (Michael Schuerig) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 8 Apr 1998 20:35:08 +0200 Organization: Completely Disorganized Message-ID: <1d757w4.17mii1i1hjria0N@rhrz-isdn3-p47.rhrz.uni-bonn.de> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <01bd6267$fa532b30$9b2168cf@test1> Todd Heberlein <todd@NetSQ.com> wrote: > As a student, you will probably be able to get the full development > environment quite inexpensively from Apple, but we will have to wait a > while for Apple to make that move. The last time I received word from > Apple (less than a month ago), they had not hammered out all the license > deals with third parties. > > Also, $500 is still quite cheap. As I mentioned in another thread, > yesterday I paid $1300+ for a license to use Digital's command-line > debugger (which is pretty lame). Okay, I am still a student. But even after that I'd like to be able to afford Apple/Mac/Rhapsody development tools. If my employer pays for them that fine. It's not likely, though. And it's small shareware developers who are driven out. If Apple thinks it doesn't need those people and that the real pros that they want can easily pay the price -- well, then go ahead. Then I'll leave sooner or later. It's rather simple: I want to be able to develop software on my system, if I can't afford to do that on an Apple system then I won't stay there. Michael -- Michael Schuerig mailto:schuerig@acm.org http://www.uni-bonn.de/~uzs90z/
From: "Paul Rekieta" <PRekieta@ix.netcom.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Tue, 7 Apr 1998 16:23:27 -0700 Organization: Apple Computer, Inc. Message-ID: <6geche$8lu$1@news2.apple.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <352AAF57.25EB@skdjsk.no> sdsd wrote in message <352AAF57.25EB@skdjsk.no>... >I want a student developers program... > >150-200 $/year for cd's... > You got it -- The Apple Developer Connection Mailing. The ADC Mailing is a 12-month subscription that includes the Developer CD Series, a wealth of technical resources, system software, development tools, technical documentation, Interactive Media Resources, SDKs, and more. More information can be found at http://developer.apple.com/programs/mailing.html.
Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6ge6hh$9mv@netaxs.com> From: John Kheit <jkheit@xtdl.com> Organization: monoChrome, Inc., NJ, USA Message-ID: <352abaf8.0@206.25.228.5> Date: 7 Apr 98 23:47:04 GMT russotto@wanda.pond.com (Matthew T. Russotto) wrote: > In article <01bc445f$744397e0$1bf0bfa8@davidsul>, macghod > <macghod@concentric.net> wrote: }WTF is up with the apple > developer program? I see a big announcement, and }all it is is: > }1st level: online stuff, just as it is now, ie NO CHANGE }2nd > level: instead of costing $250 a year its now $500 a year??!?!?? > } }As far as I can tell it adds nothing and is a 2x increase!! > Why to support }development apple :) > Yeah, sucks, doesn't it? All I really want is the CDs, and they > keep upping the price and adding things I don't need. I think there is a free online subscription and one can subscribe to get *just* the CDs in addition. I haven't bothered to check out the details yet. -- Thanks, be well, take care, later, John Kheit; self expressed... __________________________________________________________________ monoChrome, Inc. ASCII, MIME, PGP, SUN, & NeXTmail OK NeXT/OPENSTEP Developer mailto:jkheit@xtdl.com Telepathy, It's coming... http://www.xtdl.com/~jkheit Franklin Pierce Law Center You're dangerous because you're honest
Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer From: erkyrath@netcom.com (Andrew Plotkin) Subject: Re: Apple developer program Message-ID: <erkyrathEr4984.8MF@netcom.com> Followup-To: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Organization: Netcom On-Line Services References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> Date: Wed, 8 Apr 1998 22:31:16 GMT Sender: erkyrath@netcom11.netcom.com Xcanpos: shelf.1/199804230201!0032511460 Lawson English (english@primenet.com) wrote: > Rex Riley <rriley@yahoo.com> said: > > Apple's actions signal that it can't afford to get into the "handout" > > business nor support legions of "individual" developers (read welfare). > By > > "raising-the-bar" Apple introduces an entry fee and seeds the concept of > > "profit" in its Developer base. Developers who approach Rhapsody without > > resources and a profit motive will be annoyed, discouraged and drop off. > But shareware programmers often fill in the gaps that commercial software > leaves. Also shareware programmers learn the ropes via feedback from their > customers and are often high school/college students who later enter the > Mac programming field as professionals. No kidding. I'm an individual, free-time, freeware/shareware Mac developer. I've been writing freeware games (text adventures) and doing a *lot* of work in porting free games and game systems to the Mac. (Yeah, text adventures are a tiny little niche of the game world. Everybody's got their niche.) I may get back into Mac shareware someday. If Apple really wants me to "be annoyed, discouraged, and drop off"... well, then the Mac game world will get that much smaller. I don't think Apple really wants that, as you (Rex Riley) claim. I *do* think they're willing to *accept* it in exchange for higher short-term profits from their developer program. No surprise there. Is there any money in the BeOS shareware game market? --Z -- "And Aholibamah bare Jeush, and Jaalam, and Korah: these were the borogoves..."
From: G.C.Th.Wierda@AWT.nl (Gerben Wierda) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Fri, 10 Apr 1998 07:51:54 GMT Organization: Adviesraad voor het Wetenschaps- en Technologiebeleid Sender: news@AWT.NL Message-ID: <Er6tuI.Mu8@AWT.NL> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <joe.ragosta-0904980642200001@elk33.dol.net> <6gitpj$csn$1@interport.net> <joe.ragosta-0904980134360001@wil68.dol.net> joe.ragosta@dol.net (Joe Ragosta) wrote: >> Apple should be ACTIVELY ENCOURAGING these people. Some of them might >> grow up to be software developers . . . or management . . . or sysadmins . >> . . or end users . . . or, or, or . . . > >Sure. They should encourage everyone. > >But how do you allocate limited resources? Apple has decided that spending >the money on TV ads or magazine ads is a better place to put scarce >resources than subsidizing hobbyists. Which is a good point, but do we know the numbers? Is the cost of providing hardware discounts and support comparable to the cost of running TV ads? This comparison only makes sense if you know the relative cost. Price/performance is the issue, also in marketing. >And in case you've missed it, it is now possible for hobbyists to get >SDK's free. How much cheaper do they need to be? SDK's and Mac API development don't interest me. I want to know what will happen to the cost of developing for Rhapsody. --Gerben
From: G.C.Th.Wierda@AWT.nl (Gerben Wierda) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Fri, 10 Apr 1998 07:57:51 GMT Organization: Adviesraad voor het Wetenschaps- en Technologiebeleid Sender: news@AWT.NL Message-ID: <Er6u4F.Mv5@AWT.NL> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> <1d7821x.1rdi5101enh5fhN@hobbit1.injep.fr> xhsoft@injep.fr (Xavier Humbert) wrote: ><rmcassid@uci.edu> wrote: > >> and can quite likely afford a modest >> increase in costs. > >100% increase is modest for you ? Come on. If they had to stick to figures like 10% they could have increased a whopping $25. Any increase on a small amount is big in percentages. --Gerben
From: "M. Kilgore" <mkilgore@nospam.prysm.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program -shareware Date: Wed, 8 Apr 1998 14:52:21 -0500 Message-ID: <6ggkg1$90k$1@news0-alterdial.uu.net> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> There is a very good reason for Apple to discourage the small or shareware developer - the sad fact is that if it's too easy to get into the shareware market then less than knowlegeable programmers can play hell with your system. All you need do is look at the Win shareware to discover this. Pehaps Rhap isn't as bullet proof as the pr would have us believe and Apple is just trying to protect the system from becoming just another Windows. There is, however, a far better reason to encourage shareware/freeware - many of the guys that get started in that do it because they're enthusiastic about their machines. That enthusiasm is is contageous and gets even more people interested in a platform. So how bad could it really be if CNET was able to cover a new & exciting piece of MacShareWare on each episode. I'd be tempted to say that a healthy shareware market is directly translatable into new hardware sales. And no, I don't have any studies to back that up. Then again, it just may be that it's not just MS that makes PCs sell. I'm still thinking, though, that Apple's trying to get itself sold. They keep saying they're on the offensive and yet they just seem to be hunkering down to a nice marketable package for someone else to buy and rejuvenate. mark
From: support@fluxsoft.com (Maurice Volaski) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Tue, 07 Apr 1998 20:11:26 -0500 Organization: Flux Software Message-ID: <support-0704982011260001@207-172-164-184.s57.as4.loc.erols.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <j-jahnke-0704981616320001@192.168.1.3> In article <j-jahnke-0704981616320001@192.168.1.3>, j-jahnke@uchicago.edu (Jerome Jahnke) wrote: >In article <01bc445f$744397e0$1bf0bfa8@davidsul>, "macghod" ><macghod@concentric.net> wrote: > >> WTF is up with the apple developer program? I see a big announcement, and >> all it is is: >> 1st level: online stuff, just as it is now, ie NO CHANGE >> 2nd level: instead of costing $250 a year its now $500 a year??!?!?? >> >> As far as I can tell it adds nothing and is a 2x increase!! Why to support >> development apple :) > >Actually it adds nothing, removes other stuff and increases by a factor of >two... Apple did this a few years ago, gouge developers. It didn't work >then. I don't see why they think it will work now. It won't if we complain (won't it?). -- Maurice Volaski, Flux Software support@fluxsoft.com http://www.fluxsoft.com/ ftp://ftp.fluxsoft.com
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 20:27:46 GMT Organization: Digital Fix Development Message-ID: <6ggmk2$jdl$1@news.digifix.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> In-Reply-To: <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> On 04/08/98, George Graves wrote: >In article <6gg8il$6v9$1@ha2.rdc1.sdca.home.com>, >rexr@cx54440-a.dt1.sdca.home.com wrote: > >> In <352b5da2.0@206.25.228.5> John Kheit wrote: >> > j-jahnke@uchicago.edu (Jerome Jahnke) wrote: >> > > In article <trumbull-0704981900290001@net44-223.student.yale.edu>, >> > > trumbull@cs.yale.edu (Ben Trumbull) wrote: > >> What Apple Developer Program seeds is neither leadership nor a grand vision >> around which anyone can "rally" . So Apple's unilateral actions lack >> context and seed resentment and hostility... > >What this really means is less shareware (on top of the fact that >its already shrinking because the platform is shrinking), and less >software from small developers in general, you know, the people >from whom all the innovation comes. Thats a crock. The Online program is free. It gets you the samples, and the manuals. It also gets you Code Warrior Lite. From that you can learn Macintosh programming and put out freeware apps. Few, if any, of the people pissing and moaning about loosing access to the seeding program actually NEED access to it in many cases. Now, as far as Rhapsody is concerned it is leaving a gap, but only because of the time-frame.... However if you join Select you're still covered. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: kychenABC@hpl.hp.com (Kay-Yut Chen - remove ABC in email to reply) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 1998 20:40:36 GMT Organization: All USENET -- http://www.Supernews.com Message-ID: <3532e039.603512314@news.supernews.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gecns$f9p$1@news01.deltanet.com> <1d767ho.1tixfqy1dqgm03N@hobbit1.injep.fr> <6gg6af$fit$1@anvil.BLaCKSMITH.com> <6ggabn$rts$2@news01.deltanet.com> <joe.ragosta-0804981312160001@wil32.dol.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit >Maybe that was the point, but it's not how I read it, either. > >One would presume that "serious" developers (i.e. those who rely on Apple >for their livelihood) are less likely to post a build to the Warez site. >To the extent that raising the price will not affect a large company as >much as someone doing Mac OS development as a hobby, raising the price >_may_ reduce the postings to Warez sites. > >Whether that justifies the action is, of course, open for debate. > I do not think that argument holds water. It is true the "serious" developers is less likely to post stuff to a warez site. However, once something is out to the warez sites, it gets duplicates very very fast since you just need ONE pirate to get their hands on something before it gets all over the place. Thus, unless you can stop EVERY developers to leak something out (which I think is impossible by raising the price, you will only stop some), the end result will be the same. It will be distributed everywhere. Kay-Yut
From: bbennett@unixg.ubc.ca (Bruce Bennett) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 8 Apr 1998 14:09:33 -0700 Organization: University of British Columbia Message-ID: <1d765na.sy4iporb5zhoN@p060.intchg1.net.ubc.ca> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> Rex Riley <rriley@yahoo.com> wrote, among other things: > Apple's actions signal that it can't afford to get into the "handout" > business nor support legions of "individual" developers (read welfare). By > "raising-the-bar" Apple introduces an entry fee and seeds the concept of > "profit" in its Developer base. Developers who approach Rhapsody without > resources and a profit motive will be annoyed, discouraged and drop off. "Post no trifling freeware or shareware. Give us shrink-wrapped consumer products or give us death." Anyone alarmed by the "think different" ideology should be pleased by this new turn. -- Bruce Bennett
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 14:18:01 -0700 Organization: Primenet Services for the Internet Message-ID: <B15138CC-19EEA@206.165.43.150> References: <6ggiu7$auu$1@vixen.cso.uiuc.edu> To: "Guanyao Cheng" <gcheng@uni.uiuc.edu> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Guanyao Cheng <gcheng@uni.uiuc.edu> said: > Although I agree it might drive away small, > commercial developers, for the student developer, even the previous $250 > was too much. But all the SDK's were online last week for MacOS, but are no longer.. "SDK" stands for "software development kit." These give you sample code, libraries, documentation, etc. Without those, Mr. Lau might have had a hard time developing his product... [Actually, the SDK's didn't exist back then, but making use of the latest libraries and API managers from Apple absolutely REQUIRES one to have access to these kits] Thing is, the bigger houses don't download all the SDKs, but only the updates. They use the CD's which are a commercial item. However, a shareware/freeware author that only wants access to a single SDK, could download that single SDK as needed -up until this week. This action by Apple doesn't increase revenues from developers significantly and only serves to discourage use of the SDKs by the smaller, more innovative houses, like shareware games writers. I can pretty much guarantee you that any neophyte games writer will simply IGNORE the Macinitosh once he/she learns that the Games Sprocket SDK is no longer available online and that you need to pay Apple $500 to obtain it. Ask Ambrosia Software if they would be willing to pay $500 upfront if they were now starting out in the MacOS shareware business. I don't think so. --------------------------------------------------------------------- "OpenDoc may be impressive technology, but if no-one else is using it, why should Apple?" --Steve Jobs "In short: if you want to blow people away with what can be done *only* on a Mac, show them Cyberdog." --Tom Keyes <news://cyberdog.apple.com/cyberdog.general/19411> ---------------------------------------------------------------------
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 14:20:01 -0700 Organization: Primenet Services for the Internet Message-ID: <B1513924-1B3A6@206.165.43.150> References: <6ggj21$fit$6@anvil.BLaCKSMITH.com> To: "Charles Swiger" <chuck-nospam@blacksmith.com> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Charles Swiger <chuck-nospam@blacksmith.com> said: > The people for whom this price change matters are individuals and > students. And, to > be blunt, these people are the ones who are unlikely to have the resources > (unused > machine(s) to dedicate to Rhapsody, a functioning LAN with a firewall [to > meet the > RDR1 licensing terms], and a working fileserver and preferably NetInfo > server) to > provide Apple with the kinds of feedback that they could use. Such > individuals are > far more likely to require excessive support resources for little benefit. > Games Sprockets for MacOS now costs $500. It isn't just the Rhapsody beta seeding. That I don't mind paying for. --------------------------------------------------------------------- "OpenDoc may be impressive technology, but if no-one else is using it, why should Apple?" --Steve Jobs "In short: if you want to blow people away with what can be done *only* on a Mac, show them Cyberdog." --Tom Keyes <news://cyberdog.apple.com/cyberdog.general/19411> ---------------------------------------------------------------------
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: OOPS -SDKs still available (Was Re: Apple developer program Date: 8 Apr 1998 14:26:00 -0700 Organization: Primenet Services for the Internet Message-ID: <B1513A75-202AB@206.165.43.150> References: <6ggmk2$jdl$1@news.digifix.com> To: "Guanyao Cheng" <gcheng@uni.uiuc.edu>, "Charles Swiger" <chuck@blacksmith.com>, "Scott Ellsworth" <scott@eviews.com>, "Joe Ragosta" <joe.ragosta@dol.net> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit > > The Online program is free. It gets you the samples, and the > manuals. It also gets you Code Warrior Lite. > > From that you can learn Macintosh programming and put out > freeware apps. > I just checked. Apparently I had misread what the new developer's program provides. MacOS SDKs are still available via ftp. That means that everything that I've been saying is non-relevant (so what else is new?). mea culpa. --------------------------------------------------------------------- "OpenDoc may be impressive technology, but if no-one else is using it, why should Apple?" --Steve Jobs "In short: if you want to blow people away with what can be done *only* on a Mac, show them Cyberdog." --Tom Keyes <news://cyberdog.apple.com/cyberdog.general/19411> ---------------------------------------------------------------------
From: bl003@dial.oleane.com (Benoit Leraillez) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 8 Apr 1998 23:36:35 +0200 Organization: Guest of OLEANE Message-ID: <1d76xac.1dq1swxbccaqoN@dialup-15.def.oleane.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gecns$f9p$1@news01.deltanet.com> <1d767ho.1tixfqy1dqgm03N@hobbit1.injep.fr> <6gg6af$fit$1@anvil.BLaCKSMITH.com> <6ggabn$rts$2@news01.deltanet.com> <6ggj21$fit$6@anvil.BLaCKSMITH.com> Charles Swiger <chuck-nospam@blacksmith.com> wrote: > However, Apple does not want to spend the time (== money) and resources to > provide extensive support for DR1 to individual people who just want to > play around. I dont need support from Apple and I understand that asking money in exchange of info (read debugging help) is normal. I need the seeds to make sure my product will run when my clients receive the new OS version. And I dont have an extra $250 to throw away. Benoit Leraillez
Date: Wed, 08 Apr 1998 18:55:21 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-0804981855220001@elk68.dol.net> References: <6ggiu7$auu$1@vixen.cso.uiuc.edu> <B15138CC-19EEA@206.165.43.150> Xcanpos: shelf.1/199804230201!0035331913 In article <B15138CC-19EEA@206.165.43.150>, "Lawson English" <english@primenet.com> wrote: > Guanyao Cheng <gcheng@uni.uiuc.edu> said: > > > Although I agree it might drive away small, > > commercial developers, for the student developer, even the previous $250 > > was too much. > > > > But all the SDK's were online last week for MacOS, but are no longer.. > "SDK" stands for "software development kit." These give you sample code, > libraries, documentation, etc. > > Without those, Mr. Lau might have had a hard time developing his product... > Of course, the SDKs are still online at no charge. This makes your entire point moot (as usual). -- Regards, Joe Ragosta See the Complete Macintosh Advocacy Page http://www.dol.net/~Ragosta/complmac.htm
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 22:01:44 GMT Organization: P & L Systems Message-ID: <6ggs48$oop$2@ironhorse.plsys.co.uk> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <xhsof-0804981448100001@hobbit1.injep.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: xhsof@injep.fr In <xhsof-0804981448100001@hobbit1.injep.fr> Xavier Humbert wrote: > If developers have to pay 500 bucks to have Rhapsody, do you think the > will develop for it ? > If anyone is wanting to develop for Rhapsody professionally and doesn't have $500, they don't have a business plan. If you're intending to write shareware, then $500 is a reasonable investment if you believe you have something useful to contribute. If you're writing Rhapsody apps for a hobby, why should I subsidise your playtime? In the bad old days when NeXT's developement tools alone cost seven times that amount there was no shortage of freeware or shareware. Best wishes, mmalc.
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 21:56:37 GMT Organization: P & L Systems Message-ID: <6ggrql$oop$1@ironhorse.plsys.co.uk> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: english@primenet.com In <B1511189-1AF6A@206.165.43.154> "Lawson English" wrote: > But shareware programmers often fill in the gaps that commercial software > leaves. Also shareware programmers learn the ropes via feedback from their > customers and are often high school/college students who later enter the > Mac programming field as professionals. > If shareware authors can't make back $500 on their app, then they're not writing a "utility" in the sense I understand it. Apple has said that the acedemic program is to be announced later -- I'd expect it to be cheaper. Why does anyone *need* to be part of the program if they're not developing professional or semi-professional apps. Sure, access to the latest and greatest betas of the s/w is fun, but not essential if you're hacking at home for the joy of it. Apple is not a charity -- and I don't want to have to pay more for my Mac to subsidise someone else's hobby. mmalc.
From: Andre-John Mas <ama@fabre.act.qc.ca> Newsgroups: comp.sys.next.programmer Subject: Re: TIFF format used in OpenStep / Rhapsody Date: Wed, 8 Apr 1998 17:06:54 -0500 Organization: Communications Accessibles Montreal, Quebec Canada Message-ID: <Pine.LNX.3.95.980408170629.8499A-100000@fabre.act.qc.ca> References: <Pine.LNX.3.95.980403114356.19940B-100000@fabre.act.qc.ca> <6gcpdp$nr$1@lazar.select-tech.si> <marke-ya02408000R0704981357380001@17.128.100.122> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII To: Mark Eaton <marke@apple.com> In-Reply-To: <marke-ya02408000R0704981357380001@17.128.100.122> On Tue, 7 Apr 1998, Mark Eaton wrote: > In article <6gcpdp$nr$1@lazar.select-tech.si>, > andreja.vidmar@select-tech.si wrote: > > > In article <Pine.LNX.3.95.980403114356.19940B-100000@fabre.act.qc.ca> > > Andre-John Mas <ama@fabre.act.qc.ca> writes: > > > > > > Could someone tell me where I can find some specs of the > > > TIFF file format used on OpenStep. I have was given some > > > TIFF files that originated from OpenStep and I am unable > > > to read them with any programs that supports TIFF on the > > > Mac or the PC. > > > > > The TIFF files used on OpenStep are readable by QuickTime 3, on all of > QuickTime's supported platforms. You can open the tiffs with the > PictureViewer application included with QT3 and export it to some other > format if you need to. > Works a charm! Thanks AJ
From: atlauren@uci.edu (Andrew Laurence) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Tue, 07 Apr 1998 22:48:26 -0700 Organization: Office of Academic Computing, UC Irvine Message-ID: <atlauren-0704982248270001@dialin33462.slip.uci.edu> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> _3;X4o1Yg^$]=Oc\u Xcanpos: shelf.1/199804230201!0036109777 >I guess they didn't solicit input from graduate students when forming >their new policies. I'm hardly an important developer for Apple, and am >not likely to become one, but other people like me might (Aaron Giles >comes to mind as a recent example). :-| You may find that your university was given a "Scholarship" into the Associate level developer programs. This university was, and my department serves as the coordination point for participation and distribution. Ask around - I'm sure Cornell was given the opportunity. -Andrew Andrew Laurence atlauren@uci.edu Office of Academic Computing http://www.oac.uci.edu/~atlauren/ UC Irvine
From: "C.S." <caleb@diehlgraphsoft.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 1998 19:27:22 -0400 Organization: @Home Network Message-ID: <352C07D8.35ED761F@diehlgraphsoft.com> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Lawson English <english@primenet.com> Lawson English wrote: > Rex Riley <rriley@yahoo.com> said: > > Apple's actions signal that it can't afford to get into the "handout" > > business nor support legions of "individual" developers (read welfare). ... > But shareware programmers often fill in the gaps that commercial software > leaves. Also shareware programmers learn the ropes via feedback from their > customers and are often high school/college students who later enter the > Mac programming field as professionals. Sure, that's fine. Do you need the very latest development version of this API or that extension to learn the ropes? Nah. Could you write a good piece of shareware with a couple volumes of Inside Macintosh and the CodeWarrior academic (68K compilers only) release? Sure thing. > Where's the next Stuffit or Alladin Software for MacOS going to come from > if there aren't going to be any more Raymond Lau's? Probably from the next guy who's creative and insightful enough to recognize a need in the marketplace or dream up some new tool that truly makes life easier/better/faster/cheaper/more fun for users. Probably NOT from the geek with a G3 and a truckload of prerelease software and no imagination. I'm not saying great tools and good information aren't important, but I _am_ saying that you don't need a Formula 1 and a zillion sponsors to enjoy driving, or to get from one place to another. And even if you plan on someday driving that Formula 1 and getting those sponsors, you probably won't be lucky enough to start out that way. Caleb Strockbine, speaking entirely by and for himself. caleb@diehlgraphsoft.com
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 17:38:03 -0700 Organization: Primenet Services for the Internet Message-ID: <B151677A-C971F@206.165.43.150> References: <joe.ragosta-0804981855220001@elk68.dol.net> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Joe Ragosta <joe.ragosta@dol.net> said: > > Without those, Mr. Lau might have had a hard time developing his > product... > > > > Of course, the SDKs are still online at no charge. This makes your entire > point moot (as usual). As I pointed out in an e-mail to you and everyone who had participated in this thread when I found out what you say above. --------------------------------------------------------------------- "OpenDoc may be impressive technology, but if no-one else is using it, why should Apple?" --Steve Jobs "In short: if you want to blow people away with what can be done *only* on a Mac, show them Cyberdog." --Tom Keyes <news://cyberdog.apple.com/cyberdog.general/19411> ---------------------------------------------------------------------
From: philipm@NO-SPAMeecs.umich.edu (Philip Machanick) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 1998 20:51:17 -0400 Organization: Department of EE and Computer Science, The University of Michigan Message-ID: <philipm-0804982051170001@pm1-23.eecs.umich.edu> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <352AAF57.25EB@skdjsk.no> <6geche$8lu$1@news2.apple.com> For those who are defending Apple's position, you have forgotten the one key rule of marketing. Don't piss the customer off. Now stop saying the customer is wrong, and think about how to persuade Apple to change the message. The message is the key here, not the substance. But still, on to detailed responses... In article <6geche$8lu$1@news2.apple.com>, "Paul Rekieta" <PRekieta@ix.netcom.com> wrote: >You got it -- The Apple Developer Connection Mailing. The ADC Mailing is a >12-month subscription that includes the Developer CD Series, a wealth of >technical resources, system software, development tools, technical >documentation, Interactive Media Resources, SDKs, and more. More information >can be found at http://developer.apple.com/programs/mailing.html. All very well but for people who want to get into new stuff like Yellow Box, it's not much good -- it doesn't include pre-release software. In article <joe.ragosta-0804980858150001@wil77.dol.net>, joe.ragosta@dol.net (Joe Ragosta) wrote: >Actually, there were rumors that there would be a very low price >educational plan. You might want to wait for the educational pricing >before worrying too much. I don't much care about that... the problem is small developers who will feel discouraged when trying to find and entry point. If YB is to get a good start, it needs some killer Mac-only (or at least YB-only) apps and it's hard to see big companies betting on a new API. In article <6ggj21$fit$6@anvil.BLaCKSMITH.com>, cswiger@BLaCKSMITH.com wrote: >I'm not sure people understand the purpose of seeding and beta testing programs. >Rhapsody DR1 is the first "semi-public" beta release of a complex new operating >system. DR1 is not a final product; it is not ready for production usage, nor is it >suitable as a standalone platform for serious development. [...] >However, Apple does not want to spend the time (== money) and resources to provide >extensive support for DR1 to individual people who just want to play around. In the >long run, Apple's customer base is better served by Apple using their finite >resources effectively. For a perfect example, someone over the last day or so was >asking how to make DR1 into a primary nameserver and mail exchanger. That doesn't >make any sense-- DR1 is *beta*, not production. I know someone who's using it very successfully as an alternative to other UNIX platforms. It's certainly more stable than some versions of Linux. Why shouldn't people play with it in these kinds of ways if they are willing to take the risk? If 100,000 enthusiasts hammer it, bugs will show up pretty fast. I don't call that "support", I call it "free beta testing". After all, Microsoft gets customers to beta test their new OS releases in a big way, and doesn't charge them $500+ for the privilege. >On 04/08/98, George Graves wrote: > Few, if any, of the people pissing and moaning about loosing >access to the seeding program actually NEED access to it in many >cases. Few, if any, of the people pissing and moaning about the Pentium divide error actually were effected by it but look what harm it did to Intel. Perception is big and Apple doesn't have a clue in this area (for a while I thought interim CEO for life Jobs had learned something but I'm not so sure now). In article <6ggvub$oop$5@ironhorse.plsys.co.uk>, mmalcolm crawford <malcolm@plsys.co.uk> wrote: >Apple acting as a charity subsidising people's hobbies is what's insane. >Nor should they be offering a program wide open to abuse by people signing up >as developers just to get a discount on hardware. I don't think Apple is subsidising developers. Without developers, no one would buy Macs (they tried that experiment in 1984). Anyway what's wrong with having a sneaky back door way of cutting dealers out of the loop? What has a dealer ever done for you? You don't think Apple loses on hardware sold at developer prices? -- Philip Machanick Dept. of Computer Science, Univ. of Witwatersrand on sabbatical until September 1998 at Dept EECS, University of Michigan http://www.eecs.umich.edu/~philipm/ mailto:philipm@NO-SPAMeecs.umich.edu
From: "Brad Hutchings" <brad@hutchings-software.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 98 17:53:58 -0700 Organization: EarthLink Network, Inc. Message-ID: <B1516A43-1BD20@207.217.155.14> References: <1d76snt.17ecjre1ykt2xdN@ppgutkneco.lirmm.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit nntp://news.earthlink.net/comp.sys.mac.oop.powerplant, nntp://news.earthlink.net/comp.sys.mac.programmer.codewarrior, nntp://news.earthlink.net/comp.sys.next.advocacy, nntp://news.earthlink.net/comp.sys.next.programmer >Another interesting information: the FAQ on developer.apple.com reads >that Apple has been limiting access to the Associate Programs in the >last six months... In the US only I guess, as I renewed my membership >two weeks ago and have not heard a word from EDR indicating that I might >not have what I was paying for. > Not in the US. That is total CYA BS. I know one US developer (not me) who is damned thankful he got into the program and bought the G3s he needed when he bought them. Brad <mailto: "Brad Hutchings" brad@hutchings-software.com> <http://www.hutchings-software.com>
Date: Wed, 08 Apr 1998 21:09:55 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-0804982109560001@elk62.dol.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> In article <MPG.f95d7e4b8fb1622989890@news.supernews.com>, don.brown@cesoft.com (Donald Brown) wrote: > In article <joe.ragosta-0804981906460001@elk68.dol.net>, > joe.ragosta@dol.net says... > > > Unfortunately, when details do come out, you're just as screwed as ever. . . > > > > Please explain how you're so badly screwed. > > > > The SDKs are still downloadable for free. > > > > The monthly mailing is available for $50 less than the old Associate membership. > > > > The lowest paid developer status is $500 instead of $250. If you really > > need to get the beta OS versions, your cost went up by $250. But everyone > > else seems to be better off. > > > Including Microsoft, who looks better and better every day. Anyone notice that people like Mr. brown here, who would apparently never go near a Mac from reading his posts, are the ones complaining while developers like mmalc are supporting the change? -- Regards, Joe Ragosta See the Complete Macintosh Advocacy Page http://www.dol.net/~Ragosta/complmac.htm
From: Eric Hermanson <eric@alum.mit.edu> Newsgroups: comp.sys.next.programmer Subject: WebObjects 4.0? Date: Wed, 08 Apr 1998 18:15:37 -0700 Organization: Digital Universe Corporation Message-ID: <352C2138.4835AB9F@alum.mit.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Does anyone know the approximate timeframe for the release of WebObjects 4.0 (most interested in the NT platform)? Thanks, Eric
From: j-jahnke@uchicago.edu (Jerome Jahnke) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 1998 21:41:28 -0500 Organization: University of Chicago -- BSDIS/CRT Message-ID: <j-jahnke-0804982141280001@192.168.1.3> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> In article <rmcassid-0804981509160001@dante.eng.uci.edu>, rmcassid@uci.edu wrote: > In article <steve-0804981442590001@oranoco.discoverysoft.com>, > steve@discoverysoft.com (Steven Fisher) wrote: > > >In article <B1511019-15903@206.165.43.154>, "Lawson English" > ><english@primenet.com> wrote: > > > >>Trying to make developer *support* a profit center is an INSANE idea. > >>Hopefully, either Jobs will change this, or the more rational stockholders > >>will hear about it in time to vote against him because this one action, *BY > >>ITSELF*, will destroy Apple. > > > >That's the best summary I've seen yet. The real worry is that it will do > >so *SLOWLY*... consumers probably won't see the effects this has on the > >development community for months or a year yet. Suddenly new > >shareware/freeware will start getting a lot more scarce. > > Of course the counter-argument is that Apple has never been able to really > adequately support it's developers due to the low cost of the program. > While I'm sure Adobe gets a lot of attention and I get a lot for what I > pay, what about the guys in the middle - like Rich Siegal, perhaps. Or > Aladdin Sys. The devlopers in the middle could probably handle much more > in the way of devloper support and can quite likely afford a modest > increase in costs. Now this is a bunch of bull. Many of us were Associates for one reason and one reason only, access to cheaper equipment. Apple does not lose money by selling me my one machine a year at a low price. And I can have high end hardware at a reduced cost so that I can continue to develop high quality Apple Software. I didn't call them to ask them to help me with MacOS problems we have newsgroups for that, and I participate here with others to help people along. 250 bucks is cheap to send me 20 CD's a year and give me early access to new Apple Technology. What they have done is doubled the price given me a feature I don't and probably won't use, taken away the one I do use and didn't even give me a choice to get my money back. I would agree with everyone here who supports Apple on this decision if they had said. OK, all old Associates it is business as usual. Next year however, the price is going to double and you are going to get less. How much money is Apple saving by keeping me from buying a cheap machine this year? Jer,
From: j-jahnke@uchicago.edu (Jerome Jahnke) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 1998 22:46:52 -0500 Organization: University of Chicago -- BSDIS/CRT Message-ID: <j-jahnke-0804982247240001@192.168.1.3> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <6ggaju$rts$4@news01.deltanet.com> In article <6ggaju$rts$4@news01.deltanet.com>, scott@eviews.com (Scott Ellsworth) wrote: > In article <j-jahnke-0704982342480001@192.168.1.3>, j-jahnke@uchicago.edu (Jerome Jahnke) wrote: > >In article <trumbull-0704981900290001@net44-223.student.yale.edu>, > >trumbull@cs.yale.edu (Ben Trumbull) wrote: > > > >> I'm not about to stand up and say these changes are a great idea on > >> Apple's part, but you do realize, you can buy *just* the tech mailing > >> (including develop) for $199 ? > > > >Which is what we used to do, but then realized you could not get seeded > >unless you were an associate. > > I, as well. > > Hmmm. Hell may have just frozen over - I find myself agreeding > with both of your last two posts. Boy, you sure know how to hold a grudge... Jer,
From: andrew_abernathy@omnigroup.com Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 10 Apr 1998 10:08:19 GMT Organization: Omni Development, Inc. Message-ID: <6gkr2j$r9h$1@gaea.omnigroup.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> <MPG.f960500e190219d989895@news.supernews.com> <joe.ragosta-0804982250580001@wil62.dol.net> <ericb-0904981631110001@132.236.171.104> ericb@pobox.com (Eric Bennett) wrote: > I don't see any developers defending Apple's decisions. The best > "support" I've seen is some Mac news sites parroting Apple's press > release. Well, now you have. Apple just saved me $50 a year, because the $200 mailing gives me what I want. (I'm currently a member of the old $250 program, but I don't need software seeds, so the mailing is perfect for me.) My company wants a higher level, but no problem - a $500 developer program is _easily_ justifiable to anyone doing commercial development. (Note that I said "justifiable" - we'd all prefer that everything were free, but Apple is giving excellent value for that money, and I'm all for them actually making money and therefore surviving so that I can use and develop on The Superior Platform.) --- andrew_abernathy@omnigroup.com - NeXTmail & MIME ok
Date: Fri, 10 Apr 1998 06:21:29 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-1004980621290001@elk81.dol.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <joe.ragosta-0904980642200001@elk33.dol.net> <6gitpj$csn$1@interport.net> <joe.ragosta-0904980134360001@wil68.dol.net> <Er6tuI.Mu8@AWT.NL> In article <Er6tuI.Mu8@AWT.NL>, G.C.Th.Wierda@AWT.nl (Gerben Wierda) wrote: > joe.ragosta@dol.net (Joe Ragosta) wrote: > >> Apple should be ACTIVELY ENCOURAGING these people. Some of them might > >> grow up to be software developers . . . or management . . . or sysadmins . > >> . . or end users . . . or, or, or . . . > > > >Sure. They should encourage everyone. > > > >But how do you allocate limited resources? Apple has decided that spending > >the money on TV ads or magazine ads is a better place to put scarce > >resources than subsidizing hobbyists. > > Which is a good point, but do we know the numbers? Is the cost of providing > hardware discounts and support comparable to the cost of running TV ads? This > comparison only makes sense if you know the relative cost. Price/performance > is the issue, also in marketing. I don't know--but neither do you. Apple is responsible for allocating their resources. Not you or anyone else. > > >And in case you've missed it, it is now possible for hobbyists to get > >SDK's free. How much cheaper do they need to be? > > SDK's and Mac API development don't interest me. I want to know what will > happen to the cost of developing for Rhapsody. > How about zero dollars? Download MkLinux and GNUStep. -- Regards, Joe Ragosta See the Complete Macintosh Advocacy Page http://www.dol.net/~Ragosta/complmac.htm
From: xhsoft@injep.fr (Xavier Humbert) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Fri, 10 Apr 1998 12:34:02 +0200 Organization: Maquis Usenet Ouest Message-ID: <1d78ren.1dbj70n1j99pebN@hobbit1.injep.fr> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> <1d7821x.1rdi5101enh5fhN@hobbit1.injep.fr> <6girj3$6j$2@anvil.BLaCKSMITH.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Mail-Copies-to: never Charles Swiger <chuck-nospam@blacksmith.com> wrote: > the increase in costs represents roughly a day's pay That's not the point. 100% is 100%. BTW, I'm not a developer anymore. Now I'm employed as Network Engineer. And it is MY decision for my company to buy $50 000 new Macintoshes, or new Compaqs. And my boss may find that $500 is too expensive just to evaluate new solutions. QED. Xav -- Xavier HUMBERT Laboratoire Informatique INJEP Office: labo-info/AT/injep.fr Home: humbert/AT/injep.fr
From: xhsoft@injep.fr (Xavier Humbert) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 9 Apr 1998 14:25:44 +0200 Organization: Maquis Usenet Ouest Message-ID: <1d7821x.1rdi5101enh5fhN@hobbit1.injep.fr> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Mail-Copies-to: never <rmcassid@uci.edu> wrote: > and can quite likely afford a modest > increase in costs. 100% increase is modest for you ? Xav -- Xavier HUMBERT Laboratoire Informatique INJEP Office: labo-info/AT/injep.fr Home: humbert/AT/injep.fr
From: Le_Jax@iName.com (Jacques Foucry) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 9 Apr 1998 20:07:38 +0200 Organization: Correze Software Message-ID: <1998040920073876968@dialup-202.def.oleane.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gecns$f9p$1@news01.deltanet.com> <1d767ho.1tixfqy1dqgm03N@hobbit1.injep.fr> <6gg6af$fit$1@anvil.BLaCKSMITH.com> <6ggabn$rts$2@news01.deltanet.com> <1d76snt.17ecjre1ykt2xdN@ppgutkneco.lirmm.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Olivier Gutknecht <gutkneco+news@lirmm.fr> wrote (écrivait) : > Agreed. The only explanation is that Apple is not interested anymore in > small/independant development groups. Perhaps, but without small/independant developer, no "Super Clock", no "Stickies", no new Control Strip... -- Calendriers 3.0 est disponible. Vous le trouverez sur le ouèbe de la BurpTeam <http://www.burpteam.home.ml.org>
From: steve@discoverysoft.com (Steven Fisher) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 12:01:50 -0700 Organization: Discovery Software Ltd. Message-ID: <steve-0904981201500001@oranoco.discoverysoft.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <6ggs48$oop$2@ironhorse.plsys.co.uk> <6ghf6g$e1u$1@geraldo.cc.utexas.edu> <6gi8o4$oop$9@ironhorse.plsys.co.uk> In article <6gi8o4$oop$9@ironhorse.plsys.co.uk>, mmalcolm crawford <malcolm@plsys.co.uk> wrote: >In <6ghf6g$e1u$1@geraldo.cc.utexas.edu> Kurt D. Bollacker wrote: > >> : If anyone is wanting to develop for Rhapsody professionally and doesn't >have >> : $500, they don't have a business plan. If you're intending to write >> : shareware, then $500 is a reasonable investment if you believe you have >> : something useful to contribute. If you're writing Rhapsody apps for a >hobby, >> : why should I subsidise your playtime? >> >> Because I might develop some useful freeware. With your attitude there'd >> never be any freeware at all. Without freeware, perhaps NeXT would have >> been later with NeXTSTEP (no GNU tools jumpstart) and Rhapsody today would >> not exist. Not all software should be free, but some of it always should. >> >How is not being a member of the program going to stop people from producing >freeware if they want? Why does anyone producing freeware need advance >access to beta software? I would think this extends to shareware as well. Would Default Folder have been able to support MacOS 8 as it shipped if Jon Gotow hadn't had acess to betas of MacOS 8? Would Kaleidoscope have supported MacOS 8 and MacOS 8 interface elements immediately if Greg Landweber hadn't had acess to betas of MacOS 8? Don't you feel better knowing that the freeware and shareware software you use is being tested long before the next MacOS is actually released? This stuff doesn't happen by magic. Elves don't do it in the middle of the night. It's the efforts of developers such as Greg and Jon that make sure this stuff works. Now, chances are everyone has heard of Default Folder and Kaleidoscope. Chances are that both Greg and Jon can easily afford the new programs now. But they probably couldn't have when they first started. So which new software developer will Apple block that could have become the next Andrew Welch, Greg Landweber, Jon Gotow or Raymound Lou? (Hope I spelt that last one right... if not, sorry.) -- Steven Fisher; Discovery Software Ltd.; steve@discoverysoft.com "Anyone who has never made a mistake has never tried anything new." -- Albert Einstein
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 98 18:59:30 GMT Organization: QMS Message-ID: <6gj5ps$3qf$1@news01.deltanet.com> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> In article <B1511019-15903@206.165.43.154>, "Lawson English" <english@primenet.com> wrote: >Trying to make developer *support* a profit center is an INSANE idea. >Hopefully, either Jobs will change this, or the more rational stockholders >will hear about it in time to vote against him because this one action, *BY >ITSELF*, will destroy Apple. For what it is worth, my proxy is sitting on my desk now, and I have a very brief time to get it in the mail in time for the stockholder meeting. It is a pity that there is no easy way to say, even en masse "I will vote against your appiontment to the BoD because of the following move" before the day itself. Even after voting, there is little way for them to know why a group of shareholders voted the way they did. The stockholders thus have little way of communicating why they are acting the way they are. NB, I do not think stockholders should be running a company particularly, but it is good for directors to know the reasons why the stockholders react the way they do. Scott Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
From: xhsoft@injep.fr (Xavier Humbert) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Fri, 10 Apr 1998 15:08:27 +0200 Organization: Maquis Usenet Ouest Message-ID: <1d79xzo.1eithty186myblN@hobbit1.injep.fr> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <ericb-0904981701440001@132.236.171.104> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Mail-Copies-to: never Eric Bennett <ericb@pobox.com> wrote: > There are lots of good examples like this. [snip] Your forgot Glenn Anderson's Eudora Internet Mail Server Xav -- Xavier HUMBERT Laboratoire Informatique INJEP Office: labo-info/AT/injep.fr Home: humbert/AT/injep.fr
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 98 19:13:23 GMT Organization: QMS Message-ID: <6gj6js$3qf$3@news01.deltanet.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> In article <joe.ragosta-0804981906460001@elk68.dol.net>, joe.ragosta@dol.net (Joe Ragosta) wrote: >In article <mazulauf-0804981503390001@sneezy.met.utah.edu>, >mazulauf@atmos.met.utah.edu (Mike Zulauf) wrote: > >> In article <joe.ragosta-0804980858150001@wil77.dol.net>, >> joe.ragosta@dol.net (Joe Ragosta) wrote: >> >> > Actually, there were rumors that there would be a very low price >> > educational plan. You might want to wait for the educational pricing >> > before worrying too much. >> >> The eternal Joe Ragosta reply. "All is well. Don't worry. Wait for >details." >> >> Unfortunately, when details do come out, you're just as screwed as ever. . . > >Please explain how you're so badly screwed. Because they removed a primary benefit for the current Associates without warning, and did not provide at least one of the two supposed added benefits for those being transitioned. According to the contract, they can do this without warning, but my suspicion is that the current developer contract would leave them open to an interesting class action suit. I have not yet heard back from the credit card company about whether they are willing to reverse my developer renewal. Thier comment was that as the contract explicitly states they may change any programl, and will never give either a full or partial refund, it is dicey. On the other hand, the contract seems to be invalid because it includes so many weasel words about what they do not have to do that the bank is unsure whether the contract counts as enforcable. >The monthly mailing is available for $50 less than the old Associate > membership. We do not have any indicators about whether Rhapsody will be included in those mailings. >The lowest paid developer status is $500 instead of $250. If you really >need to get the beta OS versions, your cost went up by $250. But everyone >else seems to be better off. You miss the point, I think, that the loss of hardware purchase privs in mid year is a substantial change in the benefits. Scott Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 98 19:18:03 GMT Organization: QMS Message-ID: <6gj6sk$3qf$4@news01.deltanet.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> In article <joe.ragosta-0804982109560001@elk62.dol.net>, joe.ragosta@dol.net (Joe Ragosta) wrote: >In article <MPG.f95d7e4b8fb1622989890@news.supernews.com>, >don.brown@cesoft.com (Donald Brown) wrote: > >> In article <joe.ragosta-0804981906460001@elk68.dol.net>, >> joe.ragosta@dol.net says... .. >> > The lowest paid developer status is $500 instead of $250. If you really >> > need to get the beta OS versions, your cost went up by $250. But everyone >> > else seems to be better off. >> Including Microsoft, who looks better and better every day. >Anyone notice that people like Mr. brown here, who would apparently never >go near a Mac from reading his posts, are the ones complaining while >developers like mmalc are supporting the change? Don't be an ass, Joe. A fair number of Real Mac developers are rather vexed at the change. By simple extension, you seem to be claiming that those who are complaining are those who do not develop Mac apps. I submit a good 150KLOC of Macintosh C++ code and a shipping, though dying, Mac app to the contrary. I know this makes it harder for me to get Macs back in this shop. I also know that it makes it harder for me to develop for them on a personal level. This combination is not good. Scott Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
From: "Brad Hutchings" <brad@hutchings-software.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 98 12:17:27 -0700 Organization: EarthLink Network, Inc. Message-ID: <B1526CE7-65D6E@207.217.155.85> References: <01bc455b$1be404c0$18f0bfa8@davidsul> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit nntp://news.earthlink.net/comp.sys.mac.oop.powerplant, nntp://news.earthlink.net/comp.sys.mac.programmer.codewarrior, nntp://news.earthlink.net/comp.sys.next.advocacy, nntp://news.earthlink.net/comp.sys.next.programmer >I think it would be great if they get sued, Apple being the one whose legal >department threatens web sites like macintouch, and they do something like >this. I don't know anyone who has the stomach to sue over $250. And even if one could get 5000 developers to participate, it's likely they'd have to collectively assume liability for legal costs if they lose and that they'd get anything more than a paralegal working on the case. Each individual developer needs to evaluate whether $500 is a good deal for the services offered. I know I'll be making that evaluation in July. >I also hope several developers pass this on to the media, I am sure the >media would love to flame apple for this and god knows apple deserves a >good spanking for this A spanking over a fairly reasonable tactical move sends the wrong message. This move would not have been met with intense cynicism if developers thought Apple had an overall clue and direction and appreciation of how developers support Apple. It starts from the top, and unfortunately, is beyond the control of ADR. Brad <mailto: "Brad Hutchings" brad@hutchings-software.com> <http://www.hutchings-software.com>
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 98 19:31:38 GMT Organization: QMS Message-ID: <6gj7m3$3qf$5@news01.deltanet.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <6ggaju$rts$4@news01.deltanet.com> <j-jahnke-0804982247240001@192.168.1.3> In article <j-jahnke-0804982247240001@192.168.1.3>, j-jahnke@uchicago.edu (Jerome Jahnke) wrote: >In article <6ggaju$rts$4@news01.deltanet.com>, scott@eviews.com (Scott >Ellsworth) wrote: > >> In article <j-jahnke-0704982342480001@192.168.1.3>, >j-jahnke@uchicago.edu (Jerome Jahnke) wrote: >> >In article <trumbull-0704981900290001@net44-223.student.yale.edu>, >> >trumbull@cs.yale.edu (Ben Trumbull) wrote: >> > >> >> I'm not about to stand up and say these changes are a great idea on >> >> Apple's part, but you do realize, you can buy *just* the tech mailing >> >> (including develop) for $199 ? >> > >> >Which is what we used to do, but then realized you could not get seeded >> >unless you were an associate. >> >> I, as well. >> >> Hmmm. Hell may have just frozen over - I find myself agreeding >> with both of your last two posts. > >Boy, you sure know how to hold a grudge... To be fair, I was mostly in a pissed off mood regarding Apple's recent licensing, and let that slip into my phrasing on other matters. Since I said the above in a public place, the apology belongs there too. My apologies. Scott Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 12:51:01 -0700 Organization: Primenet Services for the Internet Message-ID: <B15275E3-3AE0B@206.165.43.139> References: <j-jahnke-0904980214320001@192.168.1.3> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Jerome Jahnke <j-jahnke@uchicago.edu> said: > > > > Perhaps. But they do still provide you with all the other > > stuff, the CD's, the mailings etc.. that does cost Apple money. > > It don't cost 500 bucks... Won't the latest version of Rhapsody come with the subscription? $250/year was enough to prevent most people from purchasing a developer CD subscription just to get ahold of the latest versions of MacOS when they were distributed. $500/year probably is deemed enough to prevent most people from purchasing the new developer package in order to get the latest update of Rhapsody OS. That's the only rational explanation for this that I can see since you just can't justify charging less than cost for producing the developer product but more than distribution cost unless it is to control access to Rhapsody via pricing. Of course, this is Steve Jobs that we're talking about and he may really believe that developers should "pay their fair share." [those @#$%&^&* leeches!]... --------------------------------------------------------------------- "OpenDoc may be impressive technology, but if no-one else is using it, why should Apple?" --Steve Jobs "In short: if you want to blow people away with what can be done *only* on a Mac, show them Cyberdog." --Tom Keyes <news://cyberdog.apple.com/cyberdog.general/19411> ---------------------------------------------------------------------
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 98 19:45:04 GMT Organization: QMS Message-ID: <6gj8f9$3qf$6@news01.deltanet.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> <6ggqha$9fk$1@news01.deltanet.com> <6ghg4o$np$1@news.digifix.com> In article <6ghg4o$np$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: >On 04/08/98, Scott Ellsworth wrote: >>In article <6ggmk2$jdl$1@news.digifix.com>, sanguish@digifix.com >(Scott Anguish) wrote: ><snip> .. >>I know two other developers at former cross platform shops doing >>much the same. God only knows if this will work, but it is the >>case that QMS has already decided to drop Mac development, >>so they are a lost cause if I cannot change that. >> > > If they're being so petty as to not be willing to pay the >$500, then they should talk to Apple. Although realistically I know >of several large commercial developers who never take advantage of >Apple's programs, and use CodeWarrior for all development. I may have misphrased the above. Let me try again. QMS is not being petty. They have decided to drop Mac development unless someone here convinces them otherwise. That means all of the resources expended to justify more Mac development come out of my pocket. They are also unconvinced by anything other than a demo of a working product, so the only help Apple can give me are tools that I can use to Gee-Whiz them back into interest. The seeds were a way of doing that, because Rhapsody is quite neat in concept. They were reasonably receptive to NextStep five years ago, after all. Clearly, I cannot show them the seeds before Apple ships CR1, but I can prepare what I am going to use. Having access to them gives me a chance to be ready, so that whatever media presence Apple can garner will be happening when I have something available, instead of six months later. Like I said, this may not be relevant, but it is a reason why I found it worth $250 to get the seed materials, and may find it worth $500. Were I certain, absolutely certain, that Rhapsody would be a part of the standard developer mailing CDs, then I might decide the lost time is acceptable, and switch to mailings instead. The recent changes, though, have entered some doubt in my mind - it is quite clear that they want to position Rhapsody differently, and a case might be made that the general developer does not need access. We do not know what will happen, but this is the first time that they have clearly made a delinitation between those who have cheap access, and those who have expensive, in terms of information, instead of just human contact. The unilateral drop in purchase privs also reveals that there is no way to plan your own needs, as when Rhapsody nears shipping, they will doubtless make decisions, and pre- existing agreements will hold little weight. >>>Few, if any, of the people pissing and moaning about loosing >>>access to the seeding program actually NEED access to it in many >>>cases. >> >>Of course, they changed the rules in mid year, such that anyone >>already paying for a developer subscription got substantially >>fewer benefits. (Specifically, dropping the hardware agreement >>unlaterally is far from a nice move, and changes the economic >>impact of the developer program substantially.) > > Have you purchased under this program? You can get virtually >equivalent prices elsewhere. This is a valid point. My own readings got different results, but we may be quibbling over details. Scott Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
From: trumbull@cs.yale.edu (Ben Trumbull) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 15:54:15 -0400 Organization: Yale University Message-ID: <trumbull-0904981554150001@net44-223.student.yale.edu> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <352AAF57.25EB@skdjsk.no> <6geche$8lu$1@news2.apple.com> <philipm-0804982051170001@pm1-23.eecs.umich.edu> <6gj0or$oop$12@ironhorse.plsys.co.uk> In article <6gj0or$oop$12@ironhorse.plsys.co.uk>, mmalcolm crawford <malcolm@plsys.co.uk> wrote: > > You don't think Apple loses on hardware sold at developer prices? > > > Compared to hardware sold at retail prices, yes. that's pure fiction. Retailers take a huge chunk out of Apple's net on those sales. And the hardware discounts were never very impressive, imho. Think about. Does Apple make more money if Apple sells X a Mac or if Cyberian Outpost sells X a Mac ? terminally curious, Ben ___________________________________________________________________ Benjamin Trumbull trumbull@cs.yale.edu Yale University You can't be in hell; you can still read your e-mail
From: ericb@pobox.com (Eric Bennett) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 17:03:41 -0400 Organization: Cornell University Sender: emb22@cornell.edu (Verified) Message-ID: <ericb-0904981703410001@132.236.171.104> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <352C07D8.35ED761F@diehlgraphsoft.com> In article <352C07D8.35ED761F@diehlgraphsoft.com>, "C.S." <caleb@diehlgraphsoft.com> wrote: > Probably from the next guy who's creative and insightful enough to > recognize a need in the marketplace or dream up some new tool that > truly makes life easier/better/faster/cheaper/more fun for users. > Probably NOT from the geek with a G3 and a truckload of prerelease > software and no imagination. Those people get the OS through illegal channels anyway. They don't pay Apple for developer subscriptions. -- Eric Bennett (www.pobox.com/~ericb) Cornell University Biochemistry Department
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 98 20:55:10 GMT Organization: QMS Message-ID: <6gjcin$938$1@news01.deltanet.com> References: <6gj0or$oop$12@ironhorse.plsys.co.uk> <B1527A55-4B993@206.165.43.139> In article <B1527A55-4B993@206.165.43.139>, "Lawson English" <english@primenet.com> wrote: >mmalcolm crawford <malcolm@plsys.co.uk> said: >Now, they ARE subsidizing developers on the software-tools side, by not >charging the full development costs for developing the SDKs and technical >materials, but if they were to do that, we would be back in the bad ole >days of Spindler, where an ETO subscription was many hundreds/year and you >had to pay $1500 up front to get the complete package of MPW >C/Pascal/C++/MacApp/etc. > > >Now they've raised the bar to ensure in >> part that the system isn't abused. For the sorts of commercial >> applications that I suspect you're referring to, on which the platform >> depends, if the business seriously needs access to seed software >> but cannot afford a $500 a year, it does not have a competent >> business plan. > >You may be correct about raising the bar to see that the hardware discount >system isn't abused. However, if they raise the bar sufficiently, then >there is no longer a discount on the hardware in the first place, is there? Two useful points should be brought up. First, what is the magunitude of the hardware discount, vs. the lowest price available from a mail order retailer, such as club Mac? Well, we can try to make a guess. If they were charging more than Club Mac, then developers would have been more vocal about the prices. Comparing prices to the Apple Store, it looks like the club Mac margin, at least on the two G3/266 systems I priced, runs on the order of 5%. If the developer discounts are on the order 5% in addition, then developers may have a case. If the differences are on the order %10, they almost certainly do, and if they are on the order %15, then they should be livid. I have heard on the rumors page that the Apple margin to a place like Club Mac seems to be on the order 20%, which Club Mac chooses to take as a 5% price discount, and on the order 15% revenue. This matches my experience when bargaining with them and with Creative Computers. Unfortunately, the Apple hardware price list is stamped confidential, and while it would come out in the courts, developers should probably try to keep it confidential. It does seem, though, that it should be easy for any developer to verify the analysis, and figure out just what the margin really is. Assuming that they give developers a generous 15% drop below Club Mac, they are still making as much as they would have made from a sale via a reseller. I would be surprised if any company made a program that generous. The second point: the new plan limits those getting systems to 10 per year. This makes it unlikely that developers are going to be seeing a positive expected value for the hardware purchase deal in the new program. This makes some sense, but it seems to me [OPINION] that it is till in Apple's interest for there to be developers signed up, even if they are not making money, as long as they are not losing. Scott Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
From: "David V. Duccini" <duccini@bpsi.net> Newsgroups: comp.sys.next.software,comp.sys.next.programmer Subject: Looking for Oracle Adaptor Source Date: Tue, 7 Apr 1998 10:12:24 -0500 Organization: Orbis Internet Services Message-ID: <Pine.GSO.3.96.980407100455.20831F-100000@ra.bpsi.net> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Xcanpos: shelf.1/199804211801!0028795779 Can anyone send or point me in the right direction to find the source code for the Oracle Adaptor for NextStep 3.3 ??? We have a bunch of DBKit apps that need to be upgraded to SQLNet 2.0 and because of time constraints re-writing them for EOF 1.1 (which we also have) is not really an option we would like to pursue. As far as we can tell the only real barrier to running the DBKit apps on SQLNet 2.0 is the damn login string is different. NOTE: We tried hacking up the _sanitize function in the "OracleAdaptor" by changing the text to _Sanitize so that we could include our own...but of course the 'library' is check-summed -duck ----------------------------------------------------------------------------- duccini@bpsi.net BackPack Software, Inc. www.backpack.com +1 612.645.7550 voice BPSI Internet Services www.bpsi.net +1 612.645.9798 fax 1.800eMail info@one800.net www.one800.net RoadBlock Anti-Spam System www.roadblock.net -----------------------------------------------------------------------------
From: news@net2phone.com (Net2phone Information) Newsgroups: comp.sys.next.programmer Subject: Warning about the FCC and Internet Taxes Date: Wed, 8 Apr 1998 0:34:46 -0500 Organization: NET2PHONE Message-ID: <622892010086@net2phone.com> This is not spam in the strictest sense of the definition, it's about the FCC and more taxes that are looming ahead. The method of delivery is spam-like and for that we apologize. We apologize for bothering you, but if we don't warn you right now, you might not be able to communicate over the net as freely as you do now. The Federal Communications Commission is planning to regulate the Internet by imposing new universal service fees which are likely to be passed on to you in the form of higher Internet-service charges. If these taxes are imposed, you will have to pay more every time you use the Internet. IDT Corporation (NASDAQ IDTC) a leading telecommunications and Internet access company, wants to make sure this does not happen. We are calling for a citizens uprising against government bureaucracy and taxes. IDT is going to let you voice your opinion on our nickel. We are sponsoring free phone calls to lobby your representative, senator or the FCC against this most serious infraction against affordable communications. By going to http://www.net2phone.com you can place as many free calls as you like to Congress and the FCC to demand your right to low-cost communications. We have always viewed the Internet as a fantastic medium to communicate with anyone worldwide at little or no cost. Please do not let the FCC take this right away from you. Please make that free call right now before you lose your right to unfettered Internet access. Thank you, IDT/Net2Phone Management http://www.net2phone.com
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program -shareware Organization: Is a sign of weakness Message-ID: <SCOTT.98Apr8221520@slave.doubleu.com> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggkg1$90k$1@news0-alterdial.uu.net> In-reply-to: "M. Kilgore"'s message of Wed, 8 Apr 1998 14:52:21 -0500 Date: 10 Apr 98 16:30:47 GMT In article <6ggkg1$90k$1@news0-alterdial.uu.net>, "M. Kilgore" <mkilgore@nospam.prysm.net> writes: There is a very good reason for Apple to discourage the small or shareware developer - the sad fact is that if it's too easy to get into the shareware market then less than knowlegeable programmers can play hell with your system. All you need do is look at the Win shareware to discover this. Pehaps Rhap isn't as bullet proof as the pr would have us believe and Apple is just trying to protect the system from becoming just another Windows. The problem with this argument, if someone were to use it, is that it effectively means that nothing happens in the boundary between apps and system. Nobody is going to start a large company with the goal of screwing around with extending the system in odd ways. Most shareware of that sort gets started less because someone had a goal starting out, and more because they noticed a facility and started building something based on it. If nobody ever notices the facility, nothing will ever get built on it, good _or_ evil. There is, however, a far better reason to encourage shareware/freeware - many of the guys that get started in that do it because they're enthusiastic about their machines. That enthusiasm is is contageous and gets even more people interested in a platform. It's worse than that. Shareware and freeware are often the farm teams for more commercial developers. There are the various semi-mythical beasts who have successfully made the leap from shareware/freeware status to full commercial status. But more often, shareware/freeware developers are just cutting their teeth, and later you'll find them at Apple, or some other larger, more well-funded site. Given two new hires fresh out of college, one of which has a fair covey of praised freeware, the other with good grades, which gets the more important job? Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Organization: Is a sign of weakness Message-ID: <SCOTT.98Apr9091304@slave.doubleu.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <xhsof-0804981448100001@hobbit1.injep.fr> <6ggs48$oop$2@ironhorse.plsys.co.uk> In-reply-to: mmalcolm crawford's message of 8 Apr 1998 22:01:44 GMT Date: 10 Apr 98 16:30:50 GMT In article <6ggs48$oop$2@ironhorse.plsys.co.uk>, mmalcolm crawford <malcolm@plsys.co.uk> writes: In <xhsof-0804981448100001@hobbit1.injep.fr> Xavier Humbert wrote: > If developers have to pay 500 bucks to have Rhapsody, do you > think the will develop for it ? If anyone is wanting to develop for Rhapsody professionally and doesn't have $500, they don't have a business plan. If you're intending to write shareware, then $500 is a reasonable investment if you believe you have something useful to contribute. If you're writing Rhapsody apps for a hobby, why should I subsidise your playtime? Most shareware becomes shareware after it's been written, not before. "Hey, this is a neat OS facility, I bet I could..." "What do you know, I can!" "Hey, that's a really useful little utility you've written, you should sell it." ... In the bad old days when NeXT's developement tools alone cost seven times that amount there was no shortage of freeware or shareware. Back in 1990, I wouldn't have written Stuart if I'd had to pay $500 upfront to do so. A year later, I willingly dropped ten times as much on a NeXTstation based on the results. I don't think there was much shareware released by people developing on $5k NS3.3/Intel developer licenses, and what did was spinoff material from someone with a real job (as opposed to a newbie exploring the possibilities). As I recall, most of the shareware from unknowns came off black hardware. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Organization: Is a sign of weakness Message-ID: <SCOTT.98Apr9091856@slave.doubleu.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> In-reply-to: mmalcolm crawford's message of 9 Apr 1998 01:40:49 GMT Date: 10 Apr 98 16:30:51 GMT In article <6gh8v1$oop$6@ironhorse.plsys.co.uk>, mmalcolm crawford <malcolm@plsys.co.uk> writes: In <01bc4540$d4a15e60$3af0bfa8@davidsul> "macghod" wrote: > > The lowest paid developer status is $500 instead of $250. If > > you really need to get the beta OS versions, your cost went up > > by $250. But everyone else seems to be better off. > > More Ragosta FUD. "Everyone else seems to be better off". Who > is better off? Apple's customers who no longer have to subsidise hobbyist developers. Where's the incremental cost of making developer tools freely available to all comers? The developer hardware discounts going away isn't a problem - the problem is that the hardware cost isn't in line with market costs. So far as tech support, that's generally sucked for anyone who doesn't have a rep assigned to them anyhow. But the tools themselves should be free. It's well worth Apple's time to spend a small amount to spread those tools even to people who have no apparent ability to use them. Recognizable developers are going to build recognizable apps. It's often those who aren't obvious developers who will build the apps you didn't expect. [I should note that I could care less about SDKs and whatnot. I want Rhapsody, and that's all I care about. If they sell Rhapsody with developer tools for, say, $100, then I'm not concerned. At $500, I'm concerned, but willing to play along. At $1000, I'm thinking that there's not much future in it and perhaps I should look at a more capable _and_ cheaper OS.] Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: *johnnyc*@or.psychology.dal.ca (John Christie) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Fri, 10 Apr 1998 14:11:03 -0300 Organization: ISINet, Nova Scotia Message-ID: <*johnnyc*-1004981411030001@jchristie.halifaxcable.dal.ca> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <6ggaju$rts$4@news01.deltanet.com> <j-jahnke-0804982247240001@192.168.1.3> Let's get down to brass tacks A list of all those that plan to stop developing for the Mac because of this Please? Just enter the name of the software that you have written and a "not anymore" next to it. 1) ? The worst thing Apple did here was change peoples current contracts midstream. Anything other than that that is really serious? Does that mean the end of development for the Mac platform? The only major platform other than Linux that one can just start developing on for nothing (after buying the machine and a net hook up). As I understand it it was built into the contracts that Apple could change the plan anytime they wanted. It isn't nice, but read what you sign (although I do feel that a class action suit by those people could be warranted). So, let's start filling in the list and see how this hurts the Macintosh. -- You aren't free if you CAN choose - only if you DO choose. All you are is the decisions you make. Remove "*" and ohnny (i.e. jc@) to reply via email
From: russotto@wanda.pond.com (Matthew T. Russotto) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 7 Apr 1998 21:41:04 GMT Organization: Ghotinet Message-ID: <6ge6hh$9mv@netaxs.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> In article <01bc445f$744397e0$1bf0bfa8@davidsul>, macghod <macghod@concentric.net> wrote: }WTF is up with the apple developer program? I see a big announcement, and }all it is is: }1st level: online stuff, just as it is now, ie NO CHANGE }2nd level: instead of costing $250 a year its now $500 a year??!?!?? } }As far as I can tell it adds nothing and is a 2x increase!! Why to support }development apple :) Yeah, sucks, doesn't it? All I really want is the CDs, and they keep upping the price and adding things I don't need. -- Matthew T. Russotto russotto@pond.com "Extremism in defense of liberty is no vice, and moderation in pursuit of justice is no virtue."
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Followup-To: comp.sys.mac.advocacy,comp.sys.next.advocacy Date: 10 Apr 1998 18:09:26 GMT Organization: P & L Systems Message-ID: <6gln8m$oop$13@ironhorse.plsys.co.uk> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggrql$oop$1@ironhorse.plsys.co.uk> <6gisof$bht$1@interport.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: float@interport.net NOTE: FOLLOWUPS TO COMP.SYS.MAC.ADVOCACY,COMP.SYS.NEXT.ADVOCACY In <6gisof$bht$1@interport.net> float@interport.net wrote: > mmalcolm crawford (malcolm@plsys.co.uk) wrote: > > : Why does anyone *need* to be part of the program if they're not developing > : professional or semi-professional apps. Sure, access to the latest and > : greatest betas of the s/w is fun, but not essential if you're hacking at home > : for the joy of it. Apple is not a charity -- and I don't want to have to pay > : more for my Mac to subsidise someone else's hobby. > > Have you ever heard of "mindshare"? Sure, it's a buzzword, but > dereference it and you find something important. > Yes, I've heard of mindshare. > If I had free or cheap access to Rhapsody technology, I would put it on a > spare PC. If I like Rhapsody, my next machine might be a Mac; the other > option is Sun's Darwin (the $3000 Ultra 5). > Fine. I hope Rhapsody is priced in such a way that you'll find it affordable. > I'm not about to write any killer apps for anybody's operating system, > OK, so why are you interested in Apple's developer program? Best wishes, mmalc.
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Followup-To: comp.sys.mac.advocacy,comp.sys.next.advocacy Date: 10 Apr 1998 18:36:56 GMT Organization: P & L Systems Message-ID: <6glos8$oop$17@ironhorse.plsys.co.uk> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <trumbull-0904980039090001@net44-223.student.yale.edu> <6gio5m$oop$11@ironhorse.plsys.co.uk> <ericb-0904981637520001@132.236.171.104> <6gjgm5$bfj$1@news01.deltanet.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: scott@eviews.com NOTE: FOLLOWUPS TO COMP.SYS.MAC.ADVOCACY,COMP.SYS.NEXT.ADVOCACY In <6gjgm5$bfj$1@news01.deltanet.com> Scott Ellsworth wrote: > In article <ericb-0904981637520001@132.236.171.104>, ericb@pobox.com (Eric Bennett) wrote: > >In article <6gio5m$oop$11@ironhorse.plsys.co.uk>, mmalcolm crawford > ><malcolm@plsys.co.uk> wrote: > > > >> No, I don't. I'm quite sure there are a number of hobbyists developing > >> worthwhile s/w which I might find useful at some point, however on balance > >> reports suggest that the previous system was badly abused and on average > >> Apple will have lost revenue through people signing up as developers simply > >> to get hardware at reduced prices. > > > >That doesn't give Apple the right to alter contracts. If they want to > >decrease the discount, or raise the fees for renewals or new accounts, > >fine. But they're cutting off *existing* subscribers. > > Like I have said iin other posts, had they continued the former > rights and privledges until the end of my developer contract, I > would merely think this an unwise change, but they have > changed it unilaterally, for the worse, and without sufficient > (i.e. before the last renewal) warning, imho. > According to Jordan J. Dea-Mattson Senior Partnership & Technology Solutions Manager Apple Developer Relations In the case of Associates (who got no tech support, and paid $250, they were moved to the Select program with no tech support incidents. These folks get the $100.00 Metrowerks Coupon, so they just were given $100.00. In the case of Associates Plus who got a limited number of tech support incidents, and paid $500, they were moved to the Select program with tech support incidents. These folks get the $100.00 Metrowerks Coupon, so they just were given $100.00. In the case of Partners, who paid $1500.00, they were moved to the Premier program. These folks get the $300.00 Metrowerks Coupons and a pass to WWDC ($1000.00 value), so they just were handed $1300.00. Apple has not been cashing anyone's checks on new enrollments or renewals for over six months. Everyone who has had to renew or join the program in the last six months has gotten it gratias. Jordan finishes by asking "Is anyone mentioning this fact?" The answer is clearly "No." People would rather yell at Apple and bitch about something they haven't learned enough about. mmalc.
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Followup-To: comp.sys.mac.advocacy,comp.sys.next.advocacy Date: 10 Apr 1998 18:44:38 GMT Organization: P & L Systems Message-ID: <6glpam$oop$18@ironhorse.plsys.co.uk> References: <joe.ragosta-0804982109560001@elk62.dol.net> <B151824F-7635E@207.217.155.14> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: brad@hutchings-software.com NOTE: FOLLOWUPS TO COMP.SYS.MAC.ADVOCACY,COMP.SYS.NEXT.ADVOCACY In <B151824F-7635E@207.217.155.14> "Brad Hutchings" wrote: > > mmalc, on the other hand (and to be a little picky here), is one of our new > NeXT friends. Believe me, if I were the slightest bit worried about Apple's > direction w.r.t. the technology I totally depend on for the future of my > business, I'd be kissing every butt at every turn. Not that I'm accusing > mmalc or the rest of the NeXT contingent of doing that, I'm just saying > that's what _I_ would do that if _I_ were a Rhapsody developer today. With > a number of "legitimate" commercial developers shaking their heads about > the hardware thing and the price increase, I'd do everything to encourage > Apple management to have a positive view of what I'm doing. > The implication you are making here is offensive. I ask you to retract it. I can assure you that on other matters (publicly or privately) I have been *very* critical of Apple. mmalc.
From: rriley@yahoo.com (Rex Riley) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 16:28:05 GMT Organization: @Home Network Message-ID: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: jkheit@xtdl.com In <352b5da2.0@206.25.228.5> John Kheit wrote: > j-jahnke@uchicago.edu (Jerome Jahnke) wrote: > > In article <trumbull-0704981900290001@net44-223.student.yale.edu>, > > trumbull@cs.yale.edu (Ben Trumbull) wrote: > > > I'm not about to stand up and say these changes are a great > > > idea on Apple's part, but you do realize, you can buy *just* > > > the tech mailing (including develop) for $199 ? > > > Which is what we used to do, but then realized you could not get > > seeded unless you were an associate. > > What's seeded mean? Does that mean no beta's? > Some people are under the assumption that Apple's Developer Program is tasked with the "mission" to multiply and procreate a whole cadre of wonderful developers and new applications for Rhapsody. Seeded means that Apple must grow all things naturally (ie. Ye reap what ye sow). This too is an assumption, that Apple would seed "individual" developers with "the code". Apple's actions signal that it can't afford to get into the "handout" business nor support legions of "individual" developers (read welfare). By "raising-the-bar" Apple introduces an entry fee and seeds the concept of "profit" in its Developer base. Developers who approach Rhapsody without resources and a profit motive will be annoyed, discouraged and drop off. In addition to Apple selectively choosing where to sow Rhapsody seeds, expect an aggressive Developer program which fertilizes "porting" applications and "bridging" interoperability between _platforms_, _suites_ and _drivers. Apple understands that Rhapsody consumes a "resource load" that demands both mass and momentum which established Development houses with existing product in the market can sustain. Apple's Developer program is tasked with the mission to identify and demonstrate to established Development houses "how" Rhapsody can "multiply" their profits. Apple's Dev Program must "procreate" success after success in "bellwether" industries critical to Rhapsody "positioning" in Apple's BusinessSpace. What Apple Developer Program seeds is neither leadership nor a grand vision around which anyone can "rally" . So Apple's unilateral actions lack context and seed resentment and hostility... What does "no beta" mean? It... just kidding. :-) -r Rex Riley
From: xhsoft@injep.fr (Xavier Humbert) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 8 Apr 1998 14:25:22 +0200 Organization: Maquis Usenet Ouest Message-ID: <1d767e0.8dith31p5h0pxN@hobbit1.injep.fr> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <352AAF57.25EB@skdjsk.no> <6geche$8lu$1@news2.apple.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Mail-Copies-to: never Paul Rekieta <PRekieta@ix.netcom.com> wrote: > You got it -- The Apple Developer Connection Mailing. The ADC Mailing is a > 12-month subscription that includes the Developer CD Series, a wealth of > technical resources, system software, development tools, technical > documentation, Interactive Media Resources, SDKs, and more. More information > can be found at http://developer.apple.com/programs/mailing.html. Stop your advertinsing bullshit, please. Better try to explain why Next^H^H^H^HApple decided to double the price of the programs. For less stuff ! Xav -- Xavier HUMBERT Laboratoire Informatique INJEP Office: labo-info/AT/injep.fr Home: humbert/AT/injep.fr
From: rdieter@math.unl.edu (Rex Dieter) Newsgroups: comp.sys.next.programmer Subject: OS 4.2 and libg++ static-libs/dylibs Date: 10 Apr 1998 19:00:46 GMT Organization: University of Nebraska-Lincoln Message-ID: <6glq8u$qmn$1@unlnews.unl.edu> I've ALMOST finished creating dynamic library version of libg++-2.7.2 for Openstep for Mach (4.2) (and Rhapsody), but it doesn't quite work fully. It appears that I've found a dynamic linker (dyld) bug. In brief, when linking a c++ program against the dylib version of libg++, it sucessfully creates a binary. This binary then fails to execute because of undefined variables at run-time. ??? The same binary executes successfully if linked against a static version of libg++. The undefined variables/functions are: ___builtin_vec_new ___builtin_vec_delete These are referenced by libg++, and these appear to be in /lib/libcc_dynamic.a. I have a test program that dies with "Undefined symbols" even though these symbols appear to BE defined in the binary > ./tFile dyld: ./tFile Undefined symbols: ___builtin_vec_delete ___builtin_vec_new > nm tFile | grep ___builtin_vec_ 00006790 t ___builtin_vec_delete 00006728 t ___builtin_vec_new Is this a dyld bug? Any ideas? (BTW, this error occurs similarly on Rhapsody DR1/x86) -- Rex A. Dieter rdieter@math.unl.edu (NeXT/MIME OK) Computer System Manager http://www.math.unl.edu/~rdieter/ Mathematics and Statistics University of Nebraska-Lincoln
Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program From: phenix@interpath.com (John Moreno) Message-ID: <1d761b5.13ha0xu1i2sd4wN@roxboro0-043.dyn.interpath.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <01bd6267$fa532b30$9b2168cf@test1> <352ad905.0@206.25.228.5> Organization: phenix@interpath.com ?'D1cB]L9V=\OS(\8S,5 Mail-Copies-To: never Date: Wed, 08 Apr 1998 16:10:59 GMT NNTP-Posting-Date: Wed, 08 Apr 1998 09:10:59 PST In comp.sys.next.advocacy John Kheit <jkheit@xtdl.com> wrote: > "Todd Heberlein" <todd@NetSQ.com> wrote: > > > Agreed. I might have considered getting Rhapsody under the > > > previous program. I might be able to scrounge up $250. $500 > > > is out of my price range. I guess they didn't solicit input > > > from graduate students when forming their new policies. I'm > > > hardly an important developer for Apple, and am > > > As a student, you will probably be able to get the full development > > environment quite inexpensively from Apple, but we will have to > > wait a while for Apple to make that move. The last time I received > > word from Apple (less than a month ago), they had not hammered > > out all the license deals with third parties. > > > Also, $500 is still quite cheap. As I mentioned in another > > thread, yesterday I paid $1300+ for a license to use Digital's > > command-line debugger (which is pretty lame). > > > I am also a Sun Catalyst member, but I still pay a fairly high > > price for Sun's development tools. > > Yea, and look at the marketshare and number of developers for that > platform vs. the mac. I don't know the details of this program > change yet, so I'll reserve final judgement till then. But *if* > they cut services and doubled the costs to developers, I'd say this > is the proud continuation of BOZO behavior at Apple. This about sums it up - specifically they are cutting the hardware discount program with exactly the people who need it most -- small developers (not that it's been a significant discount for years, approximately the same price as their retail dealers). I don't need the two free "incidents" of tech support - I find that I can get most of that answered via usenet, but the price increase coupled with the elimination of the discount -- why bother? -- John Moreno I am trying to convince the author of YA-NewsWatcher that the latest version should be released to the public. He doesn't think there's much interest in a new version. Help me prove him wrong. To do so, send me mail <mailto:phenix@interpath.com> with a Subject of New YA. Comments on what you like/dislike in the current version will be appreciated.
From: "NeXT Newbie" <macghod@concentric.net> Newsgroups: comp.soft-sys.nextstep,comp.sys.mac.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Newby needs help Date: 10 Apr 1998 19:02:10 GMT Distribution: inet Message-ID: <01bc46e5$6b4880a0$24f0bfa8@davidsul> I am having a real hard time getting software for OS 4.2. ppp is not working, and I guess I need to download ppp from peak first? So I download ppp and some omni stuff, save it to a fat hard disk, logonto Openstep 4.2, and try to decompress the software. Doesnt work! logon_~1.tar omnico~1.gz omnipd~1.tar next-p~1 omnifr~1.gz omniwe~1.tar localhost:10# mv ne* next-ppp localhost:11# tar -xvf next-ppp ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/PPP-2.2.pkg/: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/PPP-2.2.pkg/PPP-2.2.tar.Z: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/PPP-2.2.pkg/PPP-2.2.info: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/PPP-2.2.pkg/PPP-2.2.bom: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/PPP-2.2.pkg/PPP-2.2.sizes: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/PPP-2.2.pkg/PPP-2.2.post_install: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/README.hppa: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/hppaSerialPatch.tar.gz: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/README.NeXT.MAB.Installation: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/Persistent_Connection: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/README: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/flow-control-hints: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/ip-down.example: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/ip-up.example: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/options: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/pdial: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/ppp_multiple_hosts.tar.gz: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/pppdown: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/pppkill.c: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/pppup.annex: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/pppup.direct: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/pppup.portmaster: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/pppup.remote: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/pppup.zyxel: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/rc.ppp: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/redial.sh: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/NXHosting_with_PPP: File name too long localhost:12# ls logon_~1.tar omnico~1.gz omnipd~1.tar next-ppp omnifr~1.gz omniwe~1.tar localhost:13# gunzip omnico~* localhost:14# ls logon_~1.tar omnico~1 omnipd~1.tar next-ppp omnifr~1.gz omniwe~1.tar localhost:15# tar omnico~1 tar: n: unknown option tar: usage: tar -{txru}[cvfblmhopwBi] [tapefile] [blocksize] file1 file2... localhost:16# tar -xvf omnico~1 OmniComponents.pkg: File name too long tar: can't create OmniComponents.pkg/: File name too long OmniComponents.pkg: File name too long tar: can't create OmniComponents.pkg/OmniComponents.tar.Z: File name too long OmniComponents.pkg: File name too long tar: can't create OmniComponents.pkg/OmniComponents.info: File name too long OmniComponents.pkg: File name too long tar: can't create OmniComponents.pkg/OmniComponents.bom: File name too long OmniComponents.pkg: File name too long tar: can't create OmniComponents.pkg/OmniComponents.sizes: File name too long localhost:17# ls logon_~1.tar omnico~1 omnipd~1.tar next-ppp omnifr~1.gz omniwe~1.tar localhost:18# tar -xvf next-ppp ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/PPP-2.2.pkg/: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/PPP-2.2.pkg/PPP-2.2.tar.Z: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/PPP-2.2.pkg/PPP-2.2.info: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/PPP-2.2.pkg/PPP-2.2.bom: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/PPP-2.2.pkg/PPP-2.2.sizes: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/PPP-2.2.pkg/PPP-2.2.post_install: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/README.hppa: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/hppaSerialPatch.tar.gz: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/README.NeXT.MAB.Installation: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/Persistent_Connection: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/README: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/flow-control-hints: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/ip-down.example: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/ip-up.example: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/options: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/pdial: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/ppp_multiple_hosts.tar.gz: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/pppdown: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/pppkill.c: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/pppup.annex: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/pppup.direct: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/pppup.portmaster: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/pppup.remote: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/pppup.zyxel: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/rc.ppp: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/redial.sh: File name too long ./PPP-2.2-0.4.6-pkg: File name too long tar: can't create ./PPP-2.2-0.4.6-pkg/Examples/NXHosting_with_PPP: File name too long localhost:19# tar -xvf omnico* OmniComponents.pkg: File name too long tar: can't create OmniComponents.pkg/: File name too long OmniComponents.pkg: File name too long tar: can't create OmniComponents.pkg/OmniComponents.tar.Z: File name too long OmniComponents.pkg: File name too long tar: can't create OmniComponents.pkg/OmniComponents.info: File name too long OmniComponents.pkg: File name too long tar: can't create OmniComponents.pkg/OmniComponents.bom: File name too long OmniComponents.pkg: File name too long tar: can't create OmniComponents.pkg/OmniComponents.sizes: File name too long localhost:20# Nor can I find any help on how to set up ppp on nextanswers
From: Charles Swiger <chuck-nospam@blacksmith.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 10 Apr 1998 19:02:58 GMT Organization: BLaCKSMITH, Inc. Message-ID: <6glqd2$7j4$4@anvil.BLaCKSMITH.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> <1d7821x.1rdi5101enh5fhN@hobbit1.injep.fr> <6girj3$6j$2@anvil.BLaCKSMITH.com> <1d78ren.1dbj70n1j99pebN@hobbit1.injep.fr> xhsoft@injep.fr (Xavier Humbert) wrote: > Charles Swiger <chuck-nospam@blacksmith.com> wrote: >> the increase in costs represents roughly a day's pay > > That's not the point. 100% is 100%. Cans of soda used to cost 25 cents; they now cost 50 (or $1.00 or more on the highway). That's a 100% increase, also. People who are thirsty deal with it-- many of them don't even notice. A computer company needs to be aware of the costs of being a part of the industry. Runtime fees are very significant, since they alter the cost dynamics of every copy sold. One-time costs, such as the developer membership fees, do not change per-copy product economics. If a one-time $250 cost change makes a difference to your company as to which technologies to pursue, you were obviously considering a marginal opportunity. > BTW, I'm not a developer anymore. Now I'm employed as Network Engineer. > And it is MY decision for my company to buy $50 000 new Macintoshes, or > new Compaqs. Okay. Why does the $250 change in the accessibility of Rhapsody DR1 seeding make any difference whatsoever to your purchasing decisions? For that matter, since Rhapsody will run on your Intel-based Comaqs just fine, I don't understand what your point is. > And my boss may find that $500 is too expensive just to evaluate new > solutions. QED. Okay. See above re: "marginal opportunities". -Chuck Charles Swiger | chuck@BLaCKSMITH.com | standard disclaimer ---------------+----------------------+-------------------- "Microsoft: we make the easy almost impossible."
From: xhsof@injep.fr (Xavier Humbert) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 1998 14:42:23 +0200 Organization: INJEP Message-ID: <xhsof-0804981442230001@hobbit1.injep.fr> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gebdl$5d0$1@vixen.cso.uiuc.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit In article <6gebdl$5d0$1@vixen.cso.uiuc.edu>, gcheng@uni.uiuc.edu (Guanyao Cheng) wrote: > Can't you get the development CD's for $199?? Or do those not include > beta OS releases?? Nor does it include hadrdware discounts, neither MacOS SDKs :-( Xav -- Xavier HUMBERT Laboratoire Informatique INJEP Network Administrator humbert/AT/injep.fr labo-info/AT/injep.fr Institut National de la Jeunesse-Ministere Jeunesse et Sports
From: gupta@kcopsrm.tm.tta.moc-mirror after @ Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 10 Apr 1998 19:29:59 GMT Organization: IMEMYSELF Sender: Arun Gupta Message-ID: <6glrvn$f9j@newsb.netnews.att.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6girj3$6j$2@anvil.BLaCKSMITH.com> <1d78ren.1dbj70n1j99pebN@hobbit1.injep.fr> <6glqd2$7j4$4@anvil.blacksmith.com> Originator: gupta@tlctest >> And my boss may find that $500 is too expensive just to evaluate new >> solutions. QED. People are able to "sneak" Linux into the office though it is not an "approved" product, and have it on departmental servers, etc., because it is extremely inexpensive. Likewise, I imagine many people have the ability to do some discretionary spending without having to get approval from higher-ups. Getting early beta-software isn't all that important in this case. The ability of someone to effectively advocate Apple's products will be hurt if Apple's products are priced above certain magic points. In this regard, I don't think prospective developers should worry too much about developer subscription fees; the prices on the final products is what is going to really make a difference. [There is a developer down the aisle here, with a black Next cube gathering dust, but still on his desk, and with a Snail advertisement poster on his wall... I hope Rhapsody CR1 is released soon.] -arun gupta
From: xhsof@injep.fr (Xavier Humbert) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 1998 14:48:10 +0200 Organization: INJEP Message-ID: <xhsof-0804981448100001@hobbit1.injep.fr> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit In article <trumbull-0704981900290001@net44-223.student.yale.edu>, trumbull@cs.yale.edu (Ben Trumbull) wrote: > I'm not about to stand up and say these changes are a great idea on > Apple's part, but you do realize, you can buy *just* the tech mailing > (including develop) for $199 ? And then ? These are available for free (except Phone taxes) on Dev's Web Let's point to a REALLY annoying thing for non-US developers : we have theoretically no right to buy US system !!! Not to mention Rhapsody seeds. If developers have to pay 500 bucks to have Rhapsody, do you think the will develop for it ? Xav -- Xavier HUMBERT Laboratoire Informatique INJEP Network Administrator humbert/AT/injep.fr labo-info/AT/injep.fr Institut National de la Jeunesse-Ministere Jeunesse et Sports
From: rmcassid@uci.edu Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Fri, 10 Apr 1998 12:39:15 -0700 Organization: University of California, Irvine Message-ID: <rmcassid-1004981239150001@dante.eng.uci.edu> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <ericb-0904981701440001@132.236.171.104> In article <ericb-0904981701440001@132.236.171.104>, ericb@pobox.com (Eric Bennett) wrote: >Have any of you Mac users out there ever heard of a fellow by the name of >Peter Lewis? He started writing Mac apps in his spare time while he was >still employed at a university (do I sense a pattern here?). [snip] >Plenty of commercial or shareware apps had humble freeware beginnings. >Other good software has always been freeware. Why charge these people >$500 to get their hands on the latest system software? Do you want them >all to tell users, "Sorry, my program doesn't work with OS 8.2, because >Apple prices the developer versions out of my price range. I can't afford >to test my freeware with 8.2 until *after* it comes out, and then start >working on any necessary fixes." But this is exactly where your argument falls apart because you probably don't have to pay *anything* if you work for a univeristy. 80/20 rule time, boys and girls: It would appear that a considerable amount of freeware and shareware comes out of the .edu channel and the .edu channel doesn't have to pay. A considerable amount of the commercial software comes out of the traditional developer channel and they do have to pay. But guess what, they are commercial so $500 is no big whoop. If it is, then you are probably in need of a reality check. Another big chunk of freeware comes from commercial developers that have already paid their money, so that isn't a problem either. I don't see the problem. -Bob Cassidy
From: don_arb@wolfenet.com (Don Arbow) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Fri, 10 Apr 1998 12:07:17 -0700 Organization: EveryDay Objects, Inc. Message-ID: <don_arb-1004981207170001@sea-ts1-p29.wolfenet.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <352AAF57.25EB@skdjsk.no> <6geche$8lu$1@news2.apple.com> <philipm-0804982051170001@pm1-23.eecs.umich.edu> <6gj0or$oop$12@ironhorse.plsys.co.uk> <trumbull-0904981554150001@net44-223.student.yale.edu> In article <trumbull-0904981554150001@net44-223.student.yale.edu>, trumbull@cs.yale.edu (Ben Trumbull) wrote: : In article <6gj0or$oop$12@ironhorse.plsys.co.uk>, mmalcolm crawford : <malcolm@plsys.co.uk> wrote: : : > > You don't think Apple loses on hardware sold at developer prices? : > > : > Compared to hardware sold at retail prices, yes. : : that's pure fiction. Retailers take a huge chunk out of Apple's net on : those sales. And the hardware discounts were never very impressive, : imho. Think about. Does Apple make more money if Apple sells X a Mac or : if Cyberian Outpost sells X a Mac ? : Well, if the "hardware discounts were never very impressive", then why are all these people whining about losing access? Remember, also that Apple needed people to administer the hardware program, take orders, get the boxes shipped, etc. I'm sure that cut into any profits that Apple may have made on the boxes. Don -- Don Arbow, Partner, CTO EveryDay Objects, Inc. don_arb@wolfenet.com <-- remove underscore to reply http://www.edo-inc.com
Message-ID: <352E7B1E.CF65DC27@mci.com> From: David Hinz <David.Hinz@mci.com> Organization: MCI MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Can't find framework Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Fri, 10 Apr 1998 20:03:35 GMT NNTP-Posting-Date: Fri, 10 Apr 1998 16:03:35 EST Bing, I created a new framework for some file utilities. I built it, put it in /LocalLibrary/Frameworks on mocha. I then built a test program and in the compile/link flags used -F/LocalLibrary/Frameworks in FRAMEWORK_PATHS and added the -framework FileUtils to the list of frameworks. But when I try to run the program I get errors like: ld.so.1: ./DelimitedFile: fatal: /FileUtils.framework/Versions/A/FileUtils: can't open file: errno=2 ld.so.1: ./DelimitedFile: fatal: /FileUtils.framework/Versions/A/FileUtils: can't open file: errno=2 I've tried setting my LD_LIBRARY_PATH (didn't help). Is there something else I need to do to get my program to see the framework? dave. -- ===================================================== = David Hinz MCI Telecommunications = = Internet and New Media Development = = Email: David.Hinz@MCI.com Phone: (303) 390-6108 = = Vnet: 636-6108 Fax: (303) 390-6365 = = Pager: 1-888-900-5732 (Interactive 2-way) = =====================================================
From: j-jahnke@uchicago.edu (Jerome Jahnke) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Tue, 07 Apr 1998 16:15:45 -0500 Organization: University of Chicago -- BSDIS/CRT Message-ID: <j-jahnke-0704981616320001@192.168.1.3> References: <01bc445f$744397e0$1bf0bfa8@davidsul> In article <01bc445f$744397e0$1bf0bfa8@davidsul>, "macghod" <macghod@concentric.net> wrote: > WTF is up with the apple developer program? I see a big announcement, and > all it is is: > 1st level: online stuff, just as it is now, ie NO CHANGE > 2nd level: instead of costing $250 a year its now $500 a year??!?!?? > > As far as I can tell it adds nothing and is a 2x increase!! Why to support > development apple :) Actually it adds nothing, removes other stuff and increases by a factor of two... Apple did this a few years ago, gouge developers. It didn't work then. I don't see why they think it will work now. Jer,
From: ericb@pobox.com (Eric Bennett) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Tue, 07 Apr 1998 17:49:52 -0400 Organization: Cornell University Sender: emb22@cornell.edu (Verified) Message-ID: <ericb-0704981749520001@132.236.171.104> References: <01bc445f$744397e0$1bf0bfa8@davidsul> In article <01bc445f$744397e0$1bf0bfa8@davidsul>, "macghod" <macghod@concentric.net> wrote: > WTF is up with the apple developer program? I see a big announcement, and > all it is is: > 1st level: online stuff, just as it is now, ie NO CHANGE > 2nd level: instead of costing $250 a year its now $500 a year??!?!?? > > As far as I can tell it adds nothing and is a 2x increase!! Why to support > development apple :) Agreed. I might have considered getting Rhapsody under the previous program. I might be able to scrounge up $250. $500 is out of my price range. I guess they didn't solicit input from graduate students when forming their new policies. I'm hardly an important developer for Apple, and am not likely to become one, but other people like me might (Aaron Giles comes to mind as a recent example). :-| -- Eric Bennet (www.pobox.com/~ericb) Cornell University Biochemistry Department
From: trumbull@cs.yale.edu (Ben Trumbull) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Tue, 07 Apr 1998 19:00:29 -0400 Organization: Yale University Message-ID: <trumbull-0704981900290001@net44-223.student.yale.edu> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> In article <ericb-0704981749520001@132.236.171.104>, ericb@pobox.com (Eric Bennett) wrote: > In article <01bc445f$744397e0$1bf0bfa8@davidsul>, "macghod" > <macghod@concentric.net> wrote: > > > WTF is up with the apple developer program? I see a big announcement, and > > all it is is: > > 1st level: online stuff, just as it is now, ie NO CHANGE > > 2nd level: instead of costing $250 a year its now $500 a year??!?!?? > > > > As far as I can tell it adds nothing and is a 2x increase!! Why to support > > development apple :) > > Agreed. I might have considered getting Rhapsody under the previous > program. I might be able to scrounge up $250. $500 is out of my price > range. I'm not about to stand up and say these changes are a great idea on Apple's part, but you do realize, you can buy *just* the tech mailing (including develop) for $199 ? beats a sharp stick in the eye, sorta. terminally curious, Ben ___________________________________________________________________ Benjamin Trumbull trumbull@cs.yale.edu Yale University You can't be in hell; you can still read your e-mail
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Tue, 07 Apr 98 23:28:10 GMT Organization: QMS Message-ID: <6gecpj$f9p$2@news01.deltanet.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gebdl$5d0$1@vixen.cso.uiuc.edu> In article <6gebdl$5d0$1@vixen.cso.uiuc.edu>, gcheng@uni.uiuc.edu (Guanyao Cheng) wrote: >Can't you get the development CD's for $199?? Or do those not include >beta OS releases?? The developer CDs for $199 is not a bad deal, but they do not include the seeding support. That is, presumably, what you are paying $500 for. Scott Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Tue, 07 Apr 98 23:27:14 GMT Organization: QMS Message-ID: <6gecns$f9p$1@news01.deltanet.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> In article <01bc445f$744397e0$1bf0bfa8@davidsul>, "macghod" <macghod@concentric.net> wrote: >WTF is up with the apple developer program? I see a big announcement, and >all it is is: >1st level: online stuff, just as it is now, ie NO CHANGE >2nd level: instead of costing $250 a year its now $500 a year??!?!?? 'Tis actually worse than that. Presently, the $250 Associates get hardware discounts. As of the end of this month, they do not, which means that the contract terms were changed unilaterally in mid year. I wrote to the ADR folks, and will give them a couple of days, before I start haranguing Investor Relations. My argument is that this will substantially discourage small developers, as well as cost in the defense against the various class action suits I hope to start/join/see, and should thus have been mentioned in various SEC filings. BTW, had they announced this for future renewals, I would not have a real complaint. I would have thought it unwise, and said so, but I would not have been enraged. Changing the hardware purchase program requirements in mid year has annoyed me to the point of leaving the program. Choosing Tax Month as the month in which all such purchases need to be made is also not a great move. The seeding program, imho, served a valuable purpose, but Apple, I think, does not want any leaks of its seeds, and so it is trying to reduce the number of developers who might have access. >As far as I can tell it adds nothing and is a 2x increase!! Why to support >development apple :) Make that "takes away something of real value, and costs twice as much." It does seem to be designed to eradicate small developers. But then, that is not too dissimilar to how Next behaved. Scott Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Tue, 07 Apr 98 23:30:53 GMT Organization: QMS Message-ID: <6gecun$f9p$3@news01.deltanet.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> In article <trumbull-0704981900290001@net44-223.student.yale.edu>, trumbull@cs.yale.edu (Ben Trumbull) wrote: >In article <ericb-0704981749520001@132.236.171.104>, ericb@pobox.com (Eric >Bennett) wrote: > >> In article <01bc445f$744397e0$1bf0bfa8@davidsul>, "macghod" >> <macghod@concentric.net> wrote: >> >> > WTF is up with the apple developer program? I see a big announcement, and >> > all it is is: >> > 1st level: online stuff, just as it is now, ie NO CHANGE >> > 2nd level: instead of costing $250 a year its now $500 a year??!?!?? >> Agreed. I might have considered getting Rhapsody under the previous >> program. I might be able to scrounge up $250. $500 is out of my price >> range. > >I'm not about to stand up and say these changes are a great idea on >Apple's part, but you do realize, you can buy *just* the tech mailing >(including develop) for $199 ? It does not, though, include the seeding support. Further, they decided to change the hardware policy in mid year, and require any and all purchases to be made in the next two weeks, even if you just renewed. While it is clearly legal according to the agreement, it seems like they should send out a "would you like to bow out" letter to those for whom the hardware purchase agreement was an important part of the cost justification. But then, Apple is not really good at the cost justification game anyway. Consider what a lousy job they did selling corporate America on lower support costs. Scott Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
From: jason@jhste1.dyn.ml.org (Jason S.) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 10 Apr 1998 20:17:54 GMT Organization: I don't think so Message-ID: <slrn6isvlb.1kg.jason@jhste1.dyn.ml.org> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> <1d7821x.1rdi5101enh5fhN@hobbit1.injep.fr> <6girj3$6j$2@anvil.BLaCKSMITH.com> <1d78ren.1dbj70n1j99pebN@hobbit1.injep.fr> <6glqd2$7j4$4@anvil.BLaCKSMITH.com> Charles Swiger wrote: >If a one-time $250 cost change makes a difference to your company as to which >technologies to pursue, you were obviously considering a marginal >opportunity. Not very entrepreneurial thinking, Chuck. Not all "software companies" are well-financed (take, for example, the bright teenager with a good idea and an old Performa who produces some quality shareware). You seem to be supporting a policy that will shrink the number of developers (to the benefit of the established companies, of course). Free software is the glue that holds a platform together. Consider the following Mac programs and the impact they had in their time: FreePPP/MacPPP Anarchie/Fetch Mosaic JPEGView Sparkle NCSA Telnet TexEdit/BBEdit Lite Newswatcher and derivatives Disinfectant UUundo/Stuffit Expander/Binhex/Macbinary These titles provided basic functionality that the OS lacked for years. Now, I'd be willing to wager that many of these titles were started for the same reason that most GPL'd software is started - someone had a need that the OS/commercial offerings didn't really satisfy. If Apple restricts the number of potential developers for Rhapsody, it is less likely that stuff like this (no, not the same programs or clones, but programs to handle the needs of the future) will be developed. This doesn't bode well for the desirability of the platform to end users. On a slight different note, contrast Linux, which ships with free dev tools. I run Linux/PowerPC, which has a tiny userbase, but I have access to virtually all free UNIX software, due to the fact that I have gcc (and egcs) and all the necessary libraries (well, no Motif, but Lesstif works pretty well). I'd think that the Linux on PowerPC user experience is much better than the Rhapsody-without-dev-tools user experience will be precisely for this reason. -- If FreeBSD actually did that, I would concede that FreeBSD was any more "correct" than Linux is, but not even the FreeBSD people can justify that kind of performance loss. -- Linus Torvalds on comp.unix.advocacy
From: ylaporte@acm.org (Yan Laporte) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 12:48:37 GMT Organization: Université de Moncton Message-ID: <ylaporte-0904980952150001@139.103.176.79> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggiu7$auu$1@vixen.cso.uiuc.edu> <352c857a.2678525846@news.u-psud.fr> <joe.ragosta-0904980645050001@elk33.dol.net> > > For a student working at home, CDs are far better with his <1 to 5 > > kb/sec IP connection to US (at ~ $1/hour com cost + fixed provider > > cost). > > That is probably true. But wouldn't most students have access to their > college's facilities where they can download what they need by T1 and put > it onto a zip drive or burn a CD to take home? > Yep! that's what we do here... Or it is even better when you live on the campus (like me) and you get your 10-BaseT ethernet connection to the LAN and the Internet. At the very least you can find a firend who is in my position and get everything on a hard drive... EnCoder
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 8 Apr 1998 20:31:20 GMT Organization: Digital Fix Development Message-ID: <6ggmqo$jlv$1@news.digifix.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <j-jahnke-0704981616320001@192.168.1.3> <support-0704982011260001@207-172-164-184.s57.as4.loc.erols.com> <6ggahf$rts$3@news01.deltanet.com> In-Reply-To: <6ggahf$rts$3@news01.deltanet.com> On 04/08/98, Scott Ellsworth wrote: >I like the idea. Who do we complain to, and how? Ric Ford said on >Macintouch that the feedback was uniformly negative. And you're surprised by this? Look at MacNN, there is both positive and negative feedback there. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: don_arb@wolfenet.com (Don Arbow) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Fri, 10 Apr 1998 12:49:33 -0700 Organization: EveryDay Objects, Inc. Message-ID: <don_arb-1004981249330001@sea-ts1-p29.wolfenet.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <joe.ragosta-0904980642200001@elk33.dol.net> <6gitpj$csn$1@interport.net> <joe.ragosta-0904980134360001@wil68.dol.net> <Er6tuI.Mu8@AWT.NL> <joe.ragosta-1004980621290001@elk81.dol.net> In article <joe.ragosta-1004980621290001@elk81.dol.net>, joe.ragosta@dol.net (Joe Ragosta) wrote: : In article <Er6tuI.Mu8@AWT.NL>, G.C.Th.Wierda@AWT.nl (Gerben Wierda) wrote: : : > : > SDK's and Mac API development don't interest me. I want to know what will : > happen to the cost of developing for Rhapsody. : > : : How about zero dollars? Download MkLinux and GNUStep. Or for $150, you can buy Tenon's CodeBuilder. Still less than the $500 seeding cost. Don -- Don Arbow, Partner, CTO EveryDay Objects, Inc. don_arb@wolfenet.com <-- remove underscore to reply http://www.edo-inc.com
From: Jeff <me@home.sleeping> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 08 Apr 1998 11:40:09 -0500 Organization: a real BIG one Message-ID: <352BA869.4901@home.sleeping> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6ge6hh$9mv@netaxs.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Matthew T. Russotto wrote: - -In article <01bc445f$744397e0$1bf0bfa8@davidsul>, -macghod <macghod@concentric.net> wrote: -}WTF is up with the apple developer program? I see a big announcement, and -}all it is is: -}1st level: online stuff, just as it is now, ie NO CHANGE -}2nd level: instead of costing $250 a year its now $500 a year??!?!?? -} -}As far as I can tell it adds nothing and is a 2x increase!! Why to support -}development apple :) - - Yeah, sucks, doesn't it? All I really want is the CDs, and they keep - upping the price and adding things I don't need. - Hmmm, sounds familiar, Maybe we should call it Apple$oft?
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 10 Apr 1998 21:30:51 GMT Organization: Digital Fix Development Message-ID: <6gm32b$21o$1@news.digifix.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <352AAF57.25EB@skdjsk.no> <6geche$8lu$1@news2.apple.com> <philipm-0804982051170001@pm1-23.eecs.umich.edu> <6gj0or$oop$12@ironhorse.plsys.co.uk> <trumbull-0904981554150001@net44-223.student.yale.edu> <don_arb-1004981207170001@sea-ts1-p29.wolfenet.com> <trumbull-1004981620410001@net44-223.student.yale.edu> In-Reply-To: <trumbull-1004981620410001@net44-223.student.yale.edu> On 04/10/98, Ben Trumbull wrote: >In article <don_arb-1004981207170001@sea-ts1-p29.wolfenet.com>, >don_arb@wolfenet.com (Don Arbow) wrote: > >> Well, if the "hardware discounts were never very impressive", then why are >> all these people whining about losing access? > >*I* never whined about it. I think a lot of it has to do with the >WAY in which it was done. Apple made these people FEEL like they >were getting shafted. You have to admit the way in which it was >done .. "here, the prices are doubled, and btw, we've cut your >access to the hardware discounts" lacked any sense of finesse or >rational sense of handling customers. Even someone from Apple >admitted the notice was "badly worded". If Apple had said to every developer "We're thinking about making these changes.." The reaction would have been the same. I'd be very surprised if the ADR folks didn't talk to some developers before they made the changes. I'd be very surprised if those customers weren't split, but that a good number of them understood the rational behind the changes and accepted them. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: mcgredo@otter.mbay.net (Donald R. McGregor) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 05:41:31 GMT Organization: The Vast Right Wing Conspiracy Message-ID: <6ghn2b$noq$1@usenet11.supernews.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <352AAF57.25EB@skdjsk.no> <6geche$8lu$1@news2.apple.com> <philipm-0804982051170001@pm1-23.eecs.umich.edu> In article <philipm-0804982051170001@pm1-23.eecs.umich.edu>, Philip Machanick <philipm@NO-SPAMeecs.umich.edu> wrote: >>Actually, there were rumors that there would be a very low price >>educational plan. You might want to wait for the educational pricing >>before worrying too much. > >I don't much care about that... the problem is small developers who will >feel discouraged when trying to find and entry point. If YB is to get a >good start, it needs some killer Mac-only (or at least YB-only) apps and >it's hard to see big companies betting on a new API. It depends on how they sell Rhapsody and the development tools. If they sell it for $100, IB and ObjC and Java included, I wouldn't give two hoots about the developer program. I don't plan on buying any apple hardware and not being able to get the discount wouldn't kill me if I were; I don't need tech support (or can puzzle most of it out from the net); and I don't care much about the CD du jour that clutters up my mailbox. If I were a corporation, $500 wouldn't kill me. If I'm an after-hours hacker, all I want is the damn Rhapsody CD. -- Don McGregor | "Traditionally, conservatism is rightly suspicious of mcgredo@mbay.net | thinking, because thinking on the whole leads to wrong | conclusions, unless you think very, very hard and get | back to the point you started at." --R. Scruton.
From: j-jahnke@uchicago.edu (Jerome Jahnke) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 02:14:31 -0500 Organization: University of Chicago -- BSDIS/CRT Message-ID: <j-jahnke-0904980214320001@192.168.1.3> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> <j-jahnke-0804982141280001@192.168.1.3> <6ghg9f$o3$1@news.digifix.com> In article <6ghg9f$o3$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: > On 04/08/98, Jerome Jahnke wrote: > <snip> > >Now this is a bunch of bull. Many of us were Associates for one > >reason and one reason only, access to cheaper equipment. Apple > >does not lose money by selling me my one machine a year at a low > >price. > > Perhaps. But they do still provide you with all the other > stuff, the CD's, the mailings etc.. that does cost Apple money. It don't cost 500 bucks... Jer,
Date: Thu, 09 Apr 1998 01:35:33 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-0904980135340001@wil68.dol.net> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggrql$oop$1@ironhorse.plsys.co.uk> <6gisof$bht$1@interport.net> In article <6gisof$bht$1@interport.net>, float@interport.net wrote: > Have you ever heard of "mindshare"? Sure, it's a buzzword, but > dereference it and you find something important. > > If I had free or cheap access to Rhapsody technology, I would put it on a > spare PC. If I like Rhapsody, my next machine might be a Mac; the other > option is Sun's Darwin (the $3000 Ultra 5). Absolutely true. I wish Apple could afford to give a copy of Rhapsody to everyone who asks. It won't happen. > > I'm not about to write any killer apps for anybody's operating system, but > I am an effective advocate for any technology I like and find useful. > I've never been part of the "Apple camp" in the past, since I'm not fond > of MacOS. Apple has always relied to some extent on > enthusiasts/advocates/evangelists, and they could be getting unixers on > board right now if they cared to. > > mmalcolm crawford (malcolm@plsys.co.uk) wrote: > > : Why does anyone *need* to be part of the program if they're not developing > : professional or semi-professional apps. Sure, access to the latest and > : greatest betas of the s/w is fun, but not essential if you're hacking at home > : for the joy of it. Apple is not a charity -- and I don't want to have to pay > : more for my Mac to subsidise someone else's hobby. > > : mmalc. > -- Regards, Joe Ragosta http://www.dol.net/~Ragosta/complmac.htm
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 21:23:21 GMT Organization: Digital Fix Development Message-ID: <6gje89$q9i$1@news.digifix.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <trumbull-0904980039090001@net44-223.student.yale.edu> <6gio5m$oop$11@ironhorse.plsys.co.uk> <ericb-0904981637520001@132.236.171.104> In-Reply-To: <ericb-0904981637520001@132.236.171.104> On 04/09/98, Eric Bennett wrote: >In article <6gio5m$oop$11@ironhorse.plsys.co.uk>, mmalcolm crawford ><malcolm@plsys.co.uk> wrote: > >> No, I don't. I'm quite sure there are a number of hobbyists developing >> worthwhile s/w which I might find useful at some point, however on balance >> reports suggest that the previous system was badly abused and on average >> Apple will have lost revenue through people signing up as developers simply >> to get hardware at reduced prices. > >That doesn't give Apple the right to alter contracts. If they want to >decrease the discount, or raise the fees for renewals or new accounts, >fine. But they're cutting off *existing* subscribers. > I guess you've not READ the contract then huh? >> If you're a real hobbyist, do you *need* access to the bleeding edge s/w? >> If not, then you don't need to belong to the program, so it won't cost you >> anything. > >It would be advantageous for me to start learning Rhapsody programming >ASAP. I can't do that very well with MacOS. > You're at an educational institution. Perhaps Cornell is involved in the sponsorship program. Even if they aren't Apple has already promised an attractive development program for Students/Faculty. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 21:21:50 GMT Organization: Digital Fix Development Message-ID: <6gje5e$q8g$1@news.digifix.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> <MPG.f960500e190219d989895@news.supernews.com> <joe.ragosta-0804982250580001@wil62.dol.net> <ericb-0904981631110001@132.236.171.104> In-Reply-To: <ericb-0904981631110001@132.236.171.104> On 04/09/98, Eric Bennett wrote: >In article <joe.ragosta-0804982250580001@wil62.dol.net>, >joe.ragosta@dol.net (Joe Ragosta) wrote: > > >> As I see it (and these comments reinforce it), Donald is very much acting >> in the Lawson mold--Apple is guilty until proven innocent. He's measuring >> everything by _HIS_ standards and finds Apple at fault every time they >> don't meet his expectations. > >I don't see any developers defending Apple's decisions. The best >"support" I've seen is some Mac news sites parroting Apple's press >release. > Really? I know that Macintouch has received several positive notes... guess Ric doesn't want to print them. There has been a number of posts to Rhapsody-talk, and the Lyris rhapsody list by developers who don't understand what all the whining is about. There is an article by a YellowBox developer on Stepwise that is understanding of the need for the changes. Guess you should look with better eyes huh? -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 98 22:05:17 GMT Organization: QMS Message-ID: <6gjgm5$bfj$1@news01.deltanet.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <trumbull-0904980039090001@net44-223.student.yale.edu> <6gio5m$oop$11@ironhorse.plsys.co.uk> <ericb-0904981637520001@132.236.171.104> In article <ericb-0904981637520001@132.236.171.104>, ericb@pobox.com (Eric Bennett) wrote: >In article <6gio5m$oop$11@ironhorse.plsys.co.uk>, mmalcolm crawford ><malcolm@plsys.co.uk> wrote: > >> No, I don't. I'm quite sure there are a number of hobbyists developing >> worthwhile s/w which I might find useful at some point, however on balance >> reports suggest that the previous system was badly abused and on average >> Apple will have lost revenue through people signing up as developers simply >> to get hardware at reduced prices. > >That doesn't give Apple the right to alter contracts. If they want to >decrease the discount, or raise the fees for renewals or new accounts, >fine. But they're cutting off *existing* subscribers. Like I have said iin other posts, had they continued the former rights and privledges until the end of my developer contract, I would merely think this an unwise change, but they have changed it unilaterally, for the worse, and without sufficient (i.e. before the last renewal) warning, imho. To be fair, in the contract I signed, they reserved the right to change any and all terms and conditions unilaterally, and noted that neither full nor partial refunds are possible. The credit card company I talked with earlier said they wanted a copy of the agreement faxed to them, so that they could see whether it was, in the opinion of thier lawyers, enforcable, but it sounded (if my description was accurate, fair, and complete) like it was not. .. >I'm still not sure about whether the $199 subscription to mailings >includes things like Rhapsody DRs or not. Some people say yes, other info >seems to say "no." If it does, it's a reasonable solution for people like >me, and my only remaining beef would be that Apple has changed the terms >of active subscriptions (still, that's a serious complaint). From what I have been reading, if they class a Rhapsody build as part of the SDK or part of a released version of the software, then it will likely come on the CDs. If they classify it as a seeding release, then it will not. So far, they have never made a pre release OS part of the developer mailings, and they only rarely have released other software in other than final form. OpenDoc comes to mind as a counter example. So far, they have classified Rhapsody, and all other pre release operating systems, as seeding releases, but, except for a brief time, you could not get the developer mailing without having implicit permission to join the seeding program. If they are wise, they will provide some kind of schedule wherein developers can get Rhapsody builds prior to the general public expecting thier software to work on them, but there is little information as to whether Apple intends to change policy in that way. Scott Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
From: pointal@lure.u-psud.fr (Laurent pointal) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 08:28:44 GMT Organization: CNRS - LURE Message-ID: <352c857a.2678525846@news.u-psud.fr> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggiu7$auu$1@vixen.cso.uiuc.edu> On 8 Apr 1998 19:24:55 GMT, gcheng@uni.uiuc.edu (Guanyao Cheng) wrote: >: Although I disapprove of the move by Apple, I really question why a >: HS/college student would have to join the developer programs to write >: software. After all, they could just download MPW or buy CodeWarrior at a >: huge discount, and learn from the vast amount of online resources at >: devworld.apple.com. Do you know the bandwitdh of US servers out of the US? Do you know the cost of communications out of the US? For a student working at home, CDs are far better with his <1 to 5 kb/sec IP connection to US (at ~ $1/hour com cost + fixed provider cost). A+ Laurent. --- Laurent POINTAL - CNRS/LURE - Service Informatique Experiences Tel/fax: 01 64 46 82 80 / 01 64 46 41 48 email : pointal@lure.u-psud.fr ou lpointal@planete.net ouebe : http://www.planete.net/~lpointal
From: Joshua Winsor <joshua@alexa.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 15:19:04 -0700 Organization: Internet Archive Message-ID: <352D4957.84AE1890@alexa.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Scott Anguish wrote: > Now, as far as Rhapsody is concerned it is leaving a gap, but > only because of the time-frame.... However if you join Select you're > still covered. > Download GnuStep and MKLinux, all free, MKLinux paid for by Apple, and you are a Rhapsody developer.
From: j-jahnke@uchicago.edu (Jerome Jahnke) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 20:01:04 -0500 Organization: University of Chicago -- BSDIS/CRT Message-ID: <j-jahnke-0904982001050001@192.168.1.4> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> Xcanpos: shelf.1/199804240601!0013914045 In article <joe.ragosta-0804982109560001@elk62.dol.net>, joe.ragosta@dol.net (Joe Ragosta) wrote: > Anyone notice that people like Mr. brown here, who would apparently never > go near a Mac from reading his posts, are the ones complaining while > developers like mmalc are supporting the change? Given Mr Brown was once a developer of note for the Mac also speaks volumes, apparently he has moved on becuase Apple fucked with him one to many times... Jer,
From: "Brad Hutchings" <brad@hutchings-software.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 98 15:24:34 -0700 Organization: EarthLink Network, Inc. Message-ID: <B15298C1-10AAF2@207.217.155.85> References: <joe.ragosta-0904981624100001@wil59.dol.net> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit nntp://news.earthlink.net/comp.sys.mac.oop.powerplant, nntp://news.earthlink.net/comp.sys.mac.programmer.codewarrior, nntp://news.earthlink.net/comp.sys.next.advocacy, nntp://news.earthlink.net/comp.sys.next.programmer >>Me > Joe Ragosta >> A spanking over a fairly reasonable tactical move sends the wrong message. >> This move would not have been met with intense cynicism if developers >> thought Apple had an overall clue and direction and appreciation of how >> developers support Apple. It starts from the top, and unfortunately, is >> beyond the control of ADR. > >So, in other words, you're just assuming that, because it's Apple, they >did the wrong thing. No Joe. Read the rest of my comments. I'm observing that this isn't flying with developers because of the wider context of a company that doesn't get "developer relations". Plug yourself into one of many private Mac developers chat lists if you doubt that *real* developers are dumbfounded over this move. Take a look at MacCentral today. The reaction isn't just from hobbyists/GNUists/students. And it's not all about not being able to afford the new programs. It's about whether those programs are worth the money and meet developer needs. You wanna know something? I strongly believe that if Apple (the company, not ADR) treated its current developer crop as partners and respected their business concernd (Newton and OD/CD and Rhapsody direction are prime examples), and if they'd had the PR for the programs worded a little better, Apple would be hailed today as the most pro-developer company in history after announcing the new programs. I do not deal in simplistic FUD nor irrational support, Joe, as you suggest. Brad <mailto: "Brad Hutchings" brad@hutchings-software.com> <http://www.hutchings-software.com> Got OpenDoc? Got Cyberdog? Then try Rapid-I Button. Tried it? Then buy it! Point your Cyberdog to the Hutchings Software web site.
From: gmgraves@slip.net (George Graves) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 15:54:18 -0700 Organization: Graves Associates Message-ID: <gmgraves-0904981554180001@sf-usr1-47-175.dialup.slip.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <joe.ragosta-0904980642200001@elk33.dol.net> In article <joe.ragosta-0904980642200001@elk33.dol.net>, joe.ragosta@dol.net (Joe Ragosta) wrote: > In article <6gh8v1$oop$6@ironhorse.plsys.co.uk>, mmalcolm crawford > <malcolm@plsys.co.uk> wrote: > > > In <01bc4540$d4a15e60$3af0bfa8@davidsul> "macghod" wrote: > > > > > > > The lowest paid developer status is $500 instead of $250. If you really > > > > need to get the beta OS versions, your cost went up by $250. But everyone > > > > else seems to be better off. > > > > > > More Ragosta FUD. "Everyone else seems to be better off". Who is better > > > off? > > > > > Apple's customers who no longer have to subsidise hobbyist developers. > > > > Best wishes, > > > > mmalc. > > Or hobbyist pseudo developers--people who never wrote an app or never > planned to, but paid the $250 for access to hardware and software > discounts. \ And what's wrong with paying $250, signing an NDA, and getting the OS software early? I'm interested. I want to see what's going on. I don't mind paying the money for that privilege. I'm looking mightily forward to the DR2 Rhapsody release, and plan to use it, and upgrade to subesquent one's as they come available. I will also file bug reports (least I can do), as, I'm sure most people in this category would. That is a service Apple used to pay people good money for (product assurance) but, they fired all of those people, now its just developers. If Apple wanted to insure that the people who joined the Developer program were REALLY developers, then, instead of hiking up the price (which hurts everybody) they should screen applicants. Ask 'em what kind of software they're developing, and ask for a time-table and to be included in the loop. Those unable to comply, would then be dropped. But Apple doesn't even ask you what you want to be a developer for. They just want the money and the NDA. If Apple were smart (there's the big one again. Snort! If Apple were smart indeed, who am I kidding?) Developer programs would be FREE as would admission to WWDC. They need more developers, not fewer developers. BTW, anyone who wants the mailings can purchase a year's subscription for $199. I'm not sure whether this includes seeding or not (based, of course, on signing an NDA), but this is actually $50 cheaper than the current Associate fee. George Graves > > -- > Regards, > > Joe Ragosta > See the Complete Macintosh Advocacy Page > http://www.dol.net/~Ragosta/complmac.htm
From: don.brown@cesoft.com (Donald Brown) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 9 Apr 1998 17:50:55 -0500 Organization: CE Software Message-ID: <MPG.f970cce11b8c92e989898@news.supernews.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> <MPG.f960500e190219d989895@news.supernews.com> <joe.ragosta-0804982250580001@wil62.dol.net> In article <joe.ragosta-0804982250580001@wil62.dol.net>, joe.ragosta@dol.net says... > As I see it (and these comments reinforce it), Donald is very much acting > in the Lawson mold--Apple is guilty until proven innocent. He's measuring > everything by _HIS_ standards and finds Apple at fault every time they > don't meet his expectations. > One more thing...I can't imagine why you find it odd that I'm measuring Apple by my standards. I make comments from the point of view of small commercial developer. I assume that's understood. I am not would not never would try to enforce my standards on Apple, just giving commentary. Donald
From: don.brown@cesoft.com (Donald Brown) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 9 Apr 1998 17:48:13 -0500 Organization: CE Software Message-ID: <MPG.f970c25119c4d82989897@news.supernews.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> <MPG.f960500e190219d989895@news.supernews.com> <joe.ragosta-0804982250580001@wil62.dol.net> In article <joe.ragosta-0804982250580001@wil62.dol.net>, joe.ragosta@dol.net says... > As I see it (and these comments reinforce it), Donald is very much acting > in the Lawson mold--Apple is guilty until proven innocent. He's measuring > everything by _HIS_ standards and finds Apple at fault every time they > don't meet his expectations. > Nope. Apple is guilty when it performs a guilty action. Yes, I'm measuring Apple by my standards and making my comments appropriately-- when they meet my standards they get kudo, when they don't they get slams. I've been rolled over and left high and dry by Apple so many times that I no longer have faith in their wisdom or benificence. They often do great things and I say so when they do so, and they often screw up and I say so when they do so. After you've been bitten a few times, you no longer have to assume the dog showing his teeth is just smiling. Donald
Date: Thu, 09 Apr 1998 19:37:37 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-0904981937380001@elk64.dol.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <352D35D6.E368D52C@am.not.here> In article <352D35D6.E368D52C@am.not.here>, i@am.not.here.either wrote: > My two bits... > > It seems that all the hub-bub is w.r.t. shareware and freeware developers. > Now, I could be wrong, but it seems to me that what Apple is saying is...If you > want to develop commercial software, choose either our select or premier > programs. Otherwise, use an educational institution to gain access to our > products. And if you want to develop shareware, download the FREE SDKs. In case someone missed it, that's FREE. -- Regards, Joe Ragosta See the Complete Macintosh Advocacy Page http://www.dol.net/~Ragosta/complmac.htm
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 98 23:50:55 GMT Organization: QMS Message-ID: <6gjmsc$fg4$2@news01.deltanet.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> <MPG.f960500e190219d989895@news.supernews.com> <joe.ragosta-0804982250580001@wil62.dol.net> <MPG.f96ab2832ee6e1a9896d8@news.itg.ti.com> <MPG.f97115b6eb9d82698989a@news.supernews.com> In article <MPG.f97115b6eb9d82698989a@news.supernews.com>, don.brown@cesoft.com (Donald Brown) wrote: >In article <MPG.f96ab2832ee6e1a9896d8@news.itg.ti.com>, jmcn@msg.ti.com >says... >> In article Joe Ragosta, joe.ragosta@dol.net says... .. >> > But I don't see anything here that changes the essence. .. >> People who do rely on Apple get the shaft, a lot. >Before we all go jumping into bed together.... > >Apple can burn you, and yet Apple can bless you. Apple is neither >perfect or is it the spawn of Satan. If I didn't care any more, if I >didn't find this fascinating and love thoe machines, I wouldn't be here. This is true for me as well. As a response to a crisis of faith, I spent a bit of time browsing the Windows APIs again. While I do this for work purposes, I rarely just wander through them to get a feel for them. In my mind, they are still quite inferior. (There is, of course, room for disagreement on this. Other people have other needs.) Coupled with the crashes I have had of late on my P333 running a pretty stock install of Win95, I am still confident that the Mac has a place. If I did not find it sufficiently superior to the competition that I get angry with Microsoft, then I would not be that interested. (MS has so much cash that I have little sympathy for any crashes in thier OS. They can afford to drown it in resources, so a BSOD is a failure of will and intention.) Unfortunately, Apple does regularly do foolish things, and when they do, they pick doozies. Again, this is my own opinion as a developer of some substantial programs on both platforms, but I hope it is representative. It seems fair to me at this time to state that the present actions by Apple do not lead to a feeling of stability. Sometimes, they do good things, such as when they chose to release MPW for free. The good ones make me think they have a better chance, and are more reliable. I then increase the evangalism efforts where I can. Sometimes, they do bad things, like the unilateral change in the developer programs which reduced the current benefits to recent subscribers. These add to feelings of uncertainty. Not uncertainty in a global "Apple is doomed" sense, but the small scale uncertainty of "Apple says Rhapsody is important, and that the MacOS will be around for quite some time. Apple also said that the Newton was spun back in to preserve it, that Open Doc was an important technology, and that Copland would ship in 95. How much can I trust the statement du jour?" We know that they are going to have to make some unexpected changes, and that we as developers are going to have to react to them. That is why we get the big bucks. They should be very cautious, though, because if they do make a substantial change which harms us, without enough good changes to compensate, we will react rationally. The reintroduction of the developer mailings is a very good idea, but given the recent changes, I am not willing to take on faith that they are going to include Rhapsody release versions in the mailings. I would have said that was very likely until recently, but the changes have made me doubt their constancy. Now, they need to reassure people about thier exact plans about who gets Rhapsody, how, and when. If they are not ready to do this, then why did they pick now to make these changes? Why not wait until they had all the answers. Surely, the hostile response was not a surprise? >My objections to some actions of Apple do not mean I hate Apple. I'm >still hopining somehow things will be turned around. Exactly. I am using every tool within my grasp to convince them to change thier actions precisely because the current ones seem ill considered, in poor faith, and rapacious. Better packaging, and efforts to assure me of where I, and other developers, will fit in the future would go a long way towards making me feel they are not completely clueless. >A world without the Mac would be a much sadder world indeed. True enough. Scott Ellsworth Apple Associate Developer Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
Date: Thu, 09 Apr 1998 19:56:40 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple's Announcement Misleading. WAS :Apple developer program Message-ID: <joe.ragosta-0904981956410001@elk90.dol.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> <MPG.f960500e190219d989895@news.supernews.com> <joe.ragosta-0804982250580001@wil62.dol.net> <gmgraves-0904980952020001@sf-pm5-27-91.dialup.slip.net> In article <gmgraves-0904980952020001@sf-pm5-27-91.dialup.slip.net>, gmgraves@slip.net (George Graves) wrote: > I spoke, This AM, with someone from Apple Developer Support. > He said that the E-mail that all developer's received about > Apple's changes in their Dev program (and the source of Ric Ford's > Info), is misleading. The actual impetus, apparently, for these > changes, is to IMPROVE accessability to these resources, not to > make them more expensive. > > While notoutlining what the actual changes are, he said that at the low > end of the tier, Apple was trying to divorce the dev materials from the > support, in order to make those materials cheaper. He said that it is now > possible for small developers to get the seeding materials and the monthly > mailings at much less cost than the $250 minimum under the old program. > Unfortunately, the Apple E-mail sent to all of its developers does not put it > this way. According to my source within Apple Developer support. the E-mail > was "poorly written......." We shall see. I guess the question to answer to determine the accuracy of this is: "Was it previously possible to download the SDK's for free?" If it was not, the new program _would_ be an improvement on the low end since you could get started without spending a penny. If it was always possible to download the SDK's for free, it's hard to understand how the new program could be an improvement. -- Regards, Joe Ragosta See the Complete Macintosh Advocacy Page http://www.dol.net/~Ragosta/complmac.htm
From: cb@bartlet.df.lth.se (Christian Brunschen) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 11:50:09 +0200 Organization: The Computer Society at Lund Message-ID: <6gi5kh$ors$1@bartlet.df.lth.se> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> NNTP-Posting-User: cb In article <01bc4540$d4a15e60$3af0bfa8@davidsul>, macghod <macghod@concentric.net> wrote: > [ in an earlier article, Joe Ragost wrote: ] >> The monthly mailing is available for $50 less than the old Associate >membership. >> >> The lowest paid developer status is $500 instead of $250. If you really >> need to get the beta OS versions, your cost went up by $250. But everyone >> else seems to be better off. > >More Ragosta FUD. It would appear you are unfamiliar with the meaning of the acronym 'FUD'. 'FUD' stands for 'Fear, Uncertainty and Doubt'. The moniker 'FUD' is usually used when someone makes claims _against_ a certain technology or company or <whatever>, based not upon fact, but innuendo and opinion - and intended to create 'fear, uncertainty and doubt' about said company or technology. Ie, like when people claim that 'Apple is going to die soon', or 'there is no software for MacOS'. Joe Ragosta, above, does not spread 'FUD' in any sense of the expression. Rather, Joe indicates his own opinion that the new developer programs are not as bad as some people want to make them look; this, he is trying to reassure people, calm them, and _not_ drive them off a particular direction; he is doing the exact opposite of spreading FUD. > "Everyone else seems to be better off". Note that he says 'seems to be' - it seems, to him, that everyone else is better off. Not FUD - opinion, and clearly marked as such. > Who is better >off? Their is NOTHING that is now cheaper now than before the >announcement. >Either things are the same price as before with no added >benefits, or they have increased and in some cases have increased with less >benefits. This is, of course, your interpretation of things. // Christian Brunschen
From: rcl@ultranet.com (Rich Long) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Fri, 10 Apr 1998 22:12:36 -0400 Organization: UltraNet Communications, Inc. http://www.ultranet.com/ Message-ID: <rcl-1004982212360001@d2.dial-3.nsh.nh.ultra.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <352D35D6.E368D52C@am.not.here> <joe.ragosta-0904981937380001@elk64.dol.net> In article <joe.ragosta-0904981937380001@elk64.dol.net>, joe.ragosta@dol.net (Joe Ragosta) wrote: > And if you want to develop shareware, download the FREE SDKs. In case > someone missed it, that's FREE. As are the MPW development tools. Rich -- From the Apple PowerBook of... Richard Long -- rcl@ultranet.com My Links and Software Page: http://www.ultranet.com/~rcl/
From: "NeXT Newbie" <macghod@concentric.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple's Announcement Misleading. WAS :Apple developer program Date: 10 Apr 1998 00:34:48 GMT Message-ID: <01bc464a$a1d13d60$2af0bfa8@davidsul> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> <MPG.f960500e190219d989895@news.supernews.com> <joe.ragosta-0804982250580001@wil62.dol.net> <gmgraves-0904980952020001@sf-pm5-27-91.dialup.slip.net> <joe.ragosta-0904981956410001@elk90.dol.net> Joe Ragosta <joe.ragosta@dol.net> wrote in article <joe.ragosta-0904981956410001@elk90.dol.net>... > In article <gmgraves-0904980952020001@sf-pm5-27-91.dialup.slip.net>, > gmgraves@slip.net (George Graves) wrote: > > While notoutlining what the actual changes are, he said that at the low > > end of the tier, Apple was trying to divorce the dev materials from the > > support, in order to make those materials cheaper. He said that it is now > > possible for small developers to get the seeding materials and the monthly > > mailings at much less cost than the $250 minimum under the old program. > > Unfortunately, the Apple E-mail sent to all of its developers does not put it > > this way. According to my source within Apple Developer support. the E-mail > > was "poorly written......." We shall see. A couple of things... 1) the sdk's have been downloadable for a while now. 2) *IF* developers can get the SEEDING MATERIAL and the monthly mailings at much less cost than $250, they would not be complaining so vocally. How does one get the seedings (which rhapsody is a part of) for under $250? I think it is a factual error that you can get the seeding material for $250 or less than $250
From: j-jahnke@uchicago.edu (Jerome Jahnke) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 19:53:42 -0500 Organization: University of Chicago -- BSDIS/CRT Message-ID: <j-jahnke-0904981953430001@192.168.1.4> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <trumbull-0904980039090001@net44-223.student.yale.edu> <6gio5m$oop$11@ironhorse.plsys.co.uk> In article <6gio5m$oop$11@ironhorse.plsys.co.uk>, mmalcolm crawford <malcolm@plsys.co.uk> wrote: > In <trumbull-0904980039090001@net44-223.student.yale.edu> Ben Trumbull wrote: > > In article <6gh8v1$oop$6@ironhorse.plsys.co.uk>, mmalcolm crawford > > <malcolm@plsys.co.uk> wrote: > > > > > Apple's customers who no longer have to subsidise hobbyist developers. > > > > Apple never "subsidized" hobbyists. You don't think Apple made money off > > of every single on of them ? > > > No, I don't. I'm quite sure there are a number of hobbyists developing > worthwhile s/w which I might find useful at some point, however on balance > reports suggest that the previous system was badly abused and on average > Apple will have lost revenue through people signing up as developers simply > to get hardware at reduced prices. In order for this to be true you have to prove that Apple loses money on hardware sold to people in the Dev Program. And the fact of the matter is that they don't lose money. They make as much money off of that as they would if the sold the machine to a retailer (or proably even a discounter.) > I also hope that the additional fees will go toward giving a better service > to those who choose to stay in one of the developer programs. Lets look at the past shall we?? Every time Apple has done this (which has been three or four times since *I* have been in the program,) prices have gone up and support stayed the same or in the last case got a lot worse. Someone jumped on the Dev Support people's collective head, telling them they were losing too much money and to fix it. They scratched their heads in this was the best they could think of. What ticks me off is NOT the increase in prices, but the fact that my level of support, which I paid for on good faith, has been significantly reduced less than six months into the cycle. Given the letter states that they had not accepted any new members in the past six months, means that a bait and switch was pulled, I signed up four months ago and was not told this would happen. Jer,
From: rriley@yahoo.com (Rex Riley) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 10 Apr 1998 00:52:38 GMT Organization: @Home Network Message-ID: <6gjqgm$fd5$1@ha2.rdc1.sdca.home.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> <MPG.f960500e190219d989895@news.supernews.com> <joe.ragosta-0804982250580001@wil62.dol.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: joe.ragosta@dol.net In <joe.ragosta-0804982250580001@wil62.dol.net> Joe Ragosta wrote: > In article <MPG.f960500e190219d989895@news.supernews.com>, > don.brown@cesoft.com (Donald Brown) wrote: [snip confession of a true developer ] > > Well, Microsoft never claimed to care about us, but they haven't broken > > any promises or revised our contracts while they're still in effect > > lately. There will be some number of developers giving up on Apple and > > switching to Windows because of this action. I don't know how many, but > > in Apple's weak position, any is too many. > > I apologized to Donald in e-mail for the strength of my comments. > > But I don't see anything here that changes the essence. > > As I see it (and these comments reinforce it), Donald is very much acting > in the Lawson mold--Apple is guilty until proven innocent. He's measuring > everything by _HIS_ standards and finds Apple at fault every time they > don't meet his expectations. > > Who made you "Keeper of the Truth" ? What BULLSHIT... Sheesh, this reads like a Psych consult. The guy just shared his motivation behind his actions. It's His Story. If you don't like the reading of History, you needn't resort to accusing him of "acting" in whatever mold De Jour. Apple wrote the book, he's just reading it back to you from his living it. If you've apologized to the fellow in some prior encounter, you certainly owe him another. -r Rex Riley
From: droege@informatik.uni-koblenz.Drop_This.de (Detlev Droege) Newsgroups: comp.sys.next.programmer Subject: Re: TIFF format used in OpenStep / Rhapsody Date: 8 Apr 1998 09:51:56 GMT Organization: University Koblenz / CC Distribution: world Message-ID: <6gfhbs$hq2$1@newshost> References: <marke-ya02408000R0704981357380001@17.128.100.122> In article <marke-ya02408000R0704981357380001@17.128.100.122> marke@apple.com (Mark Eaton) writes: > In article <6gcpdp$nr$1@lazar.select-tech.si>, > andreja.vidmar@select-tech.si wrote: > > If you have access to a machine running NEXTSTEP/OPENSTEP/Rhapsody, you > > can use tiffutil command line utility to find out if JPEG compression is > > used in the TIFF files and decompress them. > The TIFF files used on OpenStep are readable by QuickTime 3, on all of > QuickTime's supported platforms. You can open the tiffs with the > PictureViewer application included with QT3 and export it to some other > format if you need to. Sure ? I saw QT3.0 (f7) fail on LZW compressed TIFFs as well as on PNG images. Tried to generate an animation from 100 (povray rendered) images. QT3 complained about "broken" images. All the images displayed just fine on OpenStep. The only way to succeed was to present _uncompressed_ TIFF to QT3. Unfortunately, QT3 didn't even bother to tell _which_ of the 100 images caused the trouble (it seemed to be different frames in TIFF-lzw and PNG). Detlev -- Detlev Droege, Universitaet Koblenz, FB Informatik | Fon:+49 261 9119-421 Rheinau 1, D-56075 Koblenz, Germany | Fax:+49 261 9119-497 NeXT/MIME/Email: droege@informatik.uni-koblenz.Drop_This.de Drop the "Drop_This." part in my Email-address to reply.
Date: Thu, 09 Apr 1998 06:40:29 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-0904980640290001@elk33.dol.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <352AAF57.25EB@skdjsk.no> In article <352AAF57.25EB@skdjsk.no>, sdd wrote: > I want a student developers program... > > 150-200 $/year for cd's... > > Anonymous.. Apple has publicly stated that there will be an attractive educational program--it's just not ready yet. -- Regards, Joe Ragosta See the Complete Macintosh Advocacy Page http://www.dol.net/~Ragosta/complmac.htm
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 18:25:00 -0700 Organization: Primenet Services for the Internet Message-ID: <B152C422-16106C@206.165.43.139> References: <joe.ragosta-0904981531590001@wil38.dol.net> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Joe Ragosta <joe.ragosta@dol.net> said: > Just like the last 20 rants that you went on were going to destroy Apple. > Wasn't abandoning QD GX supposed to destroy Apple? Yep. Abandoning superior technology for nothing in the pipeline (no upgrades to newtonOS, GX, CD, OpenDoc, etc will ever be done, meaning that any developer or user who bought into those is now orphaned) is always destructive. ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
Date: Thu, 09 Apr 1998 06:42:20 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-0904980642200001@elk33.dol.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> In article <6gh8v1$oop$6@ironhorse.plsys.co.uk>, mmalcolm crawford <malcolm@plsys.co.uk> wrote: > In <01bc4540$d4a15e60$3af0bfa8@davidsul> "macghod" wrote: > > > > > The lowest paid developer status is $500 instead of $250. If you really > > > need to get the beta OS versions, your cost went up by $250. But everyone > > > else seems to be better off. > > > > More Ragosta FUD. "Everyone else seems to be better off". Who is better > > off? > > > Apple's customers who no longer have to subsidise hobbyist developers. > > Best wishes, > > mmalc. Or hobbyist pseudo developers--people who never wrote an app or never planned to, but paid the $250 for access to hardware and software discounts. -- Regards, Joe Ragosta See the Complete Macintosh Advocacy Page http://www.dol.net/~Ragosta/complmac.htm
Date: Thu, 09 Apr 1998 06:43:36 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-0904980643360001@elk33.dol.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ylaporte-0804982005030001@139.103.176.79> <6gh9cm$oop$7@ironhorse.plsys.co.uk> In article <6gh9cm$oop$7@ironhorse.plsys.co.uk>, mmalcolm crawford <malcolm@plsys.co.uk> wrote: > In <ylaporte-0804982005030001@139.103.176.79> Yan Laporte wrote: > > What we need from Apple is an idea of when and what we will have to pay > > for the system in higher education institutions. And don't get me the > > "you got to be carefull, it is not meant to be used by normal users for > > now(...)" speech. We're a Computer Science department, we just crave for > > that stuff and we have the brain to handle it. So what are the plans for > > Rhapsody in Higher education institutions?? Anyone? I need more than > > dumb hearsays from MOSR... There was no such as a Developper rebate or any > > program available for us three weeks ago, is there now? > > > I'm not sure what I'm allowed to say here, but I've certainly seen some > public announcements by Apple (the problem being the only place I know > there's a copy is private), and have an idea from speaking with some people > on the inside that Apple is very much aware of all the points you've made > here, and will be adressing them. > > cf also the following. > > Best wishes, > > mmalc. > > > Subject: Re: Apple Developer Connection Mailing > Date: Wed, 8 Apr 1998 09:37:36 -0700 > From: Jordan Dea-Mattson <jordan@apple.com> > To: "Rhapsody Discussion List" <rhapsody@clio.lyris.net> > > [...] > > In addition, we are working to roll out programs for the academic world - > and have said so for some time. They weren't ready and this time and I > can't comment on them, except to say that we are aware of the need to > reach the academic world. Jordan later commented on RDL that Apple is trying to make the educational program very attractive. -- Regards, Joe Ragosta See the Complete Macintosh Advocacy Page http://www.dol.net/~Ragosta/complmac.htm
Date: Thu, 09 Apr 1998 06:45:04 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-0904980645050001@elk33.dol.net> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggiu7$auu$1@vixen.cso.uiuc.edu> <352c857a.2678525846@news.u-psud.fr> In article <352c857a.2678525846@news.u-psud.fr>, pointal@lure.u-psud.fr wrote: > On 8 Apr 1998 19:24:55 GMT, gcheng@uni.uiuc.edu (Guanyao Cheng) wrote: > > >: Although I disapprove of the move by Apple, I really question why a > >: HS/college student would have to join the developer programs to write > >: software. After all, they could just download MPW or buy CodeWarrior at a > >: huge discount, and learn from the vast amount of online resources at > >: devworld.apple.com. > > Do you know the bandwitdh of US servers out of the US? Do you know the > cost of communications out of the US? > > For a student working at home, CDs are far better with his <1 to 5 > kb/sec IP connection to US (at ~ $1/hour com cost + fixed provider > cost). That is probably true. But wouldn't most students have access to their college's facilities where they can download what they need by T1 and put it onto a zip drive or burn a CD to take home? -- Regards, Joe Ragosta See the Complete Macintosh Advocacy Page http://www.dol.net/~Ragosta/complmac.htm
From: Hubert HOLIN <hh@ArtQuest.fr> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 12:51:29 +0200 Organization: ArtQuest Message-ID: <352CA831.D4878417@ArtQuest.fr> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <6ggvub$oop$5@ironhorse.plsys.co.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit mmalcolm crawford wrote: [SNIP] > If you're a professional developer, or intent on producing shareware that's > of genuine utility (i,e, good enough that people will pay for it), $500 > should not be a serious impediment, else your business plans are out of > kilter. [SNIP] By that measure, freeware is junk, right? I guess TeX, GNU and their ilk are unknown to you... Hubert Holin holin@mathp7.jussieu.fr
From: j-jahnke@uchicago.edu (Jerome Jahnke) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Tue, 07 Apr 1998 23:42:46 -0500 Organization: University of Chicago -- BSDIS/CRT Message-ID: <j-jahnke-0704982342480001@192.168.1.3> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> In article <trumbull-0704981900290001@net44-223.student.yale.edu>, trumbull@cs.yale.edu (Ben Trumbull) wrote: > I'm not about to stand up and say these changes are a great idea on > Apple's part, but you do realize, you can buy *just* the tech mailing > (including develop) for $199 ? Which is what we used to do, but then realized you could not get seeded unless you were an associate. Jer,
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 10:43:16 GMT Organization: P & L Systems Message-ID: <6gi8o4$oop$9@ironhorse.plsys.co.uk> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <6ggs48$oop$2@ironhorse.plsys.co.uk> <6ghf6g$e1u$1@geraldo.cc.utexas.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: kdb@pegasus.ece.utexas.edu In <6ghf6g$e1u$1@geraldo.cc.utexas.edu> Kurt D. Bollacker wrote: > : If anyone is wanting to develop for Rhapsody professionally and doesn't have > : $500, they don't have a business plan. If you're intending to write > : shareware, then $500 is a reasonable investment if you believe you have > : something useful to contribute. If you're writing Rhapsody apps for a hobby, > : why should I subsidise your playtime? > > Because I might develop some useful freeware. With your attitude there'd > never be any freeware at all. Without freeware, perhaps NeXT would have > been later with NeXTSTEP (no GNU tools jumpstart) and Rhapsody today would > not exist. Not all software should be free, but some of it always should. > How is not being a member of the program going to stop people from producing freeware if they want? Why does anyone producing freeware need advance access to beta software? > : In the bad old days when NeXT's developement tools alone cost seven times > : that amount there was no shortage of freeware or shareware. > > Bzzzt. Wrong, but thank you for playing. > In what way wrong? The tools cost $3500. There was no shortage of freeware or shareware. > If you bought NeXTSTEP as > an academic machine or later as an academic bundle CD-ROM, the development > tools were free or *very* cheap. Remember, NeXT was orignially academic > only. Lots of cool freeware came out very quickly. > As an academic user of NeXTstep from 1989 to 1996, I am very well aware of this, thank you. As a NeXT Campus Consultant I even gave freely of my own time to spread the Good Word both in my University and, on several occasions, at others. I was the only Campus Consultant in the UK who was successful. But, to return to your "point": how much did a NeXT Cube cost when it came out? $7000+. And yet there was a lot of freeware, you say. So now that a G3 AOI costs, what, $1500, there isn't going to be any? Furthermore, you obviously missed the part about Apple not yet having announced the edu program. I can't believe that the edu program is going to be more than the commercial program. Apple *has* announced that Rhaposdy will be available at a very good price to academics, so I suspect that that will be an even better deal than the old $299 package from NeXT. Given that the cost of machines has reduced, all in all it looks to me like it's going to be cheaper for academics now than it ever has been. mmalc.
Date: Wed, 08 Apr 1998 13:12:16 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-0804981312160001@wil32.dol.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gecns$f9p$1@news01.deltanet.com> <1d767ho.1tixfqy1dqgm03N@hobbit1.injep.fr> <6gg6af$fit$1@anvil.BLaCKSMITH.com> <6ggabn$rts$2@news01.deltanet.com> In article <6ggabn$rts$2@news01.deltanet.com>, scott@eviews.com (Scott Ellsworth) wrote: > In article <6gg6af$fit$1@anvil.BLaCKSMITH.com>, cswiger@BLaCKSMITH.com wrote: > >xhsoft@injep.fr (Xavier Humbert) wrote: > >> Scott Ellsworth <scott@eviews.com> wrote: > >>> The seeding program, imho, served a valuable purpose, but > >>> Apple, I think, does not want any leaks of its seeds, and so > >>> it is trying to reduce the number of developers who might > >>> have access. > >> > >> What is obviously stupid : WarEz sites have System Seeds BEFORE the > >> official FTP :-( > >> > >> Xav, looking for a warez URL > > > >I don't have much of an opinion about Apple's change to their Dev. program > >pricing, one way or the other. I do know, however, that people who condone > >software piracy have no moral grounds whatsoever to complain about what Apple > >does. > > > The fellow's point was, I hope, that limiting distribution in this way > seems unlikely to prevent the seeds from getting to the warez sites, > because the current arcana one has to go through to get seeds > (and it is VERY arcane) has not succeeded. Maybe that was the point, but it's not how I read it, either. One would presume that "serious" developers (i.e. those who rely on Apple for their livelihood) are less likely to post a build to the Warez site. To the extent that raising the price will not affect a large company as much as someone doing Mac OS development as a hobby, raising the price _may_ reduce the postings to Warez sites. Whether that justifies the action is, of course, open for debate. -- Regards, Joe Ragosta http://www.dol.net/~Ragosta/complmac.htm
Date: Thu, 09 Apr 1998 01:34:35 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-0904980134360001@wil68.dol.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <joe.ragosta-0904980642200001@elk33.dol.net> <6gitpj$csn$1@interport.net> In article <6gitpj$csn$1@interport.net>, float@interport.net wrote: > Joe Ragosta (joe.ragosta@dol.net) wrote: > > : Or hobbyist pseudo developers--people who never wrote an app or never > : planned to, but paid the $250 for access to hardware and software > : discounts. > > Apple should be ACTIVELY ENCOURAGING these people. Some of them might > grow up to be software developers . . . or management . . . or sysadmins . > . . or end users . . . or, or, or . . . Sure. They should encourage everyone. But how do you allocate limited resources? Apple has decided that spending the money on TV ads or magazine ads is a better place to put scarce resources than subsidizing hobbyists. And in case you've missed it, it is now possible for hobbyists to get SDK's free. How much cheaper do they need to be? -- Regards, Joe Ragosta http://www.dol.net/~Ragosta/complmac.htm
From: ericb@pobox.com (Eric Bennett) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 16:37:51 -0400 Organization: Cornell University Sender: emb22@cornell.edu (Verified) Message-ID: <ericb-0904981637520001@132.236.171.104> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <trumbull-0904980039090001@net44-223.student.yale.edu> <6gio5m$oop$11@ironhorse.plsys.co.uk> In article <6gio5m$oop$11@ironhorse.plsys.co.uk>, mmalcolm crawford <malcolm@plsys.co.uk> wrote: > No, I don't. I'm quite sure there are a number of hobbyists developing > worthwhile s/w which I might find useful at some point, however on balance > reports suggest that the previous system was badly abused and on average > Apple will have lost revenue through people signing up as developers simply > to get hardware at reduced prices. That doesn't give Apple the right to alter contracts. If they want to decrease the discount, or raise the fees for renewals or new accounts, fine. But they're cutting off *existing* subscribers. > If you're a real hobbyist, do you *need* access to the bleeding edge s/w? > If not, then you don't need to belong to the program, so it won't cost you > anything. It would be advantageous for me to start learning Rhapsody programming ASAP. I can't do that very well with MacOS. > If you think you do need access to bleeding edge s/w, then that's a fairly > serious hobby, and I think it's reasonable for people to invest in it. $250 I could have invested. Not $500. I'm still not sure about whether the $199 subscription to mailings includes things like Rhapsody DRs or not. Some people say yes, other info seems to say "no." If it does, it's a reasonable solution for people like me, and my only remaining beef would be that Apple has changed the terms of active subscriptions (still, that's a serious complaint). -- Eric Bennett (www.pobox.com/~ericb) Cornell University Biochemistry Department
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Fri, 10 Apr 98 16:33:54 GMT Organization: QMS Message-ID: <6glhks$ros$1@news01.deltanet.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <trumbull-0904980039090001@net44-223.student.yale.edu> <6gio5m$oop$11@ironhorse.plsys.co.uk> <j-jahnke-0904981953430001@192.168.1.4> In article <j-jahnke-0904981953430001@192.168.1.4>, j-jahnke@uchicago.edu (Jerome Jahnke) wrote: >What ticks me off is NOT the increase in prices, but the fact that my >level of support, which I paid for on good faith, has been significantly >reduced less than six months into the cycle. Given the letter states that >they had not accepted any new members in the past six months, means that a >bait and switch was pulled, I signed up four months ago and was not told >this would happen. Interestingly, I sent in my form on Feb. 11 of 1998, my renewal date is Feb. 28, 1999, and the first and only CD I have recieved is the April, 1998 CD. As with you, what has angered me is the unilateral reduction in benefits, in my case, after only a month and change in the program. I disagree somewhat with the new pricing, but the return of the CD mailing-only level is a good idea, so it is uncertain what the balance is. IMO, they need Rhapsody seeds in thge hands of developopers as soon as possible, because of the maginitude of the change in skills, BUT this is merely a market clearing price question, amenable to analysis after the fact. If they decide they have restricted it too much, they can always seed it to the people with the CD mailings.. Changing the benefits in a substantial way is, on the other hand, very much a bad faith act that lowers trust in the long term stability (defined as continuance in a predicatble manner) of Apple. Scott Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> <j-jahnke-0804982141280001@192.168.1.3> <6ghg9f$o3$1@news.digifix.com> <j-jahnke-0904980214320001@192.168.1.3> From: John Kheit <jkheit@xtdl.com> Organization: monoChrome, Inc., NJ, USA Message-ID: <352d0347.0@206.25.228.5> Date: 9 Apr 98 17:20:07 GMT j-jahnke@uchicago.edu (Jerome Jahnke) wrote: > In article <6ghg9f$o3$1@news.digifix.com>, sanguish@digifix.com > (Scott Anguish) wrote: > > On 04/08/98, Jerome Jahnke wrote: <snip> > > >Now this is a bunch of bull. Many of us were Associates for > > >one reason and one reason only, access to cheaper equipment. > > >Apple does not lose money by selling me my one machine a year > > >at a low price. > > > > Perhaps. But they do still provide you with all the other > > stuff, the CD's, the mailings etc.. that does cost Apple money. > It don't cost 500 bucks... I'd like to get this straight. Do the mailings for $200 include betas of Rhapsody or not? If so, then they seem worthwhile. If not, I, personally, don't care. -- Thanks, be well, take care, later, John Kheit; self expressed... __________________________________________________________________ monoChrome, Inc. ASCII, MIME, PGP, SUN, & NeXTmail OK NeXT/OPENSTEP Developer mailto:jkheit@xtdl.com Telepathy, It's coming... http://www.xtdl.com/~jkheit Franklin Pierce Law Center You're dangerous because you're honest
From: ericb@pobox.com (Eric Bennett) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 16:42:58 -0400 Organization: Cornell University Sender: emb22@cornell.edu (Verified) Message-ID: <ericb-0904981642580001@132.236.171.104> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <xhsof-0804981448100001@hobbit1.injep.fr> <6ggs48$oop$2@ironhorse.plsys.co.uk> In article <6ggs48$oop$2@ironhorse.plsys.co.uk>, mmalcolm crawford <malcolm@plsys.co.uk> wrote: > If anyone is wanting to develop for Rhapsody professionally and doesn't have > $500, they don't have a business plan. If you're intending to write > shareware, then $500 is a reasonable investment if you believe you have > something useful to contribute. If you're writing Rhapsody apps for a hobby, > why should I subsidise your playtime? Because my stuff has always been freeware. Because I'm a student, and Apple needs to offer me an educational discount (NOW, not a year from now--does Apple want to encourage students to uses its latest and greatest OS or not?). I don't worry at all about major developers. It's the hobbyist programmers like myself that I worry about. Most of the software I write is to solve problems for myself. Then I realize other people might be able to use them, so I release them. But I don't really have enough free time to guarantee good support for them, they're not particularly well polished, and that's why I keep them freeware rather than charging for them. If you don't mind losing people like me, fine, but I think Apple needs hobbyist developers. -- Eric Bennett (www.pobox.com/~ericb) Cornell University Biochemistry Department
From: ericb@pobox.com (Eric Bennett) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 16:49:37 -0400 Organization: Cornell University Sender: emb22@cornell.edu (Verified) Message-ID: <ericb-0904981649370001@132.236.171.104> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> <1d7821x.1rdi5101enh5fhN@hobbit1.injep.fr> <6girj3$6j$2@anvil.BLaCKSMITH.com> In article <6girj3$6j$2@anvil.BLaCKSMITH.com>, cswiger@BLaCKSMITH.com wrote: > Let me put it to you this way: the increase in costs represents roughly a > day's pay (before taxes, anyway :-) for the average programmer. BFD. The new cost is two weeks' pay (after withholding :-) out of my stipend. :-( -- Eric Bennett (www.pobox.com/~ericb) Cornell University Biochemistry Department
From: daj@nwu.edu Newsgroups: comp.sys.next.programmer,comp.lang-objective-c Subject: Problems with changing hostname,domainname in Objective-C app Date: 10 Apr 1998 17:23:53 GMT Organization: Northwestern University, Evanston, IL, US Distribution: world Message-ID: <6glkj9$r9s@news.acns.nwu.edu> Hello everyone, I'm trying to build an app to install, maintain DHCP and provide some IP tools for OpenStep/Rhapsody/NeXTStep user. I'm having a minor problem with privileges, I think. What I'm trying to achieve is to have the app figure out what IP was assigned, then lookup what it's assigned hostname,domainname are and allow the user to choose to change their hostname,domainname to match what was assigned to that port (it should work with en0 or ppp0 for that matter). The code below compiles and runs but I it doesn't change the hostname and it reports the error: sethostname: Not owner Apr 10 11:13:40 IPMonitor[791] task already launched I set the suid bit and I've even tried running it as root, so what gives and how do I fix it? - (void)changeHostname:(id)sender { NSTask *pipeTask = [[NSTask alloc] init]; NSPipe *newPipe = [NSPipe pipe]; NSFileHandle *readHandle = [newPipe fileHandleForReading]; NSString *hostname *domainname; NSArray *futureMachineName, *cmdArgs; NSString *assignedHostname=@"testing.domain.net"; // for testing only int i; futureMachineName = [assignedHostname componentsSeparatedByString:@"."]; hostname = [futureMachineName objectAtIndex:0]; futureMachineName = [assignedHostname componentsSeparatedByString:[hostname stringByAppendingString:@"."]]; domainname = [futureMachineName objectAtIndex:0]; cmdArgs = [NSArray arrayWithObjects:[NSString stringWithCString:hostname],nil]; [pipeTask setStandardOutput:newPipe]; // write handle is closed to this process [pipeTask setLaunchPath:[NSOpenStepRootDirectory() stringByAppendingPathComponent:[NSString stringWithCString:"/bin/hostname"]]]; [pipeTask setArguments:cmdArgs]; [pipeTask launch]; cmdArgs = [NSArray arrayWithObjects:[NSString stringWithCString:domainname],nil]; [pipeTask setStandardOutput:newPipe]; // write handle is closed to this process [pipeTask setLaunchPath:[NSOpenStepRootDirectory() stringByAppendingPathComponent:[NSString stringWithCString:"/bin/domainname"]]]; [pipeTask setArguments:cmdArgs]; [pipeTask launch]; } David A. Johnson email: daj@nwu.edu Manager Computer Facility and Engineering Phone: (312) 503-0408 Feinberg Research Institute, Fax: (312) 503-0137 Northwestern University HAM call: N9HAM
From: jmcn@msg.ti.com (Jason McNorton) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 9 Apr 1998 10:54:19 -0500 Organization: Texas Instruments Message-ID: <MPG.f96ab2832ee6e1a9896d8@news.itg.ti.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> <MPG.f960500e190219d989895@news.supernews.com> <joe.ragosta-0804982250580001@wil62.dol.net> In article Joe Ragosta, joe.ragosta@dol.net says... *snip* > I apologized to Donald in e-mail for the strength of my comments. > > But I don't see anything here that changes the essence. Face it Joe, Apple has burned lots of people and they resent Apple for it. And they are right. Not everyone can be as blindly fanatical as you are. I'm starting to see that maybe you don't really have much reliance or much invested in Apple. If you did, you wouldn't support them so freely. People who do rely on Apple get the shaft, a lot. -- A world without the Mac is a step closer to utopia.
From: Charles Swiger <chuck-nospam@blacksmith.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 15:55:43 GMT Organization: BLaCKSMITH, Inc. Message-ID: <6gir1v$6j$1@anvil.BLaCKSMITH.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> "macghod" <macghod@concentric.net> wrote: >> The monthly mailing is available for $50 less than the old Associate >> membership. >> >> The lowest paid developer status is $500 instead of $250. If you really >> need to get the beta OS versions, your cost went up by $250. But everyone >> else seems to be better off. > > More Ragosta FUD. "Everyone else seems to be better off". Who is better > off? Presumably Apple, else they wouldn't have made the change. The question is whether this is at the expense of their customers, as some have claimed. I have yet to be convinced as to why this would be the case. -Chuck Charles Swiger | chuck@BLaCKSMITH.com | standard disclaimer ---------------+----------------------+-------------------- "Microsoft: we make the easy almost impossible."
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Followup-To: comp.sys.next.advocacy,comp.sys.mac.advocacy Date: 10 Apr 1998 18:24:39 GMT Organization: P & L Systems Message-ID: <6glo57$oop$14@ironhorse.plsys.co.uk> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <6ggvub$oop$5@ironhorse.plsys.co.uk> <352CA831.D4878417@ArtQuest.fr> <9APR98.16105101@fredjr.pfc.mit.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: FRIDBERG@PSFC.MIT.EDU NOTE: FOLLOWUPS TO COMP.SYS.NEXT.ADVOCACY,COMP.SYS.MAC.ADVOCACY In <9APR98.16105101@fredjr.pfc.mit.edu> FRIDBERG@PSFC.MIT.EDU wrote: > ->mmalcolm crawford wrote: > -> > ->> If you're a professional developer, or intent on producing shareware that's > ->> of genuine utility (i,e, good enough that people will pay for it), $500 > ->> should not be a serious impediment, else your business plans are out of > ->> kilter. > -> > And let's not forget such things as NewsWatcher (which probably half of people > who read this NG are using, or NSCA Telnet (which I am using) or bunch of other > internet software which made Mac so attractive to use for Internet access > and allowed us to have pretty high market share on internet. Mister MMalcolm > Crawford prbably never heard of such programs as Disinfectant of Internet > Config either. Or maybe he just afraid that all those freeware applications > going to drive him out of business? > No, I'm not afraid these will drive us out of business at all. You're displaying an astonishing lack of intelligence (in the broadest sense of the word) by suggesting this. I welcome freeware applications, just like anyone else. For your information, we have actually distributed a couple of freeware applications for Rhapsody ourselves, and, for your benefit, sponsored the development of another (OpenUp -- cf http://www.stepwise.com/ ) I fail to understand your pride in NCSA Telnet or Disinfectant; telnet comes free with Rhapsody, and viruses are unknown on Unix. There are also a number of free Unix Newsreaders, and internet configuration on Rhapsody should be straightforward. What your rather vacuous message has not addressed is the question as to why anyone writing freeware would *need* all the benefits of the developer program. mmalc.
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Followup-To: comp.sys.mac.advocacy,comp.sys.next.advocacy Date: 10 Apr 1998 18:27:31 GMT Organization: P & L Systems Message-ID: <6gloaj$oop$15@ironhorse.plsys.co.uk> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <6ggs48$oop$2@ironhorse.plsys.co.uk> <6ghf6g$e1u$1@geraldo.cc.utexas.edu> <6gi8o4$oop$9@ironhorse.plsys.co.uk> <steve-0904981201500001@oranoco.discoverysoft.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: steve@discoverysoft.com NOTE: FOLLOWUPS TO COMP.SYS.MAC.ADVOCACY,COMP.SYS.NEXT.ADVOCACY In <steve-0904981201500001@oranoco.discoverysoft.com> Steven Fisher wrote: > I would think this extends to shareware as well. > > Would Default Folder have been able to support MacOS 8 as it shipped if > Jon Gotow hadn't had acess to betas of MacOS 8? > > Would Kaleidoscope have supported MacOS 8 and MacOS 8 interface elements > immediately if Greg Landweber hadn't had acess to betas of MacOS 8? > > Don't you feel better knowing that the freeware and shareware software you > use is being tested long before the next MacOS is actually released? > No. I'm used to using an operating system where stuff doesn't break simply because a new version is released. Best wishes, mmalc.
From: Charles Swiger <chuck-nospam@blacksmith.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 16:04:51 GMT Organization: BLaCKSMITH, Inc. Message-ID: <6girj3$6j$2@anvil.BLaCKSMITH.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> <1d7821x.1rdi5101enh5fhN@hobbit1.injep.fr> xhsoft@injep.fr (Xavier Humbert) wrote: > <rmcassid@uci.edu> wrote: >> and can quite likely afford a modest >> increase in costs. > > 100% increase is modest for you ? Let me put it to you this way: the increase in costs represents roughly a day's pay (before taxes, anyway :-) for the average programmer. BFD. -Chuck Charles Swiger | chuck@BLaCKSMITH.com | standard disclaimer ---------------+----------------------+-------------------- "Microsoft: we make the easy almost impossible."
From: "NeXT Newbie" <macghod@concentric.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 10 Apr 1998 18:52:40 GMT Message-ID: <01bc46e4$1815c920$24f0bfa8@davidsul> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104><joe.ragosta-0804980858150001@wil77.dol.net><mazulauf-0804981503390001@sneezy.met.utah.edu><joe.ragosta-0804981906460001@elk68.dol.net><01bc4540$d4a15e60$3af0bfa8@davidsul><6gh8v1$oop$6@ironhorse.plsys.co.uk> <SCOTT.98Apr9091856@slave.doubleu.com> > > More Ragosta FUD. "Everyone else seems to be better off". Who > > is better off? > > Apple's customers who no longer have to subsidise hobbyist > developers. > > Where's the incremental cost of making developer tools freely > available to all comers? The developer hardware discounts going away > isn't a problem - the problem is that the hardware cost isn't in line > with market costs. So far as tech support, that's generally sucked > for anyone who doesn't have a rep assigned to them anyhow. Lets be honest here. Macapp was made available AFTER it was put in maintanence mode. IE apple said "geez, it sure isnt worth it to sell this, lets cut our losses and stop actively putting out new releases for it". And whoever thinks the cost of the program doesnt cover apples costs is greatly naive. Plus, now Apple is using the microsoft developer pricing model as its own. Oops, but microsoft has %95 of the market and apple only has %3. Oops, Microsoft has a monopoly so they can do this. I guess Apple is now "just as good" as microsoft <big grin> Except they dont have a monopoly and are figting for their lifes, so they cant use monopolistic pricing like microsoft does. Clue phone for Mr Jobs, Mr Jobs please pick up the clue phone. Mr Jobs, Apple is not Microsoft, Microsoft can charge those prices because they are a monopoly, Apple is fighting for its life so it cant treat developers so shabbily
From: rdieter@math.unl.edu (Rex Dieter) Newsgroups: comp.sys.next.programmer Subject: Re: OS 4.2 and libg++ static-libs/dylibs Date: 10 Apr 1998 19:15:32 GMT Organization: University of Nebraska-Lincoln Message-ID: <6glr4k$qrg$1@unlnews.unl.edu> References: <6glq8u$qmn$1@unlnews.unl.edu> In article <6glq8u$qmn$1@unlnews.unl.edu> rdieter@math.unl.edu (Rex Dieter) writes: > I've ALMOST finished creating dynamic library version of libg++-2.7.2 for > Openstep for Mach (4.2) (and Rhapsody), but it doesn't quite work fully. Oops, sorry to follow-up my own post, but the relavent files for testing this out on your own are at: http://www.math.unl.edu/~rdieter/OpenStep/Developer/ libg++.2.7.2.3.m.README (see CAVEATS section) libg++-dev.2.7.2.3.m.NIS.b.tar.gz libg++-dylibs.2.7.2.3.m.NIS.b.tar.gz -- Rex A. Dieter rdieter@math.unl.edu (NeXT/MIME OK) Computer System Manager http://www.math.unl.edu/~rdieter/ Mathematics and Statistics University of Nebraska-Lincoln
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 10 Apr 1998 12:55:01 -0700 Organization: Primenet Services for the Internet Message-ID: <B153C824-106D1@206.165.43.115> References: <joe.ragosta-1004980619170001@elk81.dol.net> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Joe Ragosta <joe.ragosta@dol.net> said: > > Yep. Abandoning superior technology for nothing in the pipeline (no > > upgrades to newtonOS, GX, CD, OpenDoc, etc will ever be done, meaning > that > > any developer or user who bought into those is now orphaned) is always > > destructive. > > Right. I'm still chuckling over all your statements about GX's superiority > over DPS which have been consistently wrong. No, you're chuckling over DPS-lover's propoganda. Remember that the only feedback about DPS that you are getting has been from people that use DPS, not GX. Also, the people in charge of graphics at Apple *invented* DPS, not GX, so they have a certain level of bias, also. And the only difference between "hurt Apple" and "destroy Apple" is time-frame. "Hurt Apple" often enough and it is destroyed. ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: float@interport.net Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Followup-To: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Date: 9 Apr 1998 12:24:47 -0400 Organization: Interport Communications Corp. Message-ID: <6gisof$bht$1@interport.net> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggrql$oop$1@ironhorse.plsys.co.uk> Have you ever heard of "mindshare"? Sure, it's a buzzword, but dereference it and you find something important. If I had free or cheap access to Rhapsody technology, I would put it on a spare PC. If I like Rhapsody, my next machine might be a Mac; the other option is Sun's Darwin (the $3000 Ultra 5). I'm not about to write any killer apps for anybody's operating system, but I am an effective advocate for any technology I like and find useful. I've never been part of the "Apple camp" in the past, since I'm not fond of MacOS. Apple has always relied to some extent on enthusiasts/advocates/evangelists, and they could be getting unixers on board right now if they cared to. mmalcolm crawford (malcolm@plsys.co.uk) wrote: : Why does anyone *need* to be part of the program if they're not developing : professional or semi-professional apps. Sure, access to the latest and : greatest betas of the s/w is fun, but not essential if you're hacking at home : for the joy of it. Apple is not a charity -- and I don't want to have to pay : more for my Mac to subsidise someone else's hobby. : mmalc. -- Ben <Just Another System Administrator>
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 10 Apr 1998 13:00:01 -0700 Organization: Primenet Services for the Internet Message-ID: <B153C956-14E86@206.165.43.115> References: <joe.ragosta-1004980621290001@elk81.dol.net> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Joe Ragosta <joe.ragosta@dol.net> said: > > SDK's and Mac API development don't interest me. I want to know what > will > > happen to the cost of developing for Rhapsody. > > > > How about zero dollars? Download MkLinux and GNUStep. GNUStep doesn't exist as a useable product, letalone a standard anyone could develop for. And what does MKLinux have to do with Rhaposdy developement? The fact that they are both Unices? ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: trumbull@cs.yale.edu (Ben Trumbull) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Fri, 10 Apr 1998 16:20:41 -0400 Organization: Yale University Message-ID: <trumbull-1004981620410001@net44-223.student.yale.edu> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <352AAF57.25EB@skdjsk.no> <6geche$8lu$1@news2.apple.com> <philipm-0804982051170001@pm1-23.eecs.umich.edu> <6gj0or$oop$12@ironhorse.plsys.co.uk> <trumbull-0904981554150001@net44-223.student.yale.edu> <don_arb-1004981207170001@sea-ts1-p29.wolfenet.com> In article <don_arb-1004981207170001@sea-ts1-p29.wolfenet.com>, don_arb@wolfenet.com (Don Arbow) wrote: > Well, if the "hardware discounts were never very impressive", then why are > all these people whining about losing access? *I* never whined about it. I think a lot of it has to do with the WAY in which it was done. Apple made these people FEEL like they were getting shafted. You have to admit the way in which it was done .. "here, the prices are doubled, and btw, we've cut your access to the hardware discounts" lacked any sense of finesse or rational sense of handling customers. Even someone from Apple admitted the notice was "badly worded". > Remember, also that Apple needed people to administer the hardware > program, take orders, get the boxes shipped, etc. I'm sure that cut into > any profits that Apple may have made on the boxes. That's entirely Apple mismangement, then. Everyone who sells boxes has people do this. Damn, Compaq and other clones makers make far smaller margins, yet they still sell boxes and (occassionally) make profits when they aren't busy spending all their money acquiring other companies. Why can't Apple ? This is a pretty pathetic excuse. Bottom line is Apple should still have been netting money on the hardware discounts. And how can it be bad for Apple if the program nets $10 or some small but largely irrelevant profit ? That's better than losing money, and it's helping developers write Macintosh programs .. good for Apple. The essence of my thread was that all these people who are saying that freeware/shareware developers got nothing less than they deserved and made completely unsubstantiated claims that Apple was subsidizing these developers are full of !@#$%. There's never been a hobbyist developer welfare supported entirely by Apple as some people have proposed. Increasing the costs of the dev program and restricting access to dev hardware discounts is not going to save Apple customers a dime. terminally curious, Ben ___________________________________________________________________ Benjamin Trumbull trumbull@cs.yale.edu Yale University You can't be in hell; you can still read your e-mail
From: don_arb@wolfenet.com (Don Arbow) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Fri, 10 Apr 1998 12:19:02 -0700 Organization: EveryDay Objects, Inc. Message-ID: <don_arb-1004981219030001@sea-ts1-p29.wolfenet.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gecns$f9p$1@news01.deltanet.com> <1d767ho.1tixfqy1dqgm03N@hobbit1.injep.fr> <6gg6af$fit$1@anvil.BLaCKSMITH.com> <6ggabn$rts$2@news01.deltanet.com> <6ggj21$fit$6@anvil.BLaCKSMITH.com> <ericb-0904981717420001@132.236.171.104> In article <ericb-0904981717420001@132.236.171.104>, ericb@pobox.com (Eric Bennett) wrote: : In article <6ggj21$fit$6@anvil.BLaCKSMITH.com>, cswiger@BLaCKSMITH.com wrote: : : : > However, Apple does not want to spend the time (== money) and resources : to provide : > extensive support for DR1 to individual people who just want to play around. : : They don't have to provide support. I get all the support I need from : Usenet, mailing lists, and the documentation. I've had plenty of : Macintosh programming questions, and have never directed any of them : specifically at Apple (although I have had Apple employees respond to help : requests in Internet forums). Well, since DR1 is currently under NDA, the only place to get your questions answered are from Apple itself. You cannot post questions in public forums about software which you have agreed to Apple that you will not discuss. Don -- Don Arbow, Partner, CTO EveryDay Objects, Inc. don_arb@wolfenet.com <-- remove underscore to reply http://www.edo-inc.com
From: float@interport.net Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Followup-To: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Date: 9 Apr 1998 12:35:10 -0400 Organization: Interport Communications Corp. Message-ID: <6gitbu$cg1$1@interport.net> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <6ggvub$oop$5@ironhorse.plsys.co.uk> mmalcolm crawford (malcolm@plsys.co.uk) wrote: : In <B1511019-15903@206.165.43.154> "Chicken Little" wrote: : > Trying to make developer *support* a profit center is an INSANE idea. : > : Apple acting as a charity subsidising people's hobbies is what's insane. : Nor should they be offering a program wide open to abuse by people signing up : as developers just to get a discount on hardware. The dynamics of platform adoption don't work the way you think. It doesn't matter why someone signs on to the program: every sign-on is a win for Apple, every hardware sale, discounted or not, is a win for Apple, every developer, end-user, or hobbyist who gets on the Rhapsody bandwagon is a win for Apple. -- Ben <Just Another System Administrator>
From: float@interport.net Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Followup-To: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Date: 9 Apr 1998 12:42:27 -0400 Organization: Interport Communications Corp. Message-ID: <6gitpj$csn$1@interport.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <joe.ragosta-0904980642200001@elk33.dol.net> Joe Ragosta (joe.ragosta@dol.net) wrote: : Or hobbyist pseudo developers--people who never wrote an app or never : planned to, but paid the $250 for access to hardware and software : discounts. Apple should be ACTIVELY ENCOURAGING these people. Some of them might grow up to be software developers . . . or management . . . or sysadmins . . . or end users . . . or, or, or . . . -- Ben <Just Another System Administrator>
From: gmgraves@slip.net (George Graves) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Apple's Announcement Misleading. WAS :Apple developer program Date: Thu, 09 Apr 1998 09:52:02 -0700 Organization: Graves Associates Message-ID: <gmgraves-0904980952020001@sf-pm5-27-91.dialup.slip.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> <MPG.f960500e190219d989895@news.supernews.com> <joe.ragosta-0804982250580001@wil62.dol.net> I spoke, This AM, with someone from Apple Developer Support. He said that the E-mail that all developer's received about Apple's changes in their Dev program (and the source of Ric Ford's Info), is misleading. The actual impetus, apparently, for these changes, is to IMPROVE accessability to these resources, not to make them more expensive. While notoutlining what the actual changes are, he said that at the low end of the tier, Apple was trying to divorce the dev materials from the support, in order to make those materials cheaper. He said that it is now possible for small developers to get the seeding materials and the monthly mailings at much less cost than the $250 minimum under the old program. Unfortunately, the Apple E-mail sent to all of its developers does not put it this way. According to my source within Apple Developer support. the E-mail was "poorly written......." We shall see. George Graves
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 10 Apr 1998 21:27:31 GMT Organization: Digital Fix Development Message-ID: <6gm2s3$213$1@news.digifix.com> References: <joe.ragosta-1004980619170001@elk81.dol.net> <B153C824-106D1@206.165.43.115> In-Reply-To: <B153C824-106D1@206.165.43.115> On 04/10/98, "Lawson English" wrote: >Joe Ragosta <joe.ragosta@dol.net> said: > >> > Yep. Abandoning superior technology for nothing in the pipeline (no >> > upgrades to newtonOS, GX, CD, OpenDoc, etc will ever be done, meaning >> that >> > any developer or user who bought into those is now orphaned) is always >> > destructive. >> >> Right. I'm still chuckling over all your statements about GX's >superiority >> over DPS which have been consistently wrong. > >No, you're chuckling over DPS-lover's propoganda. Remember that the only >feedback about DPS that you are getting has been from people that use DPS, >not GX. Also, the people in charge of graphics at Apple *invented* DPS, not >GX, so they have a certain level of bias, also. Horse-shit. You've never used DPS. I waited YEARS for Apple to release GX, and then abandoned the platform when it didn't materialize. The people in charge of Rhapsody graphics at Apple (Mike) KNOWS GRAPHICS. Thats all that is important to getting a great graphics environment. I'll bet we'll get it in a much more timely fashion too. > > >And the only difference between "hurt Apple" and "destroy Apple" is >time-frame. > >"Hurt Apple" often enough and it is destroyed. > Well, you are certainly hell-bend on destroying it. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: Eric Hermanson <eric@alum.mit.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Fri, 10 Apr 1998 15:13:54 -0700 Organization: Digital Universe Corporation Message-ID: <352E99A2.3E7E35FA@alum.mit.edu> References: <01bc445f$744397e0$1bf0bfa8@davidsul> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit macghod wrote: > As far as I can tell it adds nothing and is a 2x increase!! Why to support > development apple :) I don't know what everyone is bitching about? I was just upgraded (for free) from the $2,500/year Enterprise Alliance program to the $3,500/year Premium Developer program. I get two additional technical support calls, and licenses to every single piece of commercial software Apple Enterprise ships. They're also throwing in a free full pass to the Apple World Wide Developer Conference (over $1,000 value). Sounds like a good deal to me. Eric
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 10 Apr 1998 15:24:00 -0700 Organization: Primenet Services for the Internet Message-ID: <B153EB42-13174@206.165.43.22> References: <6gm2s3$213$1@news.digifix.com> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Scott Anguish <sanguish@digifix.com> said: Lawson English <english@primenet.com> said: > > > >No, you're chuckling over DPS-lover's propoganda. Remember that the > only > >feedback about DPS that you are getting has been from people that use > DPS, > >not GX. Also, the people in charge of graphics at Apple *invented* > DPS, not > >GX, so they have a certain level of bias, also. > > Horse-shit. > I presume that you justify your language below... > You've never used DPS. I waited YEARS for Apple to release > GX, and then abandoned the platform when it didn't materialize. > Nope. No justification here. > The people in charge of Rhapsody graphics at Apple (Mike) > KNOWS GRAPHICS. > Nope. No justification here. > Thats all that is important to getting a great graphics > environment. > Nope. No justification here. > I'll bet we'll get it in a much more timely fashion too. > > Nope. No justification here. > > > > > >And the only difference between "hurt Apple" and "destroy Apple" is > >time-frame. > > > >"Hurt Apple" often enough and it is destroyed. > > > > Well, you are certainly hell-bend on destroying it. > Because I dare to criticize some that NeXT invented? ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: j-jahnke@uchicago.edu (Jerome Jahnke) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Sat, 11 Apr 1998 04:44:01 -0500 Organization: University of Chicago -- BSDIS/CRT Message-ID: <j-jahnke-1104980444020001@192.168.1.3> References: <6glos8$oop$17@ironhorse.plsys.co.uk> <B154522B-157593@207.217.155.30> In article <B154522B-157593@207.217.155.30>, "Brad Hutchings" <brad@hutchings-software.com> wrote: > mmalc quoted Jordan J. Dea-Mattson > > >Apple has not been cashing anyone's checks on new enrollments or renewals > >for over six months. Everyone who has had to renew or join the program in > >the last six months has gotten it gratias. > > That statement IS NOT true. I know of one developer whose credit card was > charged in December for new Associates membership. I'd be more than happy > to put ADR reps in touch with that developer. Look, it's only $250, but if > the "Apple response" to developers who don't like this move is "RTFP" or > "deal with it" or "get a clue", then incorrect assertions like the above > are going to be criticized. > > NOTE: I'm not accusing mmalc or Jordan J. Dea-Mattson of lying. I'm just > saying that the statement quoted above is not true. I am one of those guys, my renewal date is "tada" 12/98. Jer,
From: Steve Kellener <skellener@earthlink.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.programmer Subject: Re: first day with OS Date: Fri, 10 Apr 1998 16:13:09 +0100 Organization: EarthLink Network, Inc. Message-ID: <352E3703.767B@earthlink.net> References: <01bc462d$87ca8380$40f0bfa8@davidsul> <6gl0ja$30u$1@pump1.york.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit -bat. wrote: > > 2) The mouse really annoys me. It reminds me of when I first used windows > > 3.1, I was like god I hate how jerky the mousee is. The mouse movement in > > OS 4.2 seems like how it was in windows 3.1, windows 95 definitely seems > > much better. > > Hmmm... using a serial mouse ? If so get a PS/2 or Bus mouse. There was a newer driver that should eliminate the jerkiness. But PS/2 or Bus would be better.
From: "Brad Hutchings" <brad@hutchings-software.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 10 Apr 98 17:01:41 -0700 Organization: EarthLink Network, Inc. Message-ID: <B1540102-263C7@207.217.155.30> References: <352E99A2.3E7E35FA@alum.mit.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit nntp://news.earthlink.net/comp.sys.mac.oop.powerplant, nntp://news.earthlink.net/comp.sys.mac.programmer.codewarrior, nntp://news.earthlink.net/comp.sys.next.advocacy, nntp://news.earthlink.net/comp.sys.next.programmer Eric Hermanson wrote: >I don't know what everyone is bitching about? I was just upgraded (for free) >from the $2,500/year Enterprise Alliance program to the >$3,500/year Premium Developer program. I get two additional technical support >calls, and licenses to every single piece of >commercial software Apple Enterprise ships. They're also throwing in a free >full pass to the Apple World Wide Developer Conference >(over $1,000 value). Sounds like a good deal to me. You are witnessing Classic Apple: Divide and Conquer. Clobber one constituency while quoting another (see today's Apple reply at MacCentral). Microsoft, on the other hand, is oft criticized for its assimilation practices. Scoreboard anyone? When your consituency (especially your customers/users, that is the worst) are put down and demeaned by Mr. Interim CEO, and when you're told that you're "out" despite the fact that more people are using your software than the vapor Mr. Interim CEO wants to peddle (Newton vs. MacLite comes to mind, although other examples hit closer to home), you'll know what its like to be in a clobbered constituency. And given the number of clobbered constituencies in recent times, you'll understand how otherwise rational developers can blow gaskets en masse over changes to the developer program. Developer Relations is more than just a set of programs. It's a way of doing business, and it's not clear that Apple Management understands that. Brad <mailto: "Brad Hutchings" brad@hutchings-software.com> <http://www.hutchings-software.com> Got OpenDoc? Got Cyberdog? Then try Rapid-I Button. Tried it? Then buy it! Point your Cyberdog to the Hutchings Software web site.
From: devnull@occam.com (Leon von Stauber) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Followup-To: comp.sys.mac.advocacy,comp.sys.next.advocacy Date: 11 Apr 1998 02:35:52 GMT Organization: Occam's Razor Message-ID: <6gmku8$n6j$1@news.cmc.net> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: english@primenet.com NOTE: FOLLOWUPS TO COMP.SYS.MAC.ADVOCACY,COMP.SYS.NEXT.ADVOCACY In <B1511189-1AF6A@206.165.43.154> "Lawson English" wrote: > >But shareware programmers often fill in the gaps that commercial software >leaves. Also shareware programmers learn the ropes via feedback from their >customers and are often high school/college students who later enter the >Mac programming field as professionals. > >There's one very, VERY important piece of software that wouldn't exist if >shareware wasn't easy to do on the Mac: > >Stuffit. > >Raymond Lau put himself through MIT by writing Stuffit when he was 16 and >Alladin Software is a pretty important software house in the Macintosh >community.. > >Where's the next Stuffit or Alladin Software for MacOS going to come from >if there aren't going to be any more Raymond Lau's? But I wonder... Was Raymond Lau actually a member of any of Apple's developer programs, or did he just get a Mac and hack on it like most people? Membership in Apple's developer programs is not a prerequisite to developing good software. _______________________________________________________________ Leon von Stauber http://www.occam.com/leonvs/ Occam's Razor, Game Designer <com.occam@leonvs> Metapath Software, UNIX System Admin <com.metapath@lvonstau> "We have not come to save you, but you will not die in vain!"
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 10 Apr 1998 17:57:00 -0700 Organization: Primenet Services for the Internet Message-ID: <B1540F1E-99E8B@206.165.43.22> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Brad Hutchings <brad@hutchings-software.com> said: > Developer Relations is more than just a set of programs. It's a way of > doing business, and it's not clear that Apple Management understands that. > Strikes me that "Apple Management" is an oxymoron that rivals "Apple Advertising." BTW, anyone read the new excerpts from Amelio's book about how Steve Jobs approached Amelio well before the NeXT purchase claiming that he was the ultimate CEO for Apple and that Amelio should just step aside and let the right stuff handle the job? But nyah, there was never any question that Jobs and Ellison hadn't collaborated on the Jobs takeover of Apple. ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: Bluedays <nospam@spam.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 9 Apr 1998 18:04:26 +0200 Organization: BlueDays Software Sender: ccauser@ifaedi.insa-lyon.fr Message-ID: <6giria$nng$1@ifaedi.insa-lyon.fr> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <6ggiu7$auu$1@vixen.cso.uiuc.edu> <352c857a.2678525846@news.u-psud.fr> <joe.ragosta-0904980645050001@elk33.dol.net> 5[}Lsc={Nz:qqzPHDsEk/+Kwsi|^[n$(d%oR"D7X,!&!Ma|gjHJnv$Oxxr!XA[X*/jsG iX8gev&1[W<^vSJ"'FU'r[Lxen]NMJqvS]7[x-;'jO<M!Q#:Jq;m?~7Lzb/fSwtl~f!4 L#=/n[5"1lk@G>ROW-WM<}&%z@98],c"aj:9eYn#Op-BNut;MyPMR In article <joe.ragosta-0904980645050001@elk33.dol.net>, Joe Ragosta <joe.ragosta@dol.net> wrote: >In article <352c857a.2678525846@news.u-psud.fr>, pointal@lure.u-psud.fr wrote: > >> On 8 Apr 1998 19:24:55 GMT, gcheng@uni.uiuc.edu (Guanyao Cheng) wrote: >> >> >: Although I disapprove of the move by Apple, I really question why a >> >: HS/college student would have to join the developer programs to write >> >: software. After all, they could just download MPW or buy CodeWarrior at a >> >: huge discount, and learn from the vast amount of online resources at >> >: devworld.apple.com. >> >> Do you know the bandwitdh of US servers out of the US? Do you know the >> cost of communications out of the US? >> >> For a student working at home, CDs are far better with his <1 to 5 >> kb/sec IP connection to US (at ~ $1/hour com cost + fixed provider >> cost). > >That is probably true. But wouldn't most students have access to their >college's facilities where they can download what they need by T1 and put >it onto a zip drive or burn a CD to take home? > Of course, Mrs I know everything, do you really know how are the campus in europe ? Do you really think that every campus has a internet connexion ? with free access to zip drive and CD burner ? I'm sorry but you are wrong, my school cannot afford this kind of expense, and it's the same thing in lot of schools. labs inside campus have this but it's a minority. If you read the posts today about the problem, you'll see that a lot of complains come from europe, we don't need the free pass for the WWDC because of the flight price. Even with a T1 the bandwitdh with the US is bad...1K/s and less sometimes (often) And if you think that now this will be the only way to get information for many people if you don't have the CD... Macintosh prices are higher in some places in europe than in US, so the discount lost is not a real good news.
Newsgroups: comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software From: brianw@sounds.wa.com (Brian Willoughby) Subject: OPENSTEP/NeXT (Re: C++ on NeXT?) Message-ID: <Er8B58.387.0.scream@sounds.wa.com> Organization: Sound Consulting, Bellevue, WA, USA References: <3514195A.950A6554@cs.utexas.edu> Date: Sat, 11 Apr 1998 03:03:07 GMT In article <3514195A.950A6554@cs.utexas.edu>, Makoto Sadahiro <sadahiro@cs.utexas.edu> wrote: > Hi, I am wondering if someone can help me to solve my curiousity. I >understand that you use Object-C if you use any of NeXT machine. You >can not use OPENSTEP nor developer version Rhapsody. You can use OPENSTEP on NeXT hardware, up to the latest version 4.2 OPENSTEP apps seem slow on the old 25 MHz 68040, but NEXTSTEP apps run at full speed under OPENSTEP. Any tool that has been replaced, such as ProjectBuilder and InterfaceBuilder, seem to be slow on startup, but run at a reasonable speed after that. Compiling (particularly linking) seems slow under OPENSTEP, but you do gain a lot of features over NEXTSTEP. If you can get two NeXT machines, it would be best to network them and have NEXTSTEP on one, and OPENSTEP on the other. This also makes a nice arrangement for porting NEXTSTEP software to Rhapsody - the OPENSTEP tools can do most of the conversion, and Rhapsody is very similar to OPENSTEP. You can not use Rhapsody on NeXT hardware. -- Brian Willoughby NEXTSTEP, OpenStep, Rhapsody Software Design Sound Consulting Apple Enterprise Alliance Partner NeXTmail welcome Macintosh Associate Apple is the registered trademark of Apple Computer, Inc. and Apple Records
From: rmcassid@uci.edu Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 09 Apr 1998 23:55:20 -0700 Organization: University of California, Irvine Message-ID: <rmcassid-0904982355200001@dante.eng.uci.edu> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <6ggvub$oop$5@ironhorse.plsys.co.uk> <352CA831.D4878417@ArtQuest.fr> <9APR98.16105101@fredjr.pfc.mit.edu> In article <9APR98.16105101@fredjr.pfc.mit.edu>, FRIDBERG@PSFC.MIT.EDU wrote: >And let's not forget such things as NewsWatcher (which probably half of people >who read this NG are using, or NSCA Telnet (which I am using) or bunch of other >internet software which made Mac so attractive to use for Internet access >and allowed us to have pretty high market share on internet. Mister MMalcolm >Crawford prbably never heard of such programs as Disinfectant of Internet >Config either. Or maybe he just afraid that all those freeware applications >going to drive him out of business? Uh, those are all developed at higher ed except for Internet Config. Apple seems to recognize the need to support development at higher-ed as we can get a free developer license for the campus. Clearly Apple seems to recognize where some of the most important MacOS freeware has come from in the last several years. Fear not, the sky is not falling. -Bob Cassidy
Date: Fri, 10 Apr 1998 06:19:17 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <joe.ragosta-1004980619170001@elk81.dol.net> References: <joe.ragosta-0904981531590001@wil38.dol.net> <B152C422-16106C@206.165.43.139> In article <B152C422-16106C@206.165.43.139>, "Lawson English" <english@primenet.com> wrote: > Joe Ragosta <joe.ragosta@dol.net> said: > > > Just like the last 20 rants that you went on were going to destroy Apple. > > Wasn't abandoning QD GX supposed to destroy Apple? > > > > Yep. Abandoning superior technology for nothing in the pipeline (no > upgrades to newtonOS, GX, CD, OpenDoc, etc will ever be done, meaning that > any developer or user who bought into those is now orphaned) is always > destructive. Right. I'm still chuckling over all your statements about GX's superiority over DPS which have been consistently wrong. But you need to learn the difference between "will hurt Apple" and "will destroy Apple" -- Regards, Joe Ragosta See the Complete Macintosh Advocacy Page http://www.dol.net/~Ragosta/complmac.htm
From: xhsoft@injep.fr (Xavier Humbert) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Fri, 10 Apr 1998 12:34:04 +0200 Organization: Maquis Usenet Ouest Message-ID: <1d78rmx.1lk8pnd1s3f2ybN@hobbit1.injep.fr> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Mail-Copies-to: never mmalcolm crawford <malcolm@plsys.co.uk> wrote: > Apple's customers who no longer have to subsidise hobbyist developers. Laurent Ribardiere used to be such a hobbyist. These (good ol') days, you just have to say to Apple France you had a project, they GAVE you a Mac Plus, an ImageWriter and MPW. Now, ACI/ACIUS is a worldwide major Apple Software company. Wether 4thD sucks or not is not the point :-) Xav -- Xavier HUMBERT Laboratoire Informatique INJEP Office: labo-info/AT/injep.fr Home: humbert/AT/injep.fr
From: pete@ohm.york.ac.uk (-bat.) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.programmer Subject: Re: first day with OS Date: 10 Apr 1998 11:42:34 GMT Organization: The University of York, UK Sender: pcf1@york.ac.uk Message-ID: <6gl0ja$30u$1@pump1.york.ac.uk> References: <01bc462d$87ca8380$40f0bfa8@davidsul> "NeXT Newbie" <macghod@concentric.net> writes: > Some impressions of OS 4.2 > 1) This OS sure was designed for big monitors! I have a 17 inch and it > seems to small. Seems like you would want at least a 20 inch monitor 17 inch is o.k. if you set the resolution high enough. Anything less that 1024x768 is a right pain (and even thats not very good) > 2) The mouse really annoys me. It reminds me of when I first used windows > 3.1, I was like god I hate how jerky the mousee is. The mouse movement in > OS 4.2 seems like how it was in windows 3.1, windows 95 definitely seems > much better. Hmmm... using a serial mouse ? If so get a PS/2 or Bus mouse. I always disliked the serial mice under OpenStep. They were always appalling. I haven't used on recently though. If, however, you are referring to the pproprtionality of it then once you get used to it you will find that it's a lot nicer and requires much less desk space. > 3) How do you connect to the internet with ppp? ppp is shipped with it - configuring it can be a bit of a pain. You might like to take a look at gatekeeper.app on the archive. > 4) once I get up and going with ppp, what do I use for usenet? Any UNIX newsreader. > 5) where is the option to change the monitors resolution and bits of color? run up configure.ap as root, go to the display section and theres a button marked "select" or something similar. -bat.
From: lars.farm@ite.mh.se (Lars Farm) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer,comp.sys.mac.programmer.codewarrior Subject: Re: Apple developer program Date: Sat, 11 Apr 1998 14:50:19 +0200 Organization: pv Message-ID: <1d7brgb.mh0tsh1y0bbeyN@dialup116-1-32.swipnet.se> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <trumbull-0904980039090001@net44-223.student.yale.edu> <6gio5m$oop$11@ironhorse.plsys.co.uk> <ericb-0904981637520001@132.236.171.104> <6gjgm5$bfj$1@news01.deltanet.com> <6glos8$oop$17@ironhorse.plsys.co.uk> Cache-Post-Path: nn1!s-49817@dialup116-1-32.swipnet.se mmalcolm crawford <malcolm@plsys.co.uk> wrote: > According to Jordan J. Dea-Mattson > Senior Partnership & Technology Solutions Manager > Apple Developer Relations > Apple has not been cashing anyone's checks on new enrollments or renewals > for over six months. Everyone who has had to renew or join the program in > the last six months has gotten it gratias. False. Apple cashed my renewal in late february. -- Lars Farm; lars.farm@ite.mh.se - Limt/Channelmatic: lars.farm@limt.se
From: lars.farm@ite.mh.se (Lars Farm) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Sat, 11 Apr 1998 14:50:11 +0200 Organization: pv Message-ID: <1d7bqqw.zaqaktxojgw0N@dialup116-1-32.swipnet.se> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> <1d7821x.1rdi5101enh5fhN@hobbit1.injep.fr> <6girj3$6j$2@anvil.BLaCKSMITH.com> <1d78ren.1dbj70n1j99pebN@hobbit1.injep.fr> <6glqd2$7j4$4@anvil.BLaCKSMITH.com> Cache-Post-Path: nn1!s-49817@dialup116-1-32.swipnet.se Charles Swiger <chuck-nospam@blacksmith.com> wrote: > If a one-time $250 cost change makes a difference to your company as to which > technologies to pursue, you were obviously considering a marginal > opportunity. I think you will find that the vast majority of Macintosh developers these days consider Mac development "a marginal opportunity". Many still hang on, one way or another, even though common sense says they should have abandoned ship long a go. Motivated by a strong personal interest and strong personal preferences rather than good business reasons. -- Lars Farm; lars.farm@ite.mh.se - Limt/Channelmatic: lars.farm@limt.se
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 11 Apr 1998 04:06:25 GMT Organization: Digital Fix Development Message-ID: <6gmq81$ate$1@news.digifix.com> References: <6gm2s3$213$1@news.digifix.com> <B153EB42-13174@206.165.43.22> In-Reply-To: <B153EB42-13174@206.165.43.22> On 04/10/98, "Lawson English" wrote: >Scott Anguish <sanguish@digifix.com> said: >Lawson English <english@primenet.com> said: > >> > >> >No, you're chuckling over DPS-lover's propoganda. Remember that the >> only >> >feedback about DPS that you are getting has been from people that use >> DPS, >> >not GX. Also, the people in charge of graphics at Apple *invented* >> DPS, not >> >GX, so they have a certain level of bias, also. >> >> Horse-shit. >> > >I presume that you justify your language below... > The language is due to total and complete exasperation due to YOUR GX rants. I notice that in all the rest of your reply you completely ignore the fact that you don't have any defence for your statements about DPS.. Instead, you shift the goal-posts.. Classic Lawson... -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: bl003@dial.oleane.com (Benoit Leraillez) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Sat, 11 Apr 1998 00:49:29 +0200 Organization: Guest of OLEANE Message-ID: <1d7aq7w.1yep9o018c97b2N@dialup-202.def.oleane.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <352AAF57.25EB@skdjsk.no> <6geche$8lu$1@news2.apple.com> <philipm-0804982051170001@pm1-23.eecs.umich.edu> <6gj0or$oop$12@ironhorse.plsys.co.uk> <trumbull-0904981554150001@net44-223.student.yale.edu> <don_arb-1004981207170001@sea-ts1-p29.wolfenet.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Don Arbow <don_arb@wolfenet.com> wrote: > Remember, also that Apple needed people to administer the hardware > program, take orders, get the boxes shipped, etc. I'm sure that cut into > any profits that Apple may have made on the boxes. For a punch line that's a good one. Q: Who just opened an online store? R: Dduuuhhhh. Benoît Leraillez
From: "Brad Hutchings" <brad@hutchings-software.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 10 Apr 98 22:47:59 -0700 Organization: EarthLink Network, Inc. Message-ID: <B154522B-157593@207.217.155.30> References: <6glos8$oop$17@ironhorse.plsys.co.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit nntp://news.earthlink.net/comp.sys.mac.oop.powerplant, nntp://news.earthlink.net/comp.sys.mac.programmer.codewarrior, nntp://news.earthlink.net/comp.sys.next.advocacy, nntp://news.earthlink.net/comp.sys.next.programmer mmalc quoted Jordan J. Dea-Mattson >Apple has not been cashing anyone's checks on new enrollments or renewals >for over six months. Everyone who has had to renew or join the program in >the last six months has gotten it gratias. That statement IS NOT true. I know of one developer whose credit card was charged in December for new Associates membership. I'd be more than happy to put ADR reps in touch with that developer. Look, it's only $250, but if the "Apple response" to developers who don't like this move is "RTFP" or "deal with it" or "get a clue", then incorrect assertions like the above are going to be criticized. NOTE: I'm not accusing mmalc or Jordan J. Dea-Mattson of lying. I'm just saying that the statement quoted above is not true. >> mmalc finsihes: >> Jordan finishes by asking "Is anyone mentioning this fact?" >> The answer is clearly "No." People would rather yell at Apple and bitch >> about something they haven't learned enough about. The problem here is that "facts" aren't consistent with reality. Given the reaction to this thing from some REAL COMMERCIAL developers (check out MacCentral if there's any doubt that we're talking about a segment of real commercial developers here), I find it hard to believe that ADR even beta tested this plan. Look, they rolled it out and it crashed when confronted with many developers' business models. It's like if any of us developers shipped software that didn't work on PowerBooks but worked on G3s. And then, when LONGTIME PAYING CUSTOMERS complained, we told 'em the G3 crowd was happy and suggest they go buy one. Dear Apple Management- Developers are telling you this because they love Apple and still believe in Apple. If they didn't love Apple, they'd say "see ya, wouldn't wanna be ya". Developers of all sizes still want to give you a chance to be developer friendly. Seize it before it really is too late. Brad <mailto: "Brad Hutchings" brad@hutchings-software.com> <http://www.hutchings-software.com> Got OpenDoc? Got Cyberdog? Then try Rapid-I Button. Tried it? Then buy it! Point your Cyberdog to the Hutchings Software web site.
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 11 Apr 1998 12:09:00 -0700 Organization: Primenet Services for the Internet Message-ID: <B1550F01-581B6@206.165.43.155> References: <philipm-1104981325480001@pm1-25.eecs.umich.edu> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Philip Machanick <philipm@NO-SPAMeecs.umich.edu> said: > In article <joe.ragosta-0804982109560001@elk62.dol.net>, > joe.ragosta@dol.net (Joe Ragosta) wrote: > > >Anyone notice that people like Mr. brown here, who would apparently > never > >go near a Mac from reading his posts, are the ones complaining while > >developers like mmalc are supporting the change? > > I'm on a mailing list mainly populated by developers and I can assure you > that most of them are pretty upset (including some who have been pretty > defensive of Apple in the past), not so much on the substance as on the > message. Mr. Brown is also a Mac developer and bleeds 6 colors. [hell, *I* bleed six colors (if I understand what the phrase means): I was writing a graphics library for HyperC for Apple IIs when Dave McClain hired me to work on Macintoshes in '86] ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Blocks for Obj-C Date: 11 Apr 1998 15:11:02 -0400 Organization: Virginia Tech, Blacksburg, Virginia Message-ID: <6gof86$gkk$1@crib.bevc.blacksburg.va.us> References: <6fokj8$1dp$1@usenet50.supernews.com> <6g8r4c$2uq$1@crib.bevc.blacksburg.va.us> <6gb4ib$bdj$1@news.cs.tu-berlin.de> <352FB748.41C67EA6@mundivia.es> In article <352FB748.41C67EA6@mundivia.es>, David Stes <stes@mundivia.es> wrote: > In comp.lang.objective-c Marcel Weiher wrote: > > [ ... idea to improve protocols .. ] > But anyhow, the idea seems not related to Blocks, which somehow is in > your subject line. > [description of blocks] Yeah, I think those of us involved in this thread know what blocks are (though I'm speaking only for myself).. the thread was inaptly named.. it started out when someone said they wanted blocks (I think), and I replied that Obj-C would be the perfect language if it had blocks plus this protocol category feature we've been discussing, and then no one bothered to ever change the subject line. I really wish the Apple compiler had blocks (a.k.a. "closures", "lambda" or "anonymous functions"). I really want to use them in my Rhapsody development. They're hard to appreciate at first by some ("why don't you just make a new method?"), but once you get used to them, you hate to live without them.
From: nhughes@sunflower.com (Nathan Hughes) Newsgroups: comp.soft-sys.nextstep,comp.sys.mac.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Newby needs help Date: Sat, 11 Apr 1998 19:35:20 GMT Organization: is a sign of a sick mind. Message-ID: <3530c496.8026484@198.0.0.100> References: <01bc46e5$6b4880a0$24f0bfa8@davidsul> <6god24$lq4$1@news.rt66.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit On or about 11 Apr 1998 18:33:40 GMT, in comp.sys.mac.advocacy jakeg@rt66.com (Jake Garcia) exclaimed : >NeXT Newbie (macghod@concentric.net) wrote: <s> >Doesn't the standard Windows 95 filesystem limit file/dir names to 14 chars? >I could be wrong... In reality, Win95 sits on a dos partition which is limited to 8.3 filenames. 95 works around this limitation, but Unix does not (generally). Their are utilities for Linux that allow it to see LFNs but it isn't part of the distribution, its probably the same with NeXt. You can work around the problem by zipping your tars and gzips in Win95 with LFNs and unzipping them into NeXt; at least that works in Linux. Nathan A. Hughes MFA Candidate The University Theatre KU http://sunflower.com/~nhughes
From: dcoshel@pobox.com (Dave Oshel) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program -shareware Date: Sat, 11 Apr 1998 15:29:37 -0500 Organization: Xochitl Sodality Message-ID: <dcoshel-1104981530090001@323-a-43-57.ppp.mcleodusa.net> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggkg1$90k$1@news0-alterdial.uu.net> <SCOTT.98Apr8221520@slave.doubleu.com> Mail-Copies-To: always In article <SCOTT.98Apr8221520@slave.doubleu.com>, scott@doubleu.com (Scott Hess) wrote: [whack] > Given two new hires fresh out of college, one of which has a fair > covey of praised freeware, the other with good grades, which gets the > more important job? I've always wondered that. My one brief stint in the shrinkwrap software biz kind of revealed that the guys (gals note below) who already have jobs somewhere else get hired to do "important" stuff, while the new hires fresh outa college work in q.a. or in intern slots. Once in q.a., always in q.a. Interns have possibilities. Regardless of ability, the guys (the gals are guys too, in the sexless company way of looking at things) are all unmarried "coders" who work 80 hours a week 7 days a week and who "shower" naked at 6:30 AM in the company toilet stalls with a huge bronze can of spray deodorant. The married unmarried coders soon regain company-preferred status when their divorces go through. Anyone who is a "software engineer" or what used to be called a "systems analyst" or "programmer analyst" (in the old days - yes, my hair is gray) is getting paid far too much money, and will be demoted or reassigned for bungling an innocent-looking project with impossible design criteria, such as, "must be Mac native, must use MS VC++ 4.0 Cross-compiler Edition for Macintosh, especially for look and feel." Successful team leaders will be promoted to new ways of rewriting proven software with unproven technologies and accelerated development schedules; this improves the bottom line by increasing the pool of unvested (i.e., unclaimed) stock options. The rest of the schmucks get profit-sharing, which is a way to pay out a $4000 raise in up to 4 smaller lump sums, each positioned after an implacable deadline designed to shake malcontents off the trolley. Unionize? Your average libertarian geek would rather die. Life in the in-house software trades is far less exciting, thank God. -- David C. Oshel dcoshel@pobox.com Cedar Rapids, IA http://soli.inav.net/~doshel "Tension, apprehension and dissension have begun." - Duffy Wyg&, in Alfred Bester's _The Demolished Man_
From: philipm@NO-SPAMeecs.umich.edu (Philip Machanick) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Sat, 11 Apr 1998 13:45:06 -0400 Organization: Department of EE and Computer Science, The University of Michigan Message-ID: <philipm-1104981345070001@pm1-25.eecs.umich.edu> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <352AAF57.25EB@skdjsk.no> <6geche$8lu$1@news2.apple.com> <philipm-0804982051170001@pm1-23.eecs.umich.edu> <6gj0or$oop$12@ironhorse.plsys.co.uk> In article <6gj0or$oop$12@ironhorse.plsys.co.uk>, mmalcolm crawford <malcolm@plsys.co.uk> wrote: >By offering a hardware discount of course Apple was subsidising developers -- >that's what the discount was for. Now they've raised the bar to ensure in Are you claiming that Apple was making less money on Macs sold through the developer program? I doubt it. Maybe you could argue that _dealers_ were subsidising developers since they were the ones who were really losing by allowing developers to buy direct from Apple. But I doubt the developer discounts were as low as the price to a really big dealer. Ironic that Apple should be cutting something like this just when they've accepted the principle of direct sales through the web store... -- Philip Machanick Dept. of Computer Science, Univ. of Witwatersrand on sabbatical until September 1998 at Dept EECS, University of Michigan http://www.eecs.umich.edu/~philipm/ mailto:philipm@NO-SPAMeecs.umich.edu
From: kvwright@sans.vuw.ac.nz (Ken Wright) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Sun, 12 Apr 1998 09:50:26 +1300 Organization: Victoria University of Wellington, New Zealand. Message-ID: <1d7c5ae.11ouwaz1r6e7ckN@ppp2.sans.vuw.ac.nz> References: <joe.ragosta-1004980621290001@elk81.dol.net> <B153C956-14E86@206.165.43.115> Cache-Post-Path: totara.its.vuw.ac.nz!unknown@ppp2.sans.vuw.ac.nz Lawson English <english@primenet.com> wrote: > Joe Ragosta <joe.ragosta@dol.net> said: > > > > SDK's and Mac API development don't interest me. I want to know what > > will > > > happen to the cost of developing for Rhapsody. > > > > > > > How about zero dollars? Download MkLinux and GNUStep. > > > > GNUStep doesn't exist as a useable product, letalone a standard anyone > could develop for. And what does MKLinux have to do with Rhaposdy > developement? The fact that they are both Unices? > > ---------------------------------------------------------------------- > Want Apple to license Cyberdog for third-party development? Go to: > <http://www.pcsincnet.com/petition.html> > ---------------------------------------------------------------------- -- ______________________ Breathe the pressure
From: "Earl Malmrose" <malmrose@nospam.jetcity.com> Newsgroups: comp.soft-sys.nextstep,comp.sys.mac.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Newby needs help Date: 11 Apr 1998 20:07:17 GMT Organization: ISPNews http://ispnews.com Message-ID: <01bd6585$58835660$0b0ba8c0@woohoo> References: <01bc46e5$6b4880a0$24f0bfa8@davidsul> <6god24$lq4$1@news.rt66.com> Jake Garcia <jakeg@rt66.com> wrote in article <6god24$lq4$1@news.rt66.com>... > Doesn't the standard Windows 95 filesystem limit file/dir names to 14 chars? > I could be wrong... You're wrong. Standard Win95 filesystem supports filenames whose complete file and path length is up to around 260 characters.
From: easter@egg.com Newsgroups: comp.sys.next.programmer Subject: Do you want to earn CASH $$$ ? Date: Sun, 12 Apr 1998 06:09:15 -0700 Organization: <no organization> Message-ID: <120498060915@egg.com> It Works!!!! First, I would like to start off by telling you that I am only 15 and ROLLING in cash. I'm thinking about buying a new car when I get my driver's license. This WORKS! But you must follow my instructions to the letter. Once you get the hang of it though, it's really easy. So if you want some extra cash, and would like to earn it by doing nothing, then it's worth it. All you have to do is follow the directions and you'll make a ton!. To get the details, read on....................... THIS REALLY CAN MAKE YOU EASY MONEY! A little while back, I was browsing some newsgroups and came across an article similar to this that said you could make thousands of dollars within weeks with only an initial investment of $6.00! So I thought, "OK, why not try this, if its a scam I'll only lose $6.00." Besides, it's so simple - no complicated stuff like mailing disks to people [some of the others are based on selling computer disks to people]. Anyway, it told me to send $1.00 to each of the 6 people on the list. You then place your own name and address in the bottom of the list at #6, and post the article in some newsgroups (there are literally thousands). No catch, that was it. So I thought 'hey, what have I got to lose except six stamps and a few bucks, right?' But like most of you I was still a little skeptical and a little worried about the legal aspects of it all. So I checked it out with the U.S. Post Office (1-800-725-2161) and they confirmed that it is indeed legal! I then invested the measly $6.00, and stamps. Well, GUESS WHAT.....within 7 days, I started getting money in the mail! I was shocked! I still figured it would end soon, and didn't give it another thought. I had done an experiment, and I made back my $7.92. But the money just kept coming in. In my first week, I made about $36.00 dollars. By the end of the second week I had made a total of over $1,000.00!!!!!! In the third week I had over $10,000.00 and it's still growing. This is now my fourth week and I have made a total of just over $42,000.00 and it's still coming in VERY rapidly! It's certainly worth $6.00, and 6 stamps, You could spend more than that on the lottery!! Let me tell you how this works and most importantly, WHY it works..... also, you might want to print a copy of this article, so you can refer to it as needed. The process is very simple and consists of 3 easy steps: STEP 1: Get 6 separate pieces of paper and write the following on each one: "PLEASE PUT ME ON YOUR MAILING LIST" + your name and address. Now get six American $1.00 bills and place ONE inside EACH of the 6 pieces of paper and fold it over so the bill will not be seen through the envelope (to prevent theft). Now, put one paper in each of the six envelopes and seal them. You should now have six sealed envelopes, each with a piece of paper stating the above phrase, your name, your address, and a $1.00 bill. STEP 2: Mail the 6 envelopes to the following addresses: *1 Earl Sampson 16975 Cherry Crest Ave. Lake Oswego, OR 97O34 *2 Mark Bailey 39O6 Arnold Ct. Cleveland, OH 441O9 *3 Kane McDolh 327O De la Pepiniere ave. appartment 2O8 Montreal, Quebec Canada H1N-3N4 *4 David Lane 816 Danbbury Road Cincinnati, Ohio 45240 *5 Jeremy Fulton 5804 Inman Park Circle Unit 370 Rockville, MD 20852 *6 Arkay Sajorda 1B Jalan Tun Sambanthan 4, Brickfields 50470 Kuala Lumpur, Malaysia STEP 3: Now take the #1 name off the list that you see above, move the other names up (#6 becomes #5, #5 becomes #4, etc...) and add YOUR Name as #6 on the list. STEP 4: Change anything you need to, but try to keep this article as close to original as possible. Now, post your amended article to some newsgroups (I think there are close to 24,000 newsgroups) and/or email it to some people. remember, the more you post, the more money you make! Don't know HOW to post in the newsgroups? Well do exactly the following: --------------------------------------------------------------------------- DIRECTIONS - HOW TO POST TO NEWSGROUPS --------------------------------------------------------------------------- Step 1: You do not need to re-type this entire letter to do your own posting. Simply put your cursor at the beginning of this letter, click and hold down your mouse button. While continuing to hold down the mouse button, drag your cursor to the bottom of this document and over to just after the last character, and release the mouse button. At this point the entire letter should be highlighted. Then, from the 'edit' pull down menu at the top of your screen select 'copy'. This will copy the entire letter into the computers memory. Step 2: Open a blank 'notepad' file and place your cursor at the top of the blank page. From the 'edit' pull down menu select 'paste'. This will paste a copy of the letter into notepad so that you can add your name to the list. Remember to eliminate the #1 position, move everyone up a spot (re-number everyone elses positions), and add yourself in as #6. Step 3: Save your new notepad file. --------------------------------------- FOR NETSCAPE NAVIGATOR USERS: --------------------------------------- Step 4: Within the Netscape program, go to the pull-down window entitled 'Window' select 'NetscapeNews'. Then from the pull down menu 'Options', select 'Show all Newsgroups'. After a few moments a list of all the newsgroups on your server will show up. Click on any newsgroup you desire. From within this newsgroup, click on the 'TO NEWS' button, which should be in the top left corner of the newsgroups page. This will bring up a message box. Step 5: Fill in the Subject. This will be the header that everyone sees as they scroll through the list of postings in a particular group. Step 6: Highlight the entire contents of your .txt file and copy them using the same technique as before. Go back to the newsgroup 'TO NEWS' posting you are creating and paste the letter into the body of your posting. Step 7: Hit the 'Send' Button in the upper left corner. You're done with your first one! Congratulations... ----------------------------------------------- INTERNET EXPLORER USERS: ----------------------------------------------- Step 4: Go to newsgroups and select 'Post an Article' or 'New Message.' Step 5: Fill in the subject. This will be the header that everyone sees as they scroll through the list of postings in a particular group. Step 6: Highlight the entire contents of your .txt file and copy them using the same technique as before. Go back to the newsgroup 'TO NEWS' posting you are creating and paste the letter into the body of your posting. Step 7: Hit the 'Send' Button in the upper left corner. You're done with your first one! Congratulations... ---------------------------------------------------------------------------- -------- THAT'S IT! All you have to do is jump to different newsgroupes and post away, after you get the hang of it, it will take about 30 seconds for each newsgroup! **THE MORE NEWSGROUPS YOU POST IN, THE MORE MONEY YOU WILL** **MAKE!!** That's it! You will begin reciving money from around the world within days! You may eventually want to rent a P.O.Box due to the large amount of mail you receive. If you wish to stay anonymous, you can invent a name to use, as long as the postman will deliver it. **JUST MAKE SURE ALL THE ADDRESSES ARE CORRECT.** Now the WHY part: Out of 200 postings, say I receive only 5 replies (a very low example). So then I made $5.00 with my name at #6 on the letter. Now, each of the 5 persons who just sent me $1.00 make 200 postings, each with my name at #5 and only 5 persons respond to each of the original 5, that is another $25.00 for me, now those 25 each make 200 posts with my name at #4 and only 5 replies each, I will bring in an additional $125.00! Now, those 125 persons turn around and post 200 with my name at #3 and only receive 5 replies each, I will make an additional $626.00! OK, now here is the fun part, each of those 625 persons post 200 letters with my name at #2 and they each only receive 5 replies, that just made me $3,125.00!!! Those 3,125 persons will all deliver this message to 200 newsgroups with my name at #1 and if still 5 persons per 200 newsgroups react I will receive $15,625,00! With a original investment of only $6.00! AMAZING! And as I said 5 responses is actually VERY LOW! Average is probably 20 to 30! So lets put those figures at just 15 responses per person. Here is what you could make: at #6 $15.00 at #5 $225.00 at #4 $3,375.00 at #3 $50,625.00 at #2 $759,375.00 at #1 $11,390,625.00 When your name is no longer on the list, you just take the latest posting in the newsgroups, and send out another $6.00 to names on the list, putting your name at number 6 again. And start posting again. The thing to remember is, do you realize that thousands of people all over the world are joining the internet and reading these articles everyday, JUST LIKE YOU are now!! So can you afford $6.00 and see if it really works?? I think so... People have said, "what if the plan is played out and no one sends you the money? So what! What are the chances of that happening when there are tons of new honest users and new honest people who are joining the internet and newsgroups everday and are willing to give it a try? Estimates are at 20,000 to 50,000 new users, every day, with thousands of those joining the actual internet. Remember, play FAIRLY and HONESTLY and this will work. You just have to be honest. Make sure you print this article out RIGHT NOW, also. Try to keep a list of everyone that sends you money and always keep an eye on the newsgroups to make sure everyone is playing fairly. Remember, HONESTY IS THE BEST POLICY. You don't need to cheat the basic idea to make the money!! GOOD LUCK to all and please play fairly and reap the huge rewards from this, which is tons of extra CASH. P.S. Another great way to get this out to people is to send it to your friends, relatives, etc. in e-mail. Have fun! :)
Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer From: pischke@ecf.toronto.edu (David Pischke) Subject: Re: Apple developer program Sender: news@ecf.toronto.edu (News Administrator) Message-ID: <Er9A2E.BC8@ecf.toronto.edu> Date: Sat, 11 Apr 1998 15:37:26 GMT References: <6glos8$oop$17@ironhorse.plsys.co.uk> <B154522B-157593@207.217.155.30> <j-jahnke-1104980444020001@192.168.1.3> Organization: University of Toronto, Engineering Computing Facility In article <j-jahnke-1104980444020001@192.168.1.3>, Jerome Jahnke <j-jahnke@uchicago.edu> wrote: >In article <B154522B-157593@207.217.155.30>, "Brad Hutchings" ><brad@hutchings-software.com> wrote: >> That statement IS NOT true. I know of one developer whose credit card was >> charged in December for new Associates membership. I'd be more than happy >> to put ADR reps in touch with that developer. Look, it's only $250, but if >> the "Apple response" to developers who don't like this move is "RTFP" or >> "deal with it" or "get a clue", then incorrect assertions like the above >> are going to be criticized. >> NOTE: I'm not accusing mmalc or Jordan J. Dea-Mattson of lying. I'm just >> saying that the statement quoted above is not true. > >I am one of those guys, my renewal date is "tada" 12/98. Guess when my renewal date is? March 31, 1999. They were accepting new enrolments less than a month before they made the changes.
From: schuerig@acm.org (Michael Schuerig) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Sun, 12 Apr 1998 01:03:41 +0200 Organization: Completely Disorganized Message-ID: <1d7c4mw.1thrwi2f2mbnkN@rhrz-isdn3-p50.rhrz.uni-bonn.de> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <trumbull-0904980039090001@net44-223.student.yale.edu> <6gio5m$oop$11@ironhorse.plsys.co.uk> <ericb-0904981637520001@132.236.171.104> <6gjgm5$bfj$1@news01.deltanet.com> <6glos8$oop$17@ironhorse.plsys.co.uk> <B154522B-157593@207.217.155.30> <j-jahnke-1104980444020001@192.168.1.3> <Er9A2E.BC8@ecf.toronto.edu> David Pischke <pischke@ecf.toronto.edu> wrote: > In article <j-jahnke-1104980444020001@192.168.1.3>, > Jerome Jahnke <j-jahnke@uchicago.edu> wrote: > >In article <B154522B-157593@207.217.155.30>, "Brad Hutchings" > ><brad@hutchings-software.com> wrote: > >> That statement IS NOT true. I know of one developer whose credit card was > >> charged in December for new Associates membership. I'd be more than happy > >> to put ADR reps in touch with that developer. Look, it's only $250, but if > >> the "Apple response" to developers who don't like this move is "RTFP" or > >> "deal with it" or "get a clue", then incorrect assertions like the above > >> are going to be criticized. > >> NOTE: I'm not accusing mmalc or Jordan J. Dea-Mattson of lying. I'm just > >> saying that the statement quoted above is not true. > > > >I am one of those guys, my renewal date is "tada" 12/98. > > > Guess when my renewal date is? March 31, 1999. > > They were accepting new enrolments less than a month before they made the > changes. Renewal Date for new membership (dd/mm/yy): 28/02/99 I'm not a long time member of the associate program. I just entered in Jan 1997 to have access to OpenDoc dev tools after they had been no longer publicly available (well, you know the rest of the story). In early January I was wondering how my renewal would be handled. I asked and was told by Apple's Euro Dev that I'd get a renewal form in time. I did not. Only when I called them was the renewal handled. Michael -- Michael Schuerig mailto:schuerig@acm.org http://www.uni-bonn.de/~uzs90z/
From: jason@jhste1.dyn.ml.org Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: cmsg cancel <slrn6isvlb.1kg.jason@jhste1.dyn.ml.org> Control: cancel <slrn6isvlb.1kg.jason@jhste1.dyn.ml.org> Date: 10 Apr 1998 20:20:27 GMT Organization: MindSpring Enterprises Message-ID: <6gluub$71e$1@camel15.mindspring.com> Xcanpos: shelf.1/199804120201!0103404451 ignore Article canceled by slrn 0.9.4.5
From: philipm@NO-SPAMeecs.umich.edu (Philip Machanick) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Sat, 11 Apr 1998 13:25:48 -0400 Organization: Department of EE and Computer Science, The University of Michigan Message-ID: <philipm-1104981325480001@pm1-25.eecs.umich.edu> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <MPG.f95d7e4b8fb1622989890@news.supernews.com> <joe.ragosta-0804982109560001@elk62.dol.net> In article <joe.ragosta-0804982109560001@elk62.dol.net>, joe.ragosta@dol.net (Joe Ragosta) wrote: >Anyone notice that people like Mr. brown here, who would apparently never >go near a Mac from reading his posts, are the ones complaining while >developers like mmalc are supporting the change? I'm on a mailing list mainly populated by developers and I can assure you that most of them are pretty upset (including some who have been pretty defensive of Apple in the past), not so much on the substance as on the message. -- Philip Machanick Dept. of Computer Science, Univ. of Witwatersrand on sabbatical until September 1998 at Dept EECS, University of Michigan http://www.eecs.umich.edu/~philipm/ mailto:philipm@NO-SPAMeecs.umich.edu
From: philipm@NO-SPAMeecs.umich.edu (Philip Machanick) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Sat, 11 Apr 1998 13:32:12 -0400 Organization: Department of EE and Computer Science, The University of Michigan Message-ID: <philipm-1104981332120001@pm1-25.eecs.umich.edu> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> <6ggqha$9fk$1@news01.deltanet.com> <6ghg4o$np$1@news.digifix.com> In article <6ghg4o$np$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: > If you're developing commercially, an investment of $42/month >isn't that out of reach. This is not the point. Apple makes changes that hurt some people and claims they are an improvement. It doesn't matter if the pain level is slight, or if people really could get along without the whole developer program entirely (in my opinion most people probably could, 3rd party tools have developed way beyond the state of the Mac in 1984, when Apple _really_ had to help). The point is the spin Apple has put on this. -- Philip Machanick Dept. of Computer Science, Univ. of Witwatersrand on sabbatical until September 1998 at Dept EECS, University of Michigan http://www.eecs.umich.edu/~philipm/ mailto:philipm@NO-SPAMeecs.umich.edu
From: jakeg@rt66.com (Jake Garcia) Newsgroups: comp.soft-sys.nextstep,comp.sys.mac.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Newby needs help Followup-To: comp.soft-sys.nextstep,comp.sys.mac.advocacy,comp.sys.next.misc,comp.sys.next.programmer Date: 11 Apr 1998 18:33:40 GMT Organization: Rt66.COM, New Mexico's #1 ISP Distribution: inet Message-ID: <6god24$lq4$1@news.rt66.com> References: <01bc46e5$6b4880a0$24f0bfa8@davidsul> NeXT Newbie (macghod@concentric.net) wrote: : localhost:10# mv ne* next-ppp : localhost:11# tar -xvf next-ppp : ./PPP-2.2-0.4.6-pkg: File name too long : tar: can't create ./PPP-2.2-0.4.6-pkg/: File name too long : ./PPP-2.2-0.4.6-pkg: File name too long : tar: can't create ./PPP-2.2-0.4.6-pkg/PPP-2.2.pkg/: File name too long : ./PPP-2.2-0.4.6-pkg: File name too long Doesn't the standard Windows 95 filesystem limit file/dir names to 14 chars? I could be wrong... -- Jake Garcia <jake@midnight.rt66.com> Please finger for more info...
From: David Stes <stes@mundivia.es> Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Blocks for Obj-C Date: Sat, 11 Apr 1998 18:32:40 +0000 Organization: Mundivia, Santander Message-ID: <352FB748.41C67EA6@mundivia.es> References: <6fokj8$1dp$1@usenet50.supernews.com> <6fpgbb$9o5$2@news.idiom.com> <6fphlp$iot$1@crib.bevc.blacksburg.va.us> <6fsfo4$ln@saturn.genoa.com> <6g8r4c$2uq$1@crib.bevc.blacksburg.va.us> <6gb4ib$bdj$1@news.cs.tu-berlin.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit In comp.lang.objective-c Marcel Weiher wrote: > > [ ... idea to improve protocols .. ] Protocols are specific to the Next/GNU compilers so I can't really comment (the Stepstone compiler, for example, doesn't have protocols). But anyhow, the idea seems not related to Blocks, which somehow is in your subject line. Blocks are somewhat like a "callback" in X, or maybe what one used to do with the Next appkit with selectors that were sent to a delegate (if it responded to it). It's like a piece of code that has to be evaluated when something happens ... for example, you could set an action to set or undo highlighting of selection. [window ifBecomeActive: { [selection highlight]; }]; This is actually a good example to say why a Block is not a nested function (it can be evaluated after the function where it was created, returns).
Message-ID: <352E7160.27D54D75@mci.com> From: David Hinz <David.Hinz@mci.com> Organization: MCI MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Objective-C slow or am I doing something wrong? Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Fri, 10 Apr 1998 19:22:03 GMT NNTP-Posting-Date: Fri, 10 Apr 1998 15:22:03 EST I'm creating a class for writing an array of objects (actually NSStrings, NSNumbers, NSCalendarDates) to a file where each field is delimited by a specified character and the record delimiter can also be specified. The code works but it is extremely slow. Testing with 1000 records and 9 fields it takes about 1 1/2 minutes on a Solaris system (Sparc 20) I think. On a 133 MHz Intel (OpenStep 4.2) it takes about 1:45. Is objective-C just not suited for file I/O of this type? I realize that I could probably write this by wrapping fopen/fwrite/fclose but I wanted to use the NSFileHandle just to see if it would work. I don't think the locking is causing the problem since when I'm testing only one copy of the program is running. On a different object that reads a file using NSFileHandle the objective-C program takes about 20 seconds to read a file (I don't remember the size of the file) and Perl takes 0.9 seconds. I realize that there is a lot of memory management being done creating and destroying objects, but I didn't expect the program to run 20 times slower. Is anyone else using NSFileHandle? Can it be used efficiently? Thanks, dave. - (BOOL)addRecord:(NSArray *)recordArray { NSEnumerator *arrayEnumerator; id object; NSData *recordData; NSMutableString *recordString; unsigned int arrayCount = [recordArray count]; unsigned int numberColumns = 0; NSString *classString; NSRange stringRange; NSRange dateRange; // Number of items in array must equal number of columns. if ([recordArray count] != [columnCount unsignedIntValue]) { NSLog (@"Number of columns in record must equal number defined for file."); return (NO); } // Create a string to put the column data in. recordString = [NSMutableString stringWithCapacity:0]; // Create an enumerator for the array. arrayEnumerator = [recordArray objectEnumerator]; // Loop while there are more objects in the array. while ((object = [arrayEnumerator nextObject])) { // If the object being added is a string or date type, then added quotes around the object data. classString = [[object class] description]; stringRange = [classString rangeOfString:@"String"]; dateRange = [classString rangeOfString:@"Date"]; if ((stringRange.length != 0) || (dateRange.length != 0)) { [recordString appendFormat:@"\"%@\"", object]; } else { [recordString appendFormat:@"%@", object]; } // Increment the number of columns that have been appended. numberColumns++; // If the number of columns added is less then the total number of // columns then add a column delimiter. if (numberColumns < arrayCount) { [recordString appendString:columnDelimiter]; } } // Append the record delimiter. [recordString appendString:recordDelimiter]; // Create a data object from the string. recordData = [NSData dataWithBytes:[recordString cString] length:[recordString cStringLength]]; // Obtain the lock before trying to use the data file. // Just using NSDistributedLock if ([self lock]) { // Write the data to the file. [dataHandle writeData:recordData]; // Synchronize the file with what was written. Make sure that the order is recorded. // [dataHandle synchronizeFile]; // Unlock the file. [self unlock]; } return YES; } // end addRecord: -- ===================================================== = David Hinz MCI Telecommunications = = Internet and New Media Development = = Email: David.Hinz@MCI.com Phone: (303) 390-6108 = = Vnet: 636-6108 Fax: (303) 390-6365 = = Pager: 1-888-900-5732 (Interactive 2-way) = =====================================================
From: "Michelle L. Buck" <buck.erik@mcleod.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program -shareware Date: Sat, 11 Apr 1998 22:05:15 -0500 Organization: McleodUSA - http://www.mcleodusa.net Message-ID: <6gpb5t$hjo1@odie.mcleod.net> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggkg1$90k$1@news0-alterdial.uu.net> <SCOTT.98Apr8221520@slave.doubleu.com> <dcoshel-1104981530090001@323-a-43-57.ppp.mcleodusa.net> As a person who hires hot shot programmers, I can attest that the comments below is certainly true at my company. Programmers can tend to be fanatical people. The most brilliant programmers are people who can not rest until a good idea is realized. The best are the ones who can tell a good idea from a bad one in advance. Most programmers I know are passionately libertarian. This I believe comes from a strong belief in a meritocracy. Computers don't judge programmers, and software is still as much craft as science. If you write good software, you can certainly be judged by it. The real evil of "pointy haired bosses" is that they can not discern merit. Competitive start up companies are frantic places that often demand long work weeks and sometimes cause divorces and or nervous break downs. A person must be an entreprenour and not risk averse. These two qualities are not what most spouses are looking to find. On the other hand, the rewards are high; it is not really all that risky, and potential spouces find are large bank account very attractive. With regard to the type of jobs new hires get, tallent quickly rises to the top. The really hot programmers are easily identified after the fact. It is very hard to determine talent in an interview. As a result, any kind of prior work that can be evaluated is valuable. Grades just don't matter in this profession. I have worked with college drop outs who command 6 figure saleries and preside over small clustors of deciples who daily pay homage to the guru. By the way, musicians make the best programmers. I don't know why. I always ask candidates if they play an instrument. The best programmers are seldom the best managers. Managers must be very very knowledgable, stable, and current to avoid growing pointy hairs. One bad programmer costs the company low 6 figures. One bad manager can cost millions easily. Dave Oshel wrote in message ... >In article <SCOTT.98Apr8221520@slave.doubleu.com>, scott@doubleu.com >(Scott Hess) wrote: >[whack] > >> Given two new hires fresh out of college, one of which has a fair >> covey of praised freeware, the other with good grades, which gets the >> more important job? > >I've always wondered that. My one brief stint in the shrinkwrap software >biz kind of revealed that the guys (gals note below) who already have jobs >somewhere else get hired to do "important" stuff, while the new hires >fresh outa college work in q.a. or in intern slots. Once in q.a., always >in q.a. Interns have possibilities. > >Regardless of ability, the guys (the gals are guys too, in the sexless >company way of looking at things) are all unmarried "coders" who work 80 >hours a week 7 days a week and who "shower" naked at 6:30 AM in the >company toilet stalls with a huge bronze can of spray deodorant. The >married unmarried coders soon regain company-preferred status when their >divorces go through.
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <25462891752418@digifix.com> Date: 12 Apr 1998 03:50:19 GMT Organization: Digital Fix Development Message-ID: <5697892353623@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1993. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar Subject: cmsg cancel <8c35.13e64.32@tecra> Control: cancel <8c35.13e64.32@tecra> Date: 09 Apr 1998 05:39:22 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.8c35.13e64.32@tecra> Sender: Good Info Services <goodinfoservices@yahoo.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: "NeXT Newbie" <macghod@concentric.net> Newsgroups: comp.soft-sys.nextstep,comp.sys.mac.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Newby needs help Date: 12 Apr 1998 05:34:10 GMT Distribution: inet Message-ID: <01bd65d4$674529e0$17f0bfa8@davidsul> References: <01bc46e5$6b4880a0$24f0bfa8@davidsul> <6god24$lq4$1@news.rt66.com> Thanks, I have solved this problem. I merely needed to cp the files from the partition they were on (fat, from the dos side) to the os partition. Now I am working on getting ppp up and then internet apps so I can stop having to use windows 95 to get on the internet ;-) PS does firstclass have a unix client? Also what is a good book on openstep (or the unix underneath openstep) and administering openstep? Such things as ppp, sendmail, etc? Jake Garcia <jakeg@rt66.com> wrote in article <6god24$lq4$1@news.rt66.com>... > NeXT Newbie (macghod@concentric.net) wrote: > > : localhost:10# mv ne* next-ppp > : localhost:11# tar -xvf next-ppp > : ./PPP-2.2-0.4.6-pkg: File name too long > : tar: can't create ./PPP-2.2-0.4.6-pkg/: File name too long > : ./PPP-2.2-0.4.6-pkg: File name too long > : tar: can't create ./PPP-2.2-0.4.6-pkg/PPP-2.2.pkg/: File name too long > : ./PPP-2.2-0.4.6-pkg: File name too long > > Doesn't the standard Windows 95 filesystem limit file/dir names to 14 chars? > I could be wrong... > > -- > Jake Garcia <jake@midnight.rt66.com> > > Please finger for more info... >
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program -shareware Date: 12 Apr 1998 09:08:22 GMT Organization: WARPnet, Incorporated Message-ID: <6gq0a6$l7g$1@news.idiom.com> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggkg1$90k$1@news0-alterdial.uu.net> <SCOTT.98Apr8221520@slave.doubleu.com> <dcoshel-1104981530090001@323-a-43-57.ppp.mcleodusa.net> <6gpb5t$hjo1@odie.mcleod.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: buck.erik@mcleod.net "Michelle L. Buck" may or may not have said: -> As a person who hires hot shot programmers, I can attest that the comments -> below is certainly true at my company. -> -> Programmers can tend to be fanatical people. The most brilliant programmers -> are people who can not rest until a good idea is realized. The best are the -> ones who can tell a good idea from a bad one in advance. It also helps if they can articulate why one idea is good and other idea is bad, to non-technical managers. I've see many excellent developers nearly bust a gut from frustraton, when they knew the right thing to do, and couldn't convince people of less ability but more seniority. -> Most programmers I know are passionately libertarian. This I believe comes -> from a strong belief in a meritocracy. I concur, and I would add that I've seen this in other engineering disciplines, and the hard sciences. At its root, engineering is about facts: the bridge stands or it collapses, the server runs for months, or it behaves like NT. -> Computers don't judge programmers, and software is still as much craft as -> science. Pity we can't teach software development like blacksmithing or cabinetmaking: Show that you can write something that works, and you're a journeyman; write your masterpiece, and then you're qualifed to teach your own apprentices. -> If you write good software, you can certainly be judged by it. The real -> evil of "pointy haired bosses" is that they can not discern merit. Pointy haired bosses, and Venture Capitalists who will pass over something worthwhile to fund yet another java startup. [snippage] -> With regard to the type of jobs new hires get, tallent quickly rises to the -> top. It will rise in terms of responsibility, but whether the talent is rewarded financially depends on the competence of the organization. Here in the valley, engineers know that the way to get the raise you deserve is to change jobs. It's a lot easier to land a new job at 30% more than you're getting than to get a 30% raise. ->The really hot programmers are easily identified after the fact. It -> is very hard to determine talent in an interview. Yes and no. It helps if the interviewer is himself a highly qualified developer. I've always found that it's pretty easy to screen out the wannabees. [more snippage] -> By the way, musicians make the best programmers. I don't know why. I -> always ask candidates if they play an instrument. They're highly correlated, but I've seen a few counterexamples. ->The best programmers are seldom the best managers. And the best musicians are seldom the best music teachers. -> Managers must be very very knowledgable, stable, and current to avoid -> growing pointy hairs. One bad programmer costs the company low 6 -> figures. One bad manager can cost millions easily. It's quite possible to be a good, non-technical manager, and I've had a few in my time. The most common failing I see among non-technical managers is a tendency to compensate for their lack of understanding by micro-management. -jcr -- John C. Randolph (408) 358-6732 NeXT mail preferred. Chief Technology Officer, WARPnet Incorporated. "Have a lot of Fun, make a lot of Money, then sod off to Tahiti!"
From: David Stes <stes@mundivia.es> Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Blocks for Obj-C Date: Sun, 12 Apr 1998 09:46:19 +0000 Organization: Mundivia, Santander Message-ID: <35308D6B.41C67EA6@mundivia.es> References: <6fokj8$1dp$1@usenet50.supernews.com> <6g8r4c$2uq$1@crib.bevc.blacksburg.va.us> <6gb4ib$bdj$1@news.cs.tu-berlin.de> <352FB748.41C67EA6@mundivia.es> <6gof86$gkk$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Nathan Urban wrote: > > ... blocks (a.k.a. "closures", "lambda" or "anonymous functions" ) ... Or "anonymous methods". I've been working a while ago on Blocks for Objective-C (see http://www.can.nl/~stes/block98.html) and I like the description "anonymous method". For example, you could have a function to count the number of matches : unsigned cntmatch(id c,id a) { unsigned cnt = 0; [c do:{ :what | if (what == a) cnt++; }]; return cnt; } The :what syntax is similar to the method argument syntax "+foo:what" , except that the "foo" keyword selector is absent (the method is "anonymous" or "unnamed") and only the colon remains.
From: "Hendrik Merx" <merx@pc.chemie.tu-darmstadt.de> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 12 Apr 98 15:08:32 +0200 Organization: Technische Universitaet Darmstadt Message-ID: <B1568976-2BA0A@130.83.138.138> References: <joe.ragosta-0904981624100001@wil59.dol.net> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit nntp://news.tu-darmstadt.de/comp.sys.mac.oop.powerplant, nntp://news.tu-darmstadt.de/comp.sys.mac.programmer.codewarrior, nntp://news.tu-darmstadt.de/comp.sys.next.advocacy, nntp://news.tu-darmstadt.de/comp.sys.next.programmer On Thu, Apr 09, 1998 22:24, Joe Ragosta <mailto:joe.ragosta@dol.net> wrote: >In article <B1526CE7-65D6E@207.217.155.85>, "Brad Hutchings" ><brad@hutchings-software.com> wrote: > >> This move would not have been met with intense cynicism if developers >> thought Apple had an overall clue and direction and appreciation of how >> developers support Apple. It starts from the top, and unfortunately, is >> beyond the control of ADR. > >So, in other words, you're just assuming that, because it's Apple, they >did the wrong thing. I think he suggested that Apple's engineers might have a different opinion about this topic than Apple's marketing and/or management. Hendrik merx@pc.chemie.tu-darmstadt.de
From: "Hendrik Merx" <merx@pc.chemie.tu-darmstadt.de> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 12 Apr 98 15:46:58 +0200 Organization: Technische Universitaet Darmstadt Message-ID: <B1569277-4D79F@130.83.138.138> References: <don_arb-1004981207170001@sea-ts1-p29.wolfenet.com> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="Cyberdog-MixedBoundary-0004D66F" Content-Transfer-Encoding: 7bit nntp://news.tu-darmstadt.de/comp.sys.mac.oop.powerplant, nntp://news.tu-darmstadt.de/comp.sys.mac.programmer.codewarrior, nntp://news.tu-darmstadt.de/comp.sys.next.advocacy, nntp://news.tu-darmstadt.de/comp.sys.next.programmer --Cyberdog-MixedBoundary-0004D66F X-Fontfamily: Geneva X-Fontsize: 9 Content-Type: text/enriched; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable <SMALLER><SMALLER>On Fri, Apr 10, 1998 21:07, </SMALLER></SMALLER> --Cyberdog-MixedBoundary-0004D66F Content-Type: application/X-url Content-Transfer-Encoding: base64 Content-Description: Don Arbow bWFpbHRvOmRvbl9hcmJAd29sZmVuZXQuY29t --Cyberdog-MixedBoundary-0004D66F X-Fontfamily: Geneva X-Fontsize: 9 Content-Type: text/enriched; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable <SMALLER><SMALLER> wrote:</SMALLER></SMALLER><SMALLER><SMALLER> </SMALLER></SMALLER><SMALLER><SMALLER>>Remember, also that Apple needed people to administer the hardware >program, take orders, get the boxes shipped, etc. I'm sure that cut into >any profits that Apple may have made on the boxes. Don't they have this Online Store thing? There should be enough people doing exactly that. Hendrik merx@pc.chemie.tu-darmstadt.de</SMALLER></SMALLER> --Cyberdog-MixedBoundary-0004D66F--
From: "Hendrik Merx" <merx@pc.chemie.tu-darmstadt.de> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 12 Apr 98 15:50:10 +0200 Organization: Technische Universitaet Darmstadt Message-ID: <B1569336-50497@130.83.138.138> References: <don_arb-1004981219030001@sea-ts1-p29.wolfenet.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit nntp://news.tu-darmstadt.de/comp.sys.mac.oop.powerplant, nntp://news.tu-darmstadt.de/comp.sys.mac.programmer.codewarrior, nntp://news.tu-darmstadt.de/comp.sys.next.advocacy, nntp://news.tu-darmstadt.de/comp.sys.next.programmer On Fri, Apr 10, 1998 21:19, Don Arbow <mailto:don_arb@wolfenet.com> wrote: >Well, since DR1 is currently under NDA, the only place to get your >questions answered are from Apple itself. You cannot post questions in >public forums about software which you have agreed to Apple that you will >not discuss. But he may expect some clarification on the documentation. Hendrik merx@pc.chemie.tu-darmstadt.de
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Followup-To: comp.sys.mac.advocacy,comp.sys.next.advocacy Date: 12 Apr 1998 15:07:16 GMT Organization: P & L Systems Message-ID: <6gqlb4$oop$26@ironhorse.plsys.co.uk> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggiu7$auu$1@vixen.cso.uiuc.edu> <352c857a.2678525846@news.u-psud.fr> <joe.ragosta-0904980645050001@elk33.dol.net> <ericb-0904981707160001@132.236.171.104> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: ericb@pobox.com NOTE: FOLLOWUPS TO COMP.SYS.MAC.ADVOCACY,COMP.SYS.NEXT.ADVOCACY In <ericb-0904981707160001@132.236.171.104> Eric Bennett wrote: > In article <joe.ragosta-0904980645050001@elk33.dol.net>, > joe.ragosta@dol.net (Joe Ragosta) wrote: > > I can't download Rhapsody. Not getting Rhapsody is what really burns me. > I am enthusiastic about it and I want to start learning to program on it. > MacOS 8.2 is not such a big deal, since it's not really all that different > from a basic programming perspective that the copy of MacOS 8.1 I'm > running now. Not so with Rhapsody. > So what's been stopping you getting hold of the academic version of OPENSTEP 4.2 ($299)? Note also that Apple has just announced WebObjects Academic for **$99* -- that gives you pretty much all the YellowBox API... Best wishes, mmalc.
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Followup-To: comp.sys.mac.advocacy,comp.sys.next.advocacy Date: 12 Apr 1998 15:12:27 GMT Organization: P & L Systems Message-ID: <6gqlkr$oop$27@ironhorse.plsys.co.uk> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggrql$oop$1@ironhorse.plsys.co.uk> <ericb-0904981705120001@132.236.171.104> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: ericb@pobox.com NOTE: FOLLOWUPS TO COMP.SYS.MAC.ADVOCACY,COMP.SYS.NEXT.ADVOCACY In <ericb-0904981705120001@132.236.171.104> Eric Bennett wrote: > In article <6ggrql$oop$1@ironhorse.plsys.co.uk>, mmalcolm crawford > <malcolm@plsys.co.uk> wrote: > > > Apple has said that the acedemic program is to be announced later -- I'd > > expect it to be cheaper. > > Where have they explicitly stated this? Do you have a URL? All I have > seen are rumors from "highly placed sources." And those rumors have been > around for quite awhile without having come true. > Most recently: Subject: Re: Apple Developer Connection Mailing Date: Wed, 8 Apr 1998 09:37:36 -0700 From: Jordan Dea-Mattson <jordan@apple.com> To: "Rhapsody Discussion List" <rhapsody@clio.lyris.net> [...] In addition, we are working to roll out programs for the academic world - and have said so for some time. They weren't ready and this time and I can't comment on them, except to say that we are aware of the need to reach the academic world. Yours, Jordan Jordan J. Dea-Mattson Senior Partnership & Technology Solutions Manager Apple Developer Relations --- Moreover, I again refer you to the "WebObjects/academic for $99" announcement. Best wishes, mmalc.
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 12 Apr 1998 15:17:07 GMT Organization: P & L Systems Message-ID: <6gqltj$oop$28@ironhorse.plsys.co.uk> References: <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <6ggvub$oop$5@ironhorse.plsys.co.uk> <6gitbu$cg1$1@interport.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: float@interport.net In <6gitbu$cg1$1@interport.net> float@interport.net wrote: > mmalcolm crawford (malcolm@plsys.co.uk) wrote: > : In <B1511019-15903@206.165.43.154> "Chicken Little" wrote: > > : > Trying to make developer *support* a profit center is an INSANE idea. > : > > : Apple acting as a charity subsidising people's hobbies is what's insane. > : Nor should they be offering a program wide open to abuse by people signing up > : as developers just to get a discount on hardware. > > The dynamics of platform adoption don't work the way you think. > What on earth did my post have to do with platform adoption?! > It doesn't matter why someone signs on to the program: every sign-on is a win > for Apple, every hardware sale, discounted or not, is a win for Apple, > every developer, end-user, or hobbyist who gets on the Rhapsody bandwagon > is a win for Apple. > How many people sign onto the developer programs *in total*? 10,000? An extra couple here or there are likely to be lost in the noise. mmalc.
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Followup-To: comp.sys.mac.advocacy,comp.sys.next.advocacy Date: 12 Apr 1998 15:05:14 GMT Organization: P & L Systems Message-ID: <6gql7a$oop$25@ironhorse.plsys.co.uk> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <6ggiu7$auu$1@vixen.cso.uiuc.edu> <352c857a.2678525846@news.u-psud.fr> <joe.ragosta-0904980645050001@elk33.dol.net> <6giria$nng$1@ifaedi.insa-lyon.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: nospam@spam.com NOTE: FOLLOWUPS TO COMP.SYS.MAC.ADVOCACY,COMP.SYS.NEXT.ADVOCACY In <6giria$nng$1@ifaedi.insa-lyon.fr> Bluedays wrote: > > Of course, Mrs I know everything, do you really know how are the campus > in europe ? Do you really think that every campus has a internet connexion ? > I'd be very surprised if every campus dod not now have an internet connection of some description. > with free access to zip drive and CD burner ? > Sadly you're right on this one. > If you read the posts today about the problem, you'll see that a lot > of complains come from europe, we don't need the free pass > for the WWDC because of the flight price. > So don't buy into the Premier program. > Even with a T1 the > bandwitdh with the US is bad...1K/s and less sometimes (often) > And if you think that now this will be the only way to get > information for many people if you don't have the CD... > Macintosh prices are higher in some places in europe than > in US, so the discount lost is not a real good news. > Umm, in which case what's wrong with the Developer Mailing? Which is $50 *cheaper* than the old program? Unless, of course, you're academic, in which case I'd expect the program Apple offers you soon to be even cheaper. Best wishes, mmalc.
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Followup-To: comp.sys.mac.advocacy,comp.sys.next.advocacy Date: 12 Apr 1998 15:29:22 GMT Organization: P & L Systems Message-ID: <6gqmki$oop$31@ironhorse.plsys.co.uk> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6ge6hh$9mv@netaxs.com> <352BA869.4901@home.sleeping> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: me@home.sleeping NOTE: FOLLOWUPS TO COMP.SYS.MAC.ADVOCACY,COMP.SYS.NEXT.ADVOCACY In <352BA869.4901@home.sleeping> Jeff wrote: > - Yeah, sucks, doesn't it? All I really want is the CDs, and they keep > - upping the price and adding things I don't need. > - > So what's wrong with the Developer Mailing? http://developer.apple.com/programs/mailing.html At $200 it seems to give you exactly what you want, and is $50 *cheaper* than the previous program... mmalc.
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Followup-To: comp.sys.mac.advocacy,comp.sys.next.advocacy Date: 12 Apr 1998 15:23:28 GMT Organization: P & L Systems Message-ID: <6gqm9g$oop$29@ironhorse.plsys.co.uk> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <trumbull-0904980039090001@net44-223.student.yale.edu> <6gio5m$oop$11@ironhorse.plsys.co.uk> <trumbull-0904981544360001@net44-223.student.yale.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: trumbull@cs.yale.edu NOTE: FOLLOWUPS TO COMP.SYS.MAC.ADVOCACY,COMP.SYS.NEXT.ADVOCACY In <trumbull-0904981544360001@net44-223.student.yale.edu> Ben Trumbull wrote: > In article <6gio5m$oop$11@ironhorse.plsys.co.uk>, mmalcolm crawford > <malcolm@plsys.co.uk> wrote: > > > No, I don't. I'm quite sure there are a number of hobbyists developing > > worthwhile s/w which I might find useful at some point, however on balance > > reports suggest that the previous system was badly abused and on average > > Apple will have lost revenue through people signing up as developers simply > > to get hardware at reduced prices. > > Sorry, but asserting that Apple doesn't make money off the books they sell > to Barnes&Noble and the rest of the Addison&Wesley series to "hobbyist" > developers is absurd. > Where did I make that claim?! Please refrain from putting words into my mouth -- it is most distasteful. > As for the "badly abused" hardware discount program, that's a load of > baloney. I saw the prices and was never once tempted. > So why have so many people been screaming about the fact that the hardware discount scheme is no longer as accessible? > Apple made money > off every one of those machines and they sure made a lot more money than > if some retailer sold the developer that machine instead of Apple. Which > is exactly what will happen now. Devs will pay retailers for machines, > devs will buy fewer machines overall, and Apple will see smaller margins > overall. Yeah, this is a great solution to Apple's "problem" of selling > more machines and not paying retailers for the priviledge. Let's get this > straight. Apple makes A LOT MORE MONEY when they sell direct. A lot. > Figures? Until I see figures proving it one way ot the other, each of us is untitled to our own *opinion*. > The ideas that Apple should "raise the bar" to select against small > developers and this will somehow make the Mac dev community healthier, or > that Apple EVER was "subsidizing" small or independent developers are > absurd. If Apple lost money on selling tons of development > documentations, cds, software, and hardware (true for less, but no > retailer biting into the margins), then frankly the sole reason is because > they couldn't manage their business. > So now they're managing their business better, they're doing what you people have been demanding of them for years, and now you're whingeing about it. > Trying to blame people for "abusing" > the dev program is just a bullshit excuse. However, to be fair to Apple, > I haven't actually seen anyone from Apple make that argument, just people > desperate to defend the policy changes. > It's *one* *reason* why *I* think they may have changed the Program. There are plenty of others. mmalc.
From: michal@gortel.phys.ualberta.ca (Michal Jaegermann) Newsgroups: comp.soft-sys.nextstep,comp.sys.mac.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Newby needs help Followup-To: comp.soft-sys.nextstep,comp.sys.mac.advocacy,comp.sys.next.misc,comp.sys.next.programmer Date: 12 Apr 1998 16:16:02 GMT Organization: Disorganized Bits Message-ID: <6gqpc2$tb0$1@pulp.ucs.ualberta.ca> References: <01bc46e5$6b4880a0$24f0bfa8@davidsul> <6god24$lq4$1@news.rt66.com> <3530c496.8026484@198.0.0.100> Nathan Hughes (nhughes@sunflower.com) wrote: : : In reality, Win95 sits on a dos partition which is limited to 8.3 : filenames. 95 works around this limitation, but Unix does not : (generally). 'mtools', which work across most (all?) Unix platforms support vfat extensions for quite a long time. I did have a need to hack the latest versions of 'mtools' on NeXT but I have a little doubt that I would have them working in no time at all. : Their are utilities for Linux that allow it to see LFNs : but it isn't part of the distribution, This is also false for years now. It is true that you do not have to include vfat support in your kernel (either directly or as a module) but this is a part of a standard kernel for a long time. Also 'mtools' are normally included in a distribution. "Official" Linux kernel does not have support for FAT32 format and Jolliet extensions but these are available as outside patches and are present in development kernels and will be included in the next 2.2.x kernels. --mj
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Blocks for Obj-C Date: 12 Apr 1998 16:49:24 GMT Organization: Technical University of Berlin, Germany Message-ID: <6gqrak$af8$1@news.cs.tu-berlin.de> References: <6fokj8$1dp$1@usenet50.supernews.com> <6g8r4c$2uq$1@crib.bevc.blacksburg.va.us> <6gb4ib$bdj$1@news.cs.tu-berlin.de> <352FB748.41C67EA6@mundivia.es> <6gof86$gkk$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit nurban@crib.bevc.blacksburg.va.us (Nathan Urban) writes: >In article <352FB748.41C67EA6@mundivia.es>, David Stes <stes@mundivia.es> wrote: >> In comp.lang.objective-c Marcel Weiher wrote: >> > [ ... idea to improve protocols .. ] >> But anyhow, the idea seems not related to Blocks, which somehow is in >> your subject line. >> [description of blocks] >Yeah, I think those of us involved in this thread know what blocks are >(though I'm speaking only for myself).. the thread was inaptly named.. >it started out when someone said they wanted blocks (I think), and I >replied that Obj-C would be the perfect language if it had blocks plus >this protocol category feature we've been discussing, and then no one >bothered to ever change the subject line. Yup. >I really wish the Apple compiler had blocks (a.k.a. "closures", "lambda" >or "anonymous functions"). I really want to use them in my Rhapsody >development. They're hard to appreciate at first by some ("why don't >you just make a new method?"), but once you get used to them, you hate >to live without them. I am not sure that true Smalltalk-like block can be adequately supported by a 'hack' hybrid like Objective-C (still the most wonderful hack I've seen). The protocol improvement however would require only a tiny change that wouldn't really be a new feature but more an orthogonalization of present features. As to blocks, I *love* the functor idea of taking a function and applying it to whole data-sets (FP, APL). However, in object-oriented systems we aren't really supposed to think in terms of functions (or blocks of procedural code), but in terms of messages. That's why I prefer what I call 'filters' (probably not a good name, but I had to call it something). Filters are really just a generalization of the 'perform:with...' methods. The difference is that they return values based on the computations performed on each element. There are three basic filters: filter - returns the result of sending the message select - returns the receiver-object if the message returns YES reject - returns the receiver-object if the message returns NO So you can say something like: classes = [myObjects filterWithSelector:@selector(class)]; to return an array of the classes of all the objects in the array (or enumerator) myObjects. Actually, a little hackery with proxies and forwarding allows me to say classes = [[myObjects filter] class]; (Again, I have to find a better name for the method, filterContents might be better) Filters can also be nested classDescriptions = [[[[myObjects startFilter] class] description] runFilter] The array does not have to be the target, so you could also say: values = [[myDictionary filter] objectForKey: [NSArray arrayWithObjects:@"key",@"key2",nil]]; This is much more restrictive than closures or blocks, but I think it is much closer to the spirit of OO systems than those facilities. Due to its restricted nature, it would also be easy to parallelize. It also seems to be working fairly well, at least for us. I hardly spend any time writing loops nowadays. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Blocks for Obj-C Date: 12 Apr 1998 16:54:20 GMT Organization: Technical University of Berlin, Germany Message-ID: <6gqrjs$anh$1@news.cs.tu-berlin.de> References: <6fokj8$1dp$1@usenet50.supernews.com> <6g8r4c$2uq$1@crib.bevc.blacksburg.va.us> <6gb4ib$bdj$1@news.cs.tu-berlin.de> <352FB748.41C67EA6@mundivia.es> <6gof86$gkk$1@crib.bevc.blacksburg.va.us> <35308D6B.41C67EA6@mundivia.es> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit David Stes <stes@mundivia.es> writes: >Nathan Urban wrote: >> >> ... blocks (a.k.a. "closures", "lambda" or "anonymous functions" ) ... >Or "anonymous methods". >I've been working a while ago on Blocks for Objective-C (see >http://www.can.nl/~stes/block98.html) and I like the description >"anonymous method". >For example, you could have a function to count the number of matches : > unsigned cntmatch(id c,id a) > { > unsigned cnt = 0; > [c do:{ :what | if (what == a) cnt++; }]; > return cnt; > } >The :what syntax is similar to the method argument syntax "+foo:what" , >except that the "foo" keyword selector is absent (the method is >"anonymous" or "unnamed") and only the colon remains. Doing this with filters: return [[c selectWithSelector:@selector(isEqual:) withObject:a] count]; (changing the '==' test to the isEqual: message send) In fact, I'd rather have return [[[c select] isEqual:a] count]; but Next/Apple decided to make BOOL a char, so the intermediate result will get garbled by the compiler. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: David Stes <stes@mundivia.es> Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Blocks for Obj-C Date: Sun, 12 Apr 1998 20:36:52 +0000 Organization: Mundivia, Santander Message-ID: <353125E4.41C67EA6@mundivia.es> References: <6fokj8$1dp$1@usenet50.supernews.com> <6g8r4c$2uq$1@crib.bevc.blacksburg.va.us> <6gb4ib$bdj$1@news.cs.tu-berlin.de> <352FB748.41C67EA6@mundivia.es> <6gof86$gkk$1@crib.bevc.blacksburg.va.us> <35308D6B.41C67EA6@mundivia.es> <6gqrjs$anh$1@news.cs.tu-berlin.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Marcel Weiher wrote: > > > unsigned cnt = 0; > > [c do:{ :what | if (what == a) cnt++; }]; > > return cnt; > > Doing this with filters: > > return [[c selectWithSelector:@selector(isEqual:) withObject:a] count]; > > (changing the '==' test to the isEqual: message send) But to really write the (exact) equivalent, you can't make that change, because for example, newc = [OrdCltn new]; [c do:{ :each | if (each != myObject) [newc add:each]; }]; and newc = [OrdCltn new]; [c do:{ :each | if (![each isEqual:myObject]) [newc add:each]; }]; are not the same thing ... Sometimes you want object identifier equality, sometimes you want to send a -isEqual: message. Well, one could argue that one would better always use -isEqual: . But the issue still remains that with Blocks, you could easily write it either way ( == or -isEqual:).
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <120498060915@egg.com> Control: cancel <120498060915@egg.com> Date: 11 Apr 1998 22:13:42 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.120498060915@egg.com> Sender: easter@egg.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: gutkneco+news@lirmm.fr (Olivier Gutknecht) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Mon, 13 Apr 1998 01:10:50 +0100 Organization: Elevage d'agents nourris au grain Message-ID: <1d7efs4.14d33ptutamabN@ppgutkneco.lirmm.fr> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ylaporte-0804982005030001@139.103.176.79> <6gh9cm$oop$7@ironhorse.plsys.co.uk> <joe.ragosta-0904980643360001@elk33.dol.net> Joe Ragosta <joe.ragosta@dol.net> wrote: > Jordan later commented on RDL that Apple is trying to make the educational > program very attractive. Sure, but that is not an official word from Apple... Another problem about the forthcoming (?) edu programs is about non-US educational/research communities. I hope that worldwide educational institutions will be able to enroll. Given the current status of European Developer Relations, I am a little bit dubious. Ol. -- Olivier Gutknecht ... Laboratoire d'Informatique, Robotique gutkneco@lirmm.fr ... et Micro-Electronique de Montpellier.
From: Web Surfer <surfer@www.web> Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.programmer Subject: Re: first day with OS Date: Sun, 12 Apr 1998 15:37:58 -0400 Organization: WWW Message-ID: <35311816.6449@www.web> References: <01bc462d$87ca8380$40f0bfa8@davidsul> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: NeXT Newbie <macghod@concentric.net> What OS is that? OS/2? The old NeXT OS? What? NeXT Newbie wrote: > > Some impressions of OS 4.2 > 1) This OS sure was designed for big monitors! I have a 17 inch and it > seems to small. Seems like you would want at least a 20 inch monitor > 2) The mouse really annoys me. It reminds me of when I first used windows > 3.1, I was like god I hate how jerky the mousee is. The mouse movement in > OS 4.2 seems like how it was in windows 3.1, windows 95 definitely seems > much better. > 3) How do you connect to the internet with ppp? > 4) once I get up and going with ppp, what do I use for usenet? > 5) where is the option to change the monitors resolution and bits of color?
From: Web Surfer <surfer@www.web> Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.programmer Subject: Re: first day with OS Date: Sun, 12 Apr 1998 15:38:14 -0400 Organization: WWW Message-ID: <35311826.7AD7@www.web> References: <01bc462d$87ca8380$40f0bfa8@davidsul> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit What OS is that? OS/2? The old NeXT OS? What? NeXT Newbie wrote: > > Some impressions of OS 4.2 > 1) This OS sure was designed for big monitors! I have a 17 inch and it > seems to small. Seems like you would want at least a 20 inch monitor > 2) The mouse really annoys me. It reminds me of when I first used windows > 3.1, I was like god I hate how jerky the mousee is. The mouse movement in > OS 4.2 seems like how it was in windows 3.1, windows 95 definitely seems > much better. > 3) How do you connect to the internet with ppp? > 4) once I get up and going with ppp, what do I use for usenet? > 5) where is the option to change the monitors resolution and bits of color?
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Blocks for Obj-C Date: 13 Apr 1998 13:39:21 GMT Organization: Technical University of Berlin, Germany Message-ID: <6gt4i9$835$1@news.cs.tu-berlin.de> References: <6fokj8$1dp$1@usenet50.supernews.com> <6g8r4c$2uq$1@crib.bevc.blacksburg.va.us> <6gb4ib$bdj$1@news.cs.tu-berlin.de> <352FB748.41C67EA6@mundivia.es> <6gof86$gkk$1@crib.bevc.blacksburg.va.us> <35308D6B.41C67EA6@mundivia.es> <6gqrjs$anh$1@news.cs.tu-berlin.de> <353125E4.41C67EA6@mundivia.es> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit David Stes <stes@mundivia.es> writes: >Marcel Weiher wrote: >> >> > unsigned cnt = 0; >> > [c do:{ :what | if (what == a) cnt++; }]; >> > return cnt; >> >> Doing this with filters: >> >> return [[c selectWithSelector:@selector(isEqual:) withObject:a] count]; >> >> (changing the '==' test to the isEqual: message send) >But to really write the (exact) equivalent, you can't make that change, >because for example, I wasn't trying to write an exact equivalent, just a similar enough problem to > newc = [OrdCltn new]; > [c do:{ :each | if (each != myObject) [newc add:each]; }]; > >and > newc = [OrdCltn new]; > [c do:{ :each | if (![each isEqual:myObject]) [newc add:each]; }]; >are not the same thing ... Of course they aren't, which is why I mentioned the change. However, I'd almost always prefer sending objects messages to operating on them directly. >Sometimes you want object identifier equality, sometimes you want to >send a -isEqual: message. >Well, one could argue that one would better always use -isEqual: . But >the issue still remains that with Blocks, you could easily write it >either way ( == or -isEqual:). The problem I have with blocks is probably exactly what you like about them, the fact that they are ultimately flexible because they can contain arbitrary procedural code to work on each element. The code to deal with individual elements is still visible, has to be written, read and maintained. What I like about filters is that they operate at a higher level of abstraction because the code for dealing with individual elements is hidden. For me, this has meant that I actually _think_ about collections of objects operated on simultaneously, instead of piece by piece. When I need to preserve state I use other constructs such as streams or a plain C loop. Anyway, testing for object identity is not THAT difficult: @implementation NSObject(identity) -isIdenticalTo:otherObject { return self==otherObject; } @end return [[c selectWithSelector:@selector(isIdenticalTo:) withObject:a] count]; Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Blocks for Obj-C Date: 13 Apr 1998 11:05:49 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6gt9kd$o9o$1@crib.bevc.blacksburg.va.us> References: <6fokj8$1dp$1@usenet50.supernews.com> <6gqrjs$anh$1@news.cs.tu-berlin.de> <353125E4.41C67EA6@mundivia.es> <6gt4i9$835$1@news.cs.tu-berlin.de> In article <6gt4i9$835$1@news.cs.tu-berlin.de>, marcel@cs.tu-berlin.de (Marcel Weiher) wrote: > The problem I have with blocks is probably exactly what you [David > Stes] like about them, the fact that they are ultimately flexible because > they can contain arbitrary procedural code to work on each element. > The code to deal with individual elements is still visible, has to > be written, read and maintained. What I like about filters is > that they operate at a higher level of abstraction because the > code for dealing with individual elements is hidden. The reason I like blocks is because they're sufficiently general that they often keep me from having to make new methods for little one-off routines only used by a single method. The problem with filters is that you need a selector to apply. Usually what I need done can't be done with any of the standard selectors, so I have to write a custom one. This is annoying when its implementation is only a line or two and is only used by a single method.
From: sal@panix3.panix.com (Salvatore Denaro) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 13 Apr 1998 16:23:58 GMT Organization: PANIX Public Access Internet and Unix, NYC Message-ID: <slrn6j4f0u.juc.sal@panix3.panix.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104><joe.ragosta-0804980858150001@wil77.dol.net><mazulauf-0804981503390001@sneezy.met.utah.edu><joe.ragosta-0804981906460001@elk68.dol.net><01bc4540$d4a15e60$3af0bfa8@davidsul><6gh8v1$oop$6@ironhorse.plsys.co.uk> <SCOTT.98Apr9091856@slave.doubleu.com> <01bc46e4$1815c920$24f0bfa8@davidsul> On 10 Apr 1998 18:52:40 GMT, NeXT Newbie <macghod@concentric.net> wrote: >Plus, now Apple is using the microsoft developer pricing model as its own. >Oops, but microsoft has %95 of the market and apple only has %3. Oops, >Microsoft has a monopoly so they can do this. I guess Apple is now "just >as good" as microsoft <big grin> > >Except they dont have a monopoly and are figting for their lifes, so they >cant use monopolistic pricing like microsoft does. > >Clue phone for Mr Jobs, Mr Jobs please pick up the clue phone. Mr Jobs, >Apple is not Microsoft, Microsoft can charge those prices because they are >a monopoly, Apple is fighting for its life so it cant treat developers so >shabbily Anyone with a Mac can download MPW and the SDKs and write a Mac App for $0.00, that is not the case with Windows. If you want to write a win app that can qualify for the winlogo you need to buy a compiler. The $495 MSDN subscription does not include tools, the $500 Apple dev sub does. The $195 MSDN librady rel does not include OS versions, the $199 Apple one does. The Apple program is well within the industry norms. Why are Apple users so damn whiney? It sounds like you bitch and moan over every little thing. If the $199 price includes Rhapsody CR1, then it is a bargin. For $199 you get something that cost $5k just over two years ago. I can not imagine that many freeware developers paid the $250 for the old program, or that many will pay the $199 or $500 for the new programs. I fail to see what the big deal is. -- sal@panix.com Salvatore Denaro "The reality of the software business today is that if you find something that can make you ridiculously rich, then that's something that Microsoft is going to take away from you." -- Max Metral
From: sal@panix3.panix.com (Salvatore Denaro) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 13 Apr 1998 16:37:00 GMT Organization: PANIX Public Access Internet and Unix, NYC Message-ID: <slrn6j4fpc.juc.sal@panix3.panix.com> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <ericb-0904981701440001@132.236.171.104> On Thu, 09 Apr 1998 17:01:44 -0400, Eric Bennett <ericb@pobox.com> wrote: >Plenty of commercial or shareware apps had humble freeware beginnings. >Other good software has always been freeware. Why charge these people >$500 to get their hands on the latest system software? Do you want them The $199 version gets you the latest OS versions (but not Betas) >all to tell users, "Sorry, my program doesn't work with OS 8.2, because >Apple prices the developer versions out of my price range. I can't afford >to test my freeware with 8.2 until *after* it comes out, and then start >working on any necessary fixes." $250 was well outside my budget in my .edu days. That was a whole semester of beer*. I bought a copy of Think pascal at the book store for $50 or so and that was it. That was my total software expenditure for my college career. 95%+ of my fellow students just swiped software from the student computer labs. (*) $10 bring your own Mug nights... :) -- sal@panix.com Salvatore Denaro "The reality of the software business today is that if you find something that can make you ridiculously rich, then that's something that Microsoft is going to take away from you." -- Max Metral
From: David Stes <stes@mundivia.es> Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Blocks for Obj-C Date: Mon, 13 Apr 1998 16:45:47 +0000 Organization: Mundivia, Santander Message-ID: <3532413B.41C67EA6@mundivia.es> References: <6fokj8$1dp$1@usenet50.supernews.com> <6gqrjs$anh$1@news.cs.tu-berlin.de> <353125E4.41C67EA6@mundivia.es> <6gt4i9$835$1@news.cs.tu-berlin.de> <6gt9kd$o9o$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Nathan Urban wrote: > > In article <6gt4i9$835$1@news.cs.tu-berlin.de>, marcel@cs.tu-berlin.de (Marcel Weiher) wrote: > > > The problem I have with blocks is probably exactly what you [David > > Stes] like about them, the fact that they are ultimately flexible because > > they can contain arbitrary procedural code to work on each element. > > > The code to deal with individual elements is still visible, has to > > be written, read and maintained. What I like about filters is > > that they operate at a higher level of abstraction because the > > code for dealing with individual elements is hidden. > > The reason I like blocks is because they're sufficiently general that > they often keep me from having to make new methods for little one-off > routines only used by a single method. The problem with filters is that > you need a selector to apply. Usually what I need done can't be done > with any of the standard selectors, so I have to write a custom one. > This is annoying when its implementation is only a line or two and is > only used by a single method. I agree, and, in addition, it is easy to make variables of all sorts of datatypes (int, BOOL ,char*) etc. available to a Block. For example, findstr(char *s) { [c do:{ :whatever | if (!strcmp([whatever name],s) ... }]; } where "s" is a parameter of type char*. It is accessible from within the Block (and it can even modify the variable). With a -perform:with: type of approach, you will have to cast the char* to an id to pass it to the selector which does the comparison. This is ugly and sometimes can't be done. One may think that it is an inherent difficulty with Objective-C, but I think it rather just depends on using -perform:with: for this sort of thing. There's no difficulty for a Block (as opposed to -perform:with:) to access data that is of type BOOL, int, double, id, char* etc., from within the Block. In fact the compiler promotes all these variables to compiler generated structures that are then accessible from within the Block. Without compiler support, I think that all this communication (e.g. making the parameter "s" available) between the comparison method and the method that iterates over the collection, can only be easily done by using global variables (which is not a good solution, since then things are no longer reentrant).
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Blocks for Obj-C Date: 13 Apr 1998 17:11:17 GMT Organization: Technical University of Berlin, Germany Message-ID: <6gtgvl$ctq$1@news.cs.tu-berlin.de> References: <6fokj8$1dp$1@usenet50.supernews.com> <6gqrjs$anh$1@news.cs.tu-berlin.de> <353125E4.41C67EA6@mundivia.es> <6gt4i9$835$1@news.cs.tu-berlin.de> <6gt9kd$o9o$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit nurban@crib.bevc.blacksburg.va.us (Nathan Urban) writes: >In article <6gt4i9$835$1@news.cs.tu-berlin.de>, marcel@cs.tu-berlin.de (Marcel Weiher) wrote: >> The problem I have with blocks is probably exactly what you [David >> Stes] like about them, the fact that they are ultimately flexible because >> they can contain arbitrary procedural code to work on each element. >> The code to deal with individual elements is still visible, has to >> be written, read and maintained. What I like about filters is >> that they operate at a higher level of abstraction because the >> code for dealing with individual elements is hidden. >The reason I like blocks is because they're sufficiently general that >they often keep me from having to make new methods for little one-off >routines only used by a single method. The problem with filters is that >you need a selector to apply. Usually what I need done can't be done >with any of the standard selectors, so I have to write a custom one. >This is annoying when its implementation is only a line or two and is >only used by a single method. It is true that I sometimes have to write one-off methods, but this happens only very rarely (<10% of the time). More than 80% of the time the problem can easily be cast into filter form. Sometimes I have to write a conventional loop. You mileage obviously seems to vary. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Blocks for Obj-C Date: 13 Apr 1998 18:12:40 GMT Organization: Technical University of Berlin, Germany Message-ID: <6gtkio$hh7$1@news.cs.tu-berlin.de> References: <6fokj8$1dp$1@usenet50.supernews.com> <6gqrjs$anh$1@news.cs.tu-berlin.de> <353125E4.41C67EA6@mundivia.es> <6gt4i9$835$1@news.cs.tu-berlin.de> <6gt9kd$o9o$1@crib.bevc.blacksburg.va.us> <3532413B.41C67EA6@mundivia.es> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit David Stes <stes@mundivia.es> writes: >Nathan Urban wrote: >> >> In article <6gt4i9$835$1@news.cs.tu-berlin.de>, marcel@cs.tu-berlin.de (Marcel Weiher) wrote: >> >> > The problem I have with blocks is probably exactly what you [David >> > Stes] like about them, the fact that they are ultimately flexible because >> > they can contain arbitrary procedural code to work on each element. >> >> > The code to deal with individual elements is still visible, has to >> > be written, read and maintained. What I like about filters is >> > that they operate at a higher level of abstraction because the >> > code for dealing with individual elements is hidden. >> >> The reason I like blocks is because they're sufficiently general that >> they often keep me from having to make new methods for little one-off >> routines only used by a single method. The problem with filters is that >> you need a selector to apply. Usually what I need done can't be done >> with any of the standard selectors, so I have to write a custom one. >> This is annoying when its implementation is only a line or two and is >> only used by a single method. >I agree, and, in addition, it is easy to make variables of all sorts of >datatypes (int, BOOL ,char*) etc. available to a Block. >For example, >findstr(char *s) >{ > [c do:{ :whatever | if (!strcmp([whatever name],s) ... }]; >} Why do you insist on using char* and C-lib functions? return [[c select] isEqualToString:[NSString stringWithCString:s]]; Alas, this syntax doesn't work because isEqualToString: returns a BOOL, meaning the actual id result gets truncated by gcc. So you have to write: return [c selectWithSelector:@selector(isEqualToString:) withObject:[NSString stringWithCString:s]]; >where "s" is a parameter of type char*. It is accessible from within >the Block (and it can even modify the variable). This I find one of the nastier features of blocks, passing by reference in a language that doesn't really have pass-by-reference. Unexpected side effects... >With a -perform:with: type of approach, you will have to cast the char* >to an id to pass it to the selector which does the comparison. This is >ugly and sometimes can't be done. >One may think that it is an inherent difficulty with Objective-C, but I >think it rather just depends on using -perform:with: for this sort of >thing. There's no difficulty for a Block (as opposed to -perform:with:) >to access data that is of type BOOL, int, double, id, char* etc., from >within the Block. Filters don't use perform:with:, only the 'select' convenience method does. Normally, a filter is used as follows: result = [[c filter] substringFromIndex:4]; There are no restrictions on the arguments, and one of the neater features of filters is that they can also step through parallel arrays. result = [[prefixes filter] stringByAppendingString:[suffixes filter]]; I was hoping that select/reject filters could be done the same way: result = [[strings select] isEqual:targetString]; However, the fact that 'isEqualTo:' returns a BOOL and that BOOL is typedef'ed to 'char' wreaks havoc on the return value, at least in OS 4.2 on Intel. To avoid this with the normal filter syntax, one would have to use a nested filter and stash the filter somewhere safe: f=[strings startSelect]; [f isEqualTo:targetString] result = [f runFilter]; The convenience method 'selectWithSelector:' avoids this but imposes the restriction of 'id' arguments. Another way to avoid this would be to have 32-bit BOOLs. A third way would be to either provide a bogus message definition or cast in order to convince the compiler that 'isEqual:' doesn't return a BOOL but something else instead. As a matter of fact, that *might* just work... Thanks for making me think this through! >Without compiler support, I think that all this communication (e.g. >making the parameter "s" available) between the comparison method and >the method that iterates over the collection, can only be easily done by >using global variables (which is not a good solution, since then things >are no longer reentrant). Only if you insist on fully generic blocks. Filters don't use global variables and work without additional compiler support. Anyway, I understand that blocks are by far the more powerful, flexible etc. method. Filters were nothing but a little hack to tide me over until I could get real blocks. I was quite surprised when they turned out to be this useful. Reminds me a little bit of the relationship between Smalltalk and Objective-C. As a matter of fact, wouldn't it be much more useful to get a full Smalltalk integrated with the Objective-C runtime (and thus Foundation/AppKit) rather than trying to graft more and more Smalltalk language features onto Objective-C? I've started a Squeak 'port', anybody else working on this? Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: "NeXT Newbie" <macghod@concentric.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 13 Apr 1998 19:30:34 GMT Message-ID: <01bd674c$8a576280$34f0bfa8@davidsul> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104><joe.ragosta-0804980858150001@wil77.dol.net><mazulauf-0804981503390001@sneezy.met.utah.edu><joe.ragosta-0804981906460001@elk68.dol.net><01bc4540$d4a15e60$3af0bfa8@davidsul><6gh8v1$oop$6@ironhorse.plsys.co.uk> <SCOTT.98Apr9091856@slave.doubleu.com> <01bc46e4$1815c920$24f0bfa8@davidsul> <slrn6j4f0u.juc.sal@panix3.panix.com> Salvatore Denaro <sal@panix3.panix.com> wrote in article <slrn6j4f0u.juc.sal@panix3.panix.com>... > On 10 Apr 1998 18:52:40 GMT, NeXT Newbie <macghod@concentric.net> wrote: > >Plus, now Apple is using the microsoft developer pricing model as its own. > >Oops, but microsoft has %95 of the market and apple only has %3. Oops, > >Microsoft has a monopoly so they can do this. I guess Apple is now "just > >as good" as microsoft <big grin> > > > >Except they dont have a monopoly and are figting for their lifes, so they > >cant use monopolistic pricing like microsoft does. > > > >Clue phone for Mr Jobs, Mr Jobs please pick up the clue phone. Mr Jobs, > >Apple is not Microsoft, Microsoft can charge those prices because they are > >a monopoly, Apple is fighting for its life so it cant treat developers so > >shabbily > > Anyone with a Mac can download MPW and the SDKs and write a Mac App for > $0.00, that is not the case with Windows. If you want to write a win app > that can qualify for the winlogo you need to buy a compiler. True. And this is a good feature. And I am not sure it is a detraction to apple that 1) this has been the case well before the new program was put in place (ie these arent free BECAUSE of this new program, all this was free before) 2) These tools were ONLY put out for free after apple basically gave up on them and put them in maintanence mode > The $495 MSDN subscription does not include tools, the $500 Apple dev sub > does. The $195 MSDN librady rel does not include OS versions, the $199 > Apple one does. The Apple program is well within the industry norms. Why > are Apple users so damn whiney? It sounds like you bitch and moan over > every little thing. The $500 one does not include tools either. The tools were free well before this new program was put in place. Apple users may be whiney. But you have to understand most mac programmers develop for the mac only because they love the mac, and IN SPITE of the financial push NOT TO develop for the mac. > If the $199 price includes Rhapsody CR1, then it is a bargin. For $199 > you get something that cost $5k just over two years ago. And that was 2 years ago. Many years ago a 2fx cost $10,000. My selling you one for $400 NOW is not a bargain What is really upsetting about all of this is that Apple is raising the costs by twofold and giving nothing in return. The free tools? They have been free for some time, you cant try to say this program now gives you these tools for free when they have already been free. The $200 price which doesnt give you seeds? Hasnt this been in place BEFORE this new program ? Apple NEEDS developers! They need more developers! THey need more software! One may say BFD that quickbooks isnt out for the mac, there are plenty of other choices. BFD that quake 3 is out for the pc (is it?) and not for the mac, you can play marathon instead. But it is a BFD because mac users are leaving the mac to use pc's because of this. The last several quarters apple's marketshare has consequatively DROPPED! Hopefully come this wednesday apple will report that its marketshare has skyrocketed from the %2.6 it had last quarter to a number LARGER than what it had the same quarter last year.
From: see-below@not-my-address.com (Matthew Vaughan) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.programmer Subject: Re: first day with OS Date: Mon, 13 Apr 1998 14:06:10 -0700 Message-ID: <see-below-1304981406110001@209.24.241.220> References: <01bc462d$87ca8380$40f0bfa8@davidsul> <35311826.7AD7@www.web> In article <35311826.7AD7@www.web>, Web Surfer <surfer@www.web> wrote: > What OS is that? OS/2? The old NeXT OS? What? The latter. > > NeXT Newbie wrote: > > > > Some impressions of OS 4.2 > > 1) This OS sure was designed for big monitors! I have a 17 inch and it > > seems to small. Seems like you would want at least a 20 inch monitor > > 2) The mouse really annoys me. It reminds me of when I first used windows > > 3.1, I was like god I hate how jerky the mousee is. The mouse movement in > > OS 4.2 seems like how it was in windows 3.1, windows 95 definitely seems > > much better. > > 3) How do you connect to the internet with ppp? > > 4) once I get up and going with ppp, what do I use for usenet? > > 5) where is the option to change the monitors resolution and bits of color? .................................................... MATTHEW VAUGHAN matthewv at best dot com (damn spammers...) http://www.best.com/~matthewv/ ....................................................
From: spagiola@usa.net Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.programmer Subject: Re: first day with OS Date: Mon, 13 Apr 1998 16:49:51 -0600 Organization: Deja News - The Leader in Internet Discussion Message-ID: <6gu19v$7e3$1@nnrp1.dejanews.com> References: <01bc462d$87ca8380$40f0bfa8@davidsul> <35311826.7AD7@www.web> Web Surfer <surfer@www.web> wrote: > NeXT Newbie wrote: > > Some impressions of OS 4.2 [snip] > What OS is that? OS/2? The old NeXT OS? What? In this newsgroup, it's safe to assume OS = OPENSTEP, especially when followed by version number 4.2. But you're right, it can be confusing with OS as a general term. Stefano Pagiola My opinions alone Longtime MacOS and NeXTSTEP user, reluctant Windows 95/NT user, future Rhapsody user -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: David Stes <stes@mundivia.es> Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Blocks for Obj-C Date: Mon, 13 Apr 1998 19:26:35 +0000 Organization: Mundivia, Santander Message-ID: <353266EB.41C67EA6@mundivia.es> References: <6fokj8$1dp$1@usenet50.supernews.com> <6gqrjs$anh$1@news.cs.tu-berlin.de> <353125E4.41C67EA6@mundivia.es> <6gt4i9$835$1@news.cs.tu-berlin.de> <6gt9kd$o9o$1@crib.bevc.blacksburg.va.us> <3532413B.41C67EA6@mundivia.es> <6gtkio$hh7$1@news.cs.tu-berlin.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Marcel Weiher wrote: > > This I find one of the nastier features of blocks, passing by reference > in a language that doesn't really have pass-by-reference. Unexpected > side effects... The technique is in fact a bit different than what I think you mean by passing by reference here. It's not working by references towards (ie. pointers towards) the variables on the caller's stack frame (that would only work for last in first out LIFO Blocks). > As a matter of fact, wouldn't it be much more useful to > get a full Smalltalk integrated with the Objective-C > runtime. Sounds a bit like you're asking "why Blocks?" here. In my opinion : for error handling. It's really convenient to be able to pass a piece of error handler code "down" to the method that sends an -error: message. An error handler Block could then take two arguments, for example like: [ { ... perform some code ... } ifError:{ :msg :rcv | ... handler ... }]; where "msg" is a message string and "rcv" the object that received the -error: message. While the first block is evaluated, if an error occurs, the second Block is passed "down" to the code that sends the -error: message. Any Objective-C program has to do error handling, so this is an important area of functionality.
From: <ItalianBoy@Italy.com> Newsgroups: comp.sys.next.programmer Subject: $600$ OFFER $600$ REWARD-OFFER $600$ REWARD-OFFER$600$.. REWARD- Date: 13 Apr 1998 22:14:14 GMT Organization: Customer of Flashnet S.p.A. - http://www.flashnet.it Message-ID: <6gu2nm$bo0$8977@news.flashnet.it> I'M CURRENTLY LOOKING FOR THESE 2 HONG KONG CITIZENS: ------------------------------------------------------------------------------ KULRAO RATHOUR TEL 00852/5702170-----FAX 00852/8878627 ------------------------------------------------------------------------------- SO TAK ON POSSIBLE PHONES NUMS: 23480445 SOU MOU PING KOWLOON 23647531 HO MANTINI KOWLOON 24570679 TUEN MUN NEW TERRITORIES -------------------------------------------------------------------------------------------- OFFERING REWARD----------->600 US$<----------- FOR ANYONE GIVES INFORMATIONS AND ACTIONS LEADING TO HAVE BACK A CERTAIN AMMOUNT OF MONEY EMAIL IF YOU CAN HELP US: amonici@hotmail.com
From: <ItalianBoy@Italy.com> Newsgroups: comp.sys.next.programmer Subject: $600$ OFFER $600$ REWARD-OFFER $600$ REWARD-OFFER$600$.. REWARD- Date: 13 Apr 1998 22:14:14 GMT Organization: Customer of Flashnet S.p.A. - http://www.flashnet.it Message-ID: <6gu2nm$bo0$8976@news.flashnet.it> I'M CURRENTLY LOOKING FOR THESE 2 HONG KONG CITIZENS: ------------------------------------------------------------------------------ KULRAO RATHOUR TEL 00852/5702170-----FAX 00852/8878627 ------------------------------------------------------------------------------- SO TAK ON POSSIBLE PHONES NUMS: 23480445 SOU MOU PING KOWLOON 23647531 HO MANTINI KOWLOON 24570679 TUEN MUN NEW TERRITORIES -------------------------------------------------------------------------------------------- OFFERING REWARD----------->600 US$<----------- FOR ANYONE GIVES INFORMATIONS AND ACTIONS LEADING TO HAVE BACK A CERTAIN AMMOUNT OF MONEY EMAIL IF YOU CAN HELP US: amonici@hotmail.com
From: bl003@dial.oleane.com (Benoit Leraillez) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Mon, 13 Apr 1998 20:00:52 +0200 Organization: Guest of OLEANE Message-ID: <1d7fwrb.t2p83i1fjkkdtN@dialup-96.def.oleane.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ylaporte-0804982005030001@139.103.176.79> <6gh9cm$oop$7@ironhorse.plsys.co.uk> <joe.ragosta-0904980643360001@elk33.dol.net> <1d7efs4.14d33ptutamabN@ppgutkneco.lirmm.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Olivier Gutknecht <gutkneco+news@lirmm.fr> wrote: > Another problem about the forthcoming (?) edu programs is about non-US > educational/research communities. I hope that worldwide educational > institutions will be able to enroll. Given the current status of > European Developer Relations, I am a little bit dubious. Espaecially when their recommendation is to develop for Windows :-( Benoît Leraillez PS This is unfortunatly true, it happened to one of my partners.
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Blocks for Obj-C Date: 14 Apr 1998 00:27:03 GMT Organization: Technical University of Berlin, Germany Message-ID: <6guagn$ppl$1@news.cs.tu-berlin.de> References: <6fokj8$1dp$1@usenet50.supernews.com> <6gqrjs$anh$1@news.cs.tu-berlin.de> <353125E4.41C67EA6@mundivia.es> <6gt4i9$835$1@news.cs.tu-berlin.de> <6gt9kd$o9o$1@crib.bevc.blacksburg.va.us> <3532413B.41C67EA6@mundivia.es> <6gtkio$hh7$1@news.cs.tu-berlin.de> <353266EB.41C67EA6@mundivia.es> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit David Stes <stes@mundivia.es> writes: >Marcel Weiher wrote: >> >> This I find one of the nastier features of blocks, passing by reference >> in a language that doesn't really have pass-by-reference. Unexpected >> side effects... >The technique is in fact a bit different than what I think you mean by >passing by reference here. It's not working by references towards (ie. >pointers towards) the variables on the caller's stack frame (that would >only work for last in first out LIFO Blocks). I am talking about an inside scope modifying a variable defined in an outside scope without explicitly getting a pointer to that variable. This is not something that usually happens in C, so it can lead to surprising behaviour. Code should try not to be surprising like that. >> As a matter of fact, wouldn't it be much more useful to >> get a full Smalltalk integrated with the Objective-C >> runtime. >Sounds a bit like you're asking "why Blocks?" here. No. I am saying that if you want Smalltalk-like features, it makes sense to use Smalltalk instead of trying to graft features onto Objective-C. I think Smalltalk would make much more sense as an additional language in Yellow-Box than, let's say, Java, Perl or Tcl. >In my opinion : for error handling. New subject. >It's really convenient to be able to pass a piece of error handler code >"down" to the method that sends an -error: message. An error handler >Block could then take two arguments, for example like: > [ { ... perform some code ... } ifError:{ :msg :rcv | ... handler ... >}]; >where "msg" is a message string and "rcv" the object that received the >-error: message. >While the first block is evaluated, if an error occurs, the second Block >is passed "down" to the code that sends the -error: message. >Any Objective-C program has to do error handling, so this is an >important area of functionality. Well, exceptions handle the normal case of unwinding a call with an error condition, and the NSException class allows attaching arbitrary information. For the case where your error handler might want to *retry* the operation, I'd been thinking about a custom NSException subclass that would raise only if the retry fails. Blocks may be useful in this context, I don't know. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: ericb@pobox.com (Eric Bennett) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Mon, 13 Apr 1998 20:32:15 -0400 Organization: Cornell University Sender: emb22@cornell.edu (Verified) Message-ID: <ericb-1304982032150001@132.236.171.104> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <B1511019-15903@206.165.43.154> <steve-0804981442590001@oranoco.discoverysoft.com> <rmcassid-0804981509160001@dante.eng.uci.edu> <1d7821x.1rdi5101enh5fhN@hobbit1.injep.fr> <6girj3$6j$2@anvil.BLaCKSMITH.com> <1d78ren.1dbj70n1j99pebN@hobbit1.injep.fr> <6glqd2$7j4$4@anvil.BLaCKSMITH.com> In article <6glqd2$7j4$4@anvil.BLaCKSMITH.com>, cswiger@BLaCKSMITH.com wrote: > xhsoft@injep.fr (Xavier Humbert) wrote: > > Charles Swiger <chuck-nospam@blacksmith.com> wrote: > >> the increase in costs represents roughly a day's pay > > > > That's not the point. 100% is 100%. > > Cans of soda used to cost 25 cents; they now cost 50 (or $1.00 or more on the > highway). That's a 100% increase, also. People who are thirsty deal with > it-- many of them don't even notice. I notice it. My local grocery store had 2-liter bottles of their generic root beer for fifty cents a couple months ago. I refuses to pay $1.00 for a can of soda when I can get so much more for so much less. As for what happens when I'm thirsty, where there's a soda machine, there's almost always a water fountain. -- Eric Bennett (www.pobox.com/~ericb) Cornell University Biochemistry Department
From: tom_e@usa.net (Thomas Engelmeier) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Mon, 13 Apr 1998 12:13:11 +0200 Organization: University of Rostock Message-ID: <1d78dx8.1olb8yi6si1r4N@desktop.tom-e.private> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <trumbull-0904980039090001@net44-223.student.yale.edu> <6gio5m$oop$11@ironhorse.plsys.co.uk> mmalcolm crawford <malcolm@plsys.co.uk> wrote: > If you're a real hobbyist, do you *need* access to the bleeding edge s/w? > If not, then you don't need to belong to the program, so it won't cost you > anything. I do Apple development as a "hobbyist" (not too many Mac programming jobs @ Munich at the moment), and especially for this I need access to the latest stuff. Unlike when programming for profit I need to consider the latest, "in progress" SDKs for this or that small project that I wouldn't have the time to implement otherwise. The big companies wont bother to reinvent the wheel, I simply neither have the manpower nor the time. And yes, programming Multimedia and grafics is sometimes quite complex. > (I didn't have the legs to play Doris); incidental expenses including travel > to and from rehearsals are likely to cost as much as subscription to the > Select program for the next two months. Maybe I should send in my petrol > receipts to Apple...? A comparison of the cost of hobbies is completely off the point. Just because I go scuba diving every week it is not reasonable to bill you for Inline-Skating $120 / weekend. > more? No. Does this mean that I think Apple's taking a sensible business > attitude to concentrating its resources where they matter? Yes. For the last two years, they didn't provide too much resources. Or maybe someone has better experiences? > I also hope that the additional fees will go toward giving a better service > to those who choose to stay in one of the developer programs. And I guess you belive in the Easter bunny and Santa claus. IMO it is a better bussiness practice first to establish a better service and then charge for it accordingly. Regards, TomE -- 3 KINDS OF PEOPLE: THOSE WHO CAN COUNT AND THOSE WHO CAN'T.
From: David Stes <stes@mundivia.es> Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Blocks for Obj-C Date: Tue, 14 Apr 1998 06:44:24 +0000 Organization: Mundivia, Santander Message-ID: <353305C8.41C67EA6@mundivia.es> References: <6fokj8$1dp$1@usenet50.supernews.com> <6gqrjs$anh$1@news.cs.tu-berlin.de> <353125E4.41C67EA6@mundivia.es> <6gt4i9$835$1@news.cs.tu-berlin.de> <6gt9kd$o9o$1@crib.bevc.blacksburg.va.us> <3532413B.41C67EA6@mundivia.es> <6gtkio$hh7$1@news.cs.tu-berlin.de> <353266EB.41C67EA6@mundivia.es> <6guagn$ppl$1@news.cs.tu-berlin.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Marcel Weiher wrote: > > I am saying that if you want Smalltalk-like features, > it makes sense to use Smalltalk instead of trying to graft > features onto Objective-C. I think Smalltalk would make > much more sense as an additional language in Yellow-Box > than, let's say, Java, Perl or Tcl. That's fine by me, but it is a topic for comp.sys.next.programmer, totally unrelated to comp.lang.objective-c. I just happened to note your message about protocols (titled "Blocks for Obj-C") in the comp.lang.objective-c group, and that group is concerned with the development of Objective-C.
From: kbenrmcaNICKI@BORED.COM Newsgroups: comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.nsc,comp.sys.nsc.32k Subject: Bored? Looking to do something really different? Date: 14 Apr 1998 09:10:07 GMT Organization: Road Runner Message-ID: <6gv95f$hqp$1021@proxye2.nycap.rr.com> **************************************************************** * This Article was Posted By an unregistered version of: * * Newsgroup AutoPoster 95 * * Send email address for info! Fax: +46-31-470588 * **************************************************************** Why don't you come see me at WWW.HOTXXXBABES.COM
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Followup-To: comp.sys.mac.advocacy,comp.sys.next.advocacy Date: 14 Apr 1998 12:36:01 GMT Organization: P & L Systems Message-ID: <6gvl7h$oop$43@ironhorse.plsys.co.uk> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <trumbull-0904980039090001@net44-223.student.yale.edu> <6gio5m$oop$11@ironhorse.plsys.co.uk> <1d78dx8.1olb8yi6si1r4N@desktop.tom-e.private> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: tom_e@usa.net NOTE: FOLLOWUPS TO COMP.SYS.MAC.ADVOCACY,COMP.SYS.NEXT.ADVOCACY In <1d78dx8.1olb8yi6si1r4N@desktop.tom-e.private> Thomas Engelmeier wrote: > mmalcolm crawford <malcolm@plsys.co.uk> wrote: > > > If you're a real hobbyist, do you *need* access to the bleeding edge s/w? > > If not, then you don't need to belong to the program, so it won't cost you > > anything. > > I do Apple development as a "hobbyist" (not too many Mac programming > jobs @ Munich at the moment), and especially for this I need access to > the latest stuff. Unlike when programming for profit I need to consider > the latest, "in progress" SDKs for this or that small project that I > wouldn't have the time to implement otherwise. > I was arguing the reverse; it's more likely to be commercial organisations who need access to the latest releases of s/w. If programming is a *hobby* for you, why do you need access to beta code, or if you do, why shouldn't you have to pay for it like other people do for materials for their hobbies? Better still ask the question below... > > I also hope that the additional fees will go toward giving a better service > > to those who choose to stay in one of the developer programs. > > And I guess you belive in the Easter bunny and Santa claus. IMO it is a > better bussiness practice first to establish a better service and then > charge for it accordingly. > Are Apple providing good value for money with the new programs? I believe they are. mmalc.
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 14 Apr 1998 11:28:27 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6gvh8r$8u$1@ns3.vrx.net> References: <B1540F1E-99E8B@206.165.43.22> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: english@primenet.com In <B1540F1E-99E8B@206.165.43.22> "Lawson English" claimed: (innappropriate newsgroups removed) > BTW, anyone read the new excerpts from Amelio's book about how Steve Jobs > approached Amelio well before the NeXT purchase claiming that he was the > ultimate CEO for Apple and that Amelio should just step aside and let the > right stuff handle the job? Ever hear that aliens built a statue on Mars? Even though the latest Mars Orbiter photos show it to be nothing more than a jumble of rock, that's just NASA massaging the data and lying to the world. Regardless, I think he is the right person for the job, and it would appear that for the first time in about a decade both Wall St. and Fleet St. agree. > But nyah, there was never any question that Jobs and Ellison hadn't > collaborated on the Jobs takeover of Apple. Let's see if I'm deciphering "Lawgic" correctly: since: Steve Jobs says he is the best man for the job. therefore: Jobs and Ellison were in cahoots. How clear. Other questions aside from the obvious non-sequiter of this argument exist: so what, and who cares? Maury
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 14 Apr 1998 11:33:31 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6gvhib$8u$2@ns3.vrx.net> References: <joe.ragosta-1004980619170001@elk81.dol.net> <B153C824-106D1@206.165.43.115> <6gm2s3$213$1@news.digifix.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: sanguish@digifix.com In <6gm2s3$213$1@news.digifix.com> Scott Anguish claimed: > >No, you're chuckling over DPS-lover's propoganda. No Lawson, that would be "professional DPS user's comments". In GX's defence you've offered comments solely from people (well, you and Rex) who as far as I can tell have never used either GX or DPS in a production setting. In other words I'm quite willing to write off your entirely argument set, notably when it proves to be wrong so often. > >feedback about DPS that you are getting has been from people that use > DPS, > >not GX. Whereas the only feedback we're getting about GX is from someone who has never releases a commercial product in it. If GX were so studly, why is it that the developers who actually use it can't be bothered to talk about it? > Also, the people in charge of graphics at Apple *invented* > DPS, not > >GX, so they have a certain level of bias, also. Let me see if I understand this... Team 1 from Apple creates DPS Team 2 from Apple creates GX I'm supposed to believe that... a) Team 1 is biased b) Team 2 is better than Team 1 More "Lawgic". Maury
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 14 Apr 1998 11:55:00 -0700 Organization: Primenet Services for the Internet Message-ID: <B1590035-5B0E7@206.165.43.147> References: <6gvhib$8u$2@ns3.vrx.net> nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Maury Markowitz <maury@remove_this.istar.ca> > > Also, the people in charge of graphics at Apple *invented* > > DPS, not > > >GX, so they have a certain level of bias, also. > > Let me see if I understand this... > > Team 1 from Apple creates DPS > Team 2 from Apple creates GX > > I'm supposed to believe that... > > a) Team 1 is biased > b) Team 2 is better than Team 1 > > More "Lawgic". No. Team 1 is in charge. Team 2 is disbanded. Team 1 makes comments. I suggest that since Team 1 is the only one left to make comments, that they are biased. Team 2 is no longer available, so it CANNOT make comments, and hence can't even be biased (although that I'm sure that they would be if they were still around and DID make comments). The point is: you're getting one side of the story from Apple and assuming that since it is Apple engineers making the comments, that GX must be no good since Apple engineers also made GX. 'taint necessarily so. ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 14 Apr 1998 15:57:24 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6h0114$bnb$2@ns3.vrx.net> References: <6gvhib$8u$2@ns3.vrx.net> <B1590035-5B0E7@206.165.43.147> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: english@primenet.com In <B1590035-5B0E7@206.165.43.147> "Lawson English" claimed: > No. Team 1 is in charge. Team 2 is disbanded. Team 1 makes comments. I see no comments from Team 1 or Team 2, aside from those I've received from the later in e-mail. > I suggest that since Team 1 is the only one left to make comments, that > they are biased. Team 2 is no longer available, so it CANNOT make comments, > and hence can't even be biased (although that I'm sure that they would be > if they were still around and DID make comments). Ahh yes, it's all revisionist history at work! > The point is: you're getting one side of the story from Apple I get NONE of the story from Apple. Maury
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Organization: Is a sign of weakness Message-ID: <SCOTT.98Apr10103112@slave.doubleu.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> <352D4957.84AE1890@alexa.com> In-reply-to: Joshua Winsor's message of Thu, 09 Apr 1998 15:19:04 -0700 Date: 14 Apr 98 20:53:07 GMT In article <352D4957.84AE1890@alexa.com>, Joshua Winsor <joshua@alexa.com> writes: Scott Anguish wrote: > Now, as far as Rhapsody is concerned it is leaving a gap, but > only because of the time-frame.... However if you join Select > you're still covered. Download GnuStep and MKLinux, all free, MKLinux paid for by Apple, and you are a Rhapsody developer. A couple people have mentioned this, and it almost couldn't be further from the truth. Downloading the above _might_ make you an "OpenStep" developer. But GnuStep is behind OpenStep, and OpenStep is being obsoleted by Rhapsody. Let me make sure I'm being understood. OpenStep, GnuStep, and Rhapsody will all share a certain core API. You will be able to write an application which will compile and run on all three. You _also_ will be able to write applications on all three that won't compile on the other two. It's very much like saying that downloading Sun's JDK makes you a Microsoft J++ developer, or running J++ makes you a Java developer. True in one way, very untrue in another, more important way, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: "NeXT Newbie" <macghod@concentric.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.hardware.misc,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: letter to macnn about developer article Date: 14 Apr 1998 21:32:19 GMT Message-ID: <01bd67ec$731a7b20$2ef0bfa8@davidsul> >by Oliver Dueck > >Last week, Apple announced some drastic changes to its developer program. Prices were changed, services shuffled. No, prices were increased with no new services offered, or services deleted >At first, all we heard were complaints. After two or three days, more people came forward and said that the new program wasn't that bad. Personally, I was kind of mad at first - until I realized that Apple wasn't hurting developers. True, its not ALL THAT BAD to lots of people. Still, their is nothing good >The biggest complaint was the new pricing structure. Previously, you could enter the Associate program for a mere $250. Not only did you get pre-release software at that price; you even got hardware discounts. It was a great bargain. It was a so so deal. Apple was definitely making money on the hardware sales (the deals were comparative to what some mail order places made, and when you realize that apple sold the hardware to mail order places at LESS than what the mail order places sold them for (the mail order places obviously have to make a profit) you realize apple was selling the hardware to developers with a BIGGER profit than if the developers bought from mail order > >Now, the minimum you have to pay to get pre-release software (read: Rhapsody DR1) is $500 per year, for the Select Program. That gets you the following: > > > >•Apple Developer Connection Mailing (monthly CDs) •Two technical support incidents •$100 Metrowerks coupon •Apple Software Seeding Program (Rhapsody DR1) •SDKs •Access to Compatibility Labs I dont know what strings are attached to the coupon so I cant comment on it. >All in all, not a bad deal. The Premier Program costs a lot more - $3500. But look at what you get: > > > >•Apple Developer Connection Mailing •Eight technical support incidents •$300 Metrowerks coupon •Full conference pass to the Worldwide Developers Conference •Hardware discounts •Service Source CD •Apple Technology Seeding •SDKs •Access to Compatibility Labs > > > >Now, the arguments against the new programs usually center on two things: hardware discounts and shareware developers. > >It used to be that a lot of people would join the program simply because it could pay for itself when you bought a new computer. But hey, that's not really fair to Apple - they only lose, because those people weren't developing anything. Apple makes more money selling this way than if someone bought mailorder and you say apple loses out? HUH???? > >And besides, the discounts aren't that steep anymore. The prices are often close to what you would find at a typical mail order store. By offering hardware discounts only to Premier Program members, Apple ensures that people don't enter the program simply to get a cheap Mac. You admit that the discount is about the same as mail order and then say it ensures that people dont enter the program to get a cheap mac, when you already admit their isnt a saving??? Is Steve Jobs paying you to defend apple?? > >People are also complaining that the new prices/features hurt shareware developers. Do they? No. WRONG, some definitely are hurt. Those that want to develop useful apps for rhapsody and have this software out when rhapsody cr1 is out AND cant afford the increase are hurt. >You don't need to be part of the developer program to make Mac or Rhapsody shareware. Sure, it can help, but if you put the $100 Metrowerks coupon to use, you only have to pay $400 - or forty $10 shareware registrations. It is still quite affordable. If you want your software to be out when rhapsody cr1 is released, then yes you either need to be part of the developer program OR you need to ILLEGALLY BREAK THE LAW and PIRATE rhapsody >Is Apple doing anything wrong here? Actually, yes; Apple is hurting the hobbyists who previously joined the developer program so that they could play with prerelease software, and maybe develop a few things for personal use. > >These people don't want to pay $500, and they don't need access to compatibility labs, tech support, or even the Metrowerks discount. So, Apple, bring back that $250 price point in a package that gets you only the software, and I'll be happy. If for $250 one could get the seeds that they got in the old program, plus the cd's with the programming tools and stuff, but no metrowerks discount, no hardware discount, no tech support, no compatibility labs access, apple would not hurt people as badly as they are. In fact, if apple was really smart, they would price this at $150 or $200. WHy? See bellow. Steve Jobs seems to be acting like he is Bill Gates and Apple is Microsoft. I am sorry apple, you are no microsoft. You are a STRUGGLING company whose developers are LEAVING YOU and who has a problem with LACK OF SOFTWARE. Microsoft has a monopoly, developers develop for ms windows because financially they HAVE to. FInancially developers are DISCOURAGED from developing for the mac. Mac developers develop DESPITE this because they love the mac. Lots of mac developers develop mac products in their spare time because their company doesnt support the mac because they believe it is not profitable enough, and wont support the developers. They developers have to pay for the apple program out of their own pocket. Apple needs to do everything they can to encourage development for the mac, because without software for the mac, the mac will die. And every day mac users are turning to pc's because the mac does not have some software package they need, such as quickbooks. And saying "well they could use MYOB instead" does no good because they ARE SWITCHING, and what good does it do that they could use another software package when they will still switch? Mac developers develop IN SPITE OF THE FINANCIAL forces, and Apple by doubling the fees while keeping the same benefits or reducing benefits is making it even harder for some developers, and this is plain stupid. Just like the story of Steve Jobs, when seeing a prominent game developer being given a tour of apple, having armed security guards with guns escort him out of the premises and threatened with arrest was stupid of JObs to do. I am sorry Apple, you do not have a monopoly like microsoft does, stop thinking you can use monopolistic pricing when people are leaving the mac in droves. Tommorow we will find out what the worldwide marketshare is, *HOPEFULLY* it will be higher than what it was the same quarter last year, but even if it is, this is only a start Emailed to oliver@macnn.com (the author) and also posted to usenet
From: a2663376@athena.rrz.uni-koeln.de (Markus Vollmert) Newsgroups: comp.sys.next.programmer Subject: dirname command Date: Tue, 14 Apr 1998 22:13:15 GMT Organization: Regional Computing Center, University of Cologne Message-ID: <3533dea9.833271@news.uni-koeln.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit hi i am trying to make some unix source code under next, but some scripts do not execude. in one script the command "dirname" is used. i looked it up in a unix documentation but cannot find the command under next. is there an alternative? i found the basename command, which is also used in the script, but that did not help me. bye markus Markus Vollmert - a2663376@athena.rrz.uni-koeln.de
From: M Rassbach <mark@milestonerdl.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Tue, 14 Apr 1998 16:36:07 -0500 Organization: Inc.Net http://www.inc.net Message-ID: <3533D6C7.95AECD96@milestonerdl.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> <352D4957.84AE1890@alexa.com> <SCOTT.98Apr10103112@slave.doubleu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Scott Hess wrote: > In article <352D4957.84AE1890@alexa.com>, > Joshua Winsor <joshua@alexa.com> writes: > Scott Anguish wrote: > > Now, as far as Rhapsody is concerned it is leaving a gap, but > > only because of the time-frame.... However if you join Select > > you're still covered. > > Download GnuStep and MKLinux, all free, MKLinux paid for by Apple, > and you are a Rhapsody developer. > > A couple people have mentioned this, and it almost couldn't be further > from the truth. Downloading the above _might_ make you an "OpenStep" > developer. But GnuStep is behind OpenStep, and OpenStep is being > obsoleted by Rhapsody. But the GnuStep/MKLinux route won't leave you shafted the way Apple just shafted the Newton Developers. And, keep in mind, Apple TODAY is a niche market player. The niche for NeXTSTEP/OpenSTEP has just grown, but it's still a niche. And Apple is unlikly to get beyond %25 of the whole market. And Apple has formally not committed to any FUTURE support, save it be MacOS. (if you are willing to wait till May, you may see formal support for NT[5-x]/win[98-x], perhaps other Unixes) So, if you are liking niches, AND willing to take the risk that, at any moment, you can get 'Steved' (like the newton developers did) then, go nutz on Rhapsody. And, if you go the Rhapsody route, there IS an exit route of GnuSTEP. (kinda like the Windows folks have willows)
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 14 Apr 1998 18:50:07 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6h0b4v$h5v$3@ns3.vrx.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> <352D4957.84AE1890@alexa.com> <SCOTT.98Apr10103112@slave.doubleu.com> <3533D6C7.95AECD96@milestonerdl.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: mark@milestonerdl.com In <3533D6C7.95AECD96@milestonerdl.com> M Rassbach claimed: > But the GnuStep/MKLinux route won't leave you shafted the way Apple just > shafted the Newton Developers. No, it will just leave you in the forever-limbo of projects that never go final, or even approach true usability. GNUStep, from a user perspective, does not exist. For the most part it's a mostly-complete FoundationKit and an incomplete to not-even-really-working everything else. > And, if you go the Rhapsody route, there IS an exit route of GnuSTEP. > (kinda like the Windows folks have willows) Not a single one of my apps will work under GNUStep because the libraries are not complete. How is this an "out"? Maury
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6gu2nm$bo0$8976@news.flashnet.it> Control: cancel <6gu2nm$bo0$8976@news.flashnet.it> Date: 13 Apr 1998 22:31:09 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6gu2nm$bo0$8976@news.flashnet.it> Sender: <ItalianBoy@Italy.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6gu2nm$bo0$8977@news.flashnet.it> Control: cancel <6gu2nm$bo0$8977@news.flashnet.it> Date: 13 Apr 1998 22:31:51 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6gu2nm$bo0$8977@news.flashnet.it> Sender: <ItalianBoy@Italy.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 14 Apr 1998 21:48:00 -0700 Organization: Primenet Services for the Internet Message-ID: <B1598B35-5BAB4@206.165.43.216> References: <6h0114$bnb$2@ns3.vrx.net> nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Maury Markowitz <maury@remove_this.istar.ca> said: > In <B1590035-5B0E7@206.165.43.147> "Lawson English" claimed: > > No. Team 1 is in charge. Team 2 is disbanded. Team 1 makes comments. > > I see no comments from Team 1 or Team 2, aside from those I've received > from the later in e-mail. > You talk to former GX team members in e-mail? Oh, you mean Dan Lipton. Mike Paquette of Team 1 contributes to these threads all the time. > > I suggest that since Team 1 is the only one left to make comments, that > > they are biased. Team 2 is no longer available, so it CANNOT make > comments, > > and hence can't even be biased (although that I'm sure that they would be > > if they were still around and DID make comments). > > Ahh yes, it's all revisionist history at work! ??? Team 1 makes comments that I assert are biased. Team 2 CAN'T make comments because it isn't around any more and I suggest that if they WERE still around, that they would be biased also. How is this "revisionist history?" People tend to be biased about things that they are/were involved in. My point isn't that one team would be biased while the other wouldn't be, but only that you are hearing from ONLY one team simply because it is the only one left to make comments of any kind and that these comments are biased, not because of some conspiracy, but because that is the nature of human beings: to present THEIR side of the arguement rather than the other side. > > > The point is: you're getting one side of the story from Apple > > I get NONE of the story from Apple. Mike Paquette works for Apple. Other people program using DPS and OpenStep. You don't hear from many GX programmers because there aren't that many contributing to comp.sys.mac.advocacy or comp.sys.next.advocacy or comp.sys.next.programmer. ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: Henry McGilton <henry@trilithon.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Tue, 14 Apr 1998 22:18:21 -0700 Organization: Trilithon Research and Trading Message-ID: <3534431D.AD9C2AA6@trilithon.com> References: <6h0114$bnb$2@ns3.vrx.net> <B1598B35-5BAB4@206.165.43.216> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lawson English wrote: <<<<< standard nonsense snipped >>>>> * Mike Paquette works for Apple. And works bloody hard, too. * Other people program using DPS very few people "program using DPS". This is one of the standard items of mythology promulgated by the anti-OpenStep faction. The amount of *PostScript* programming necessary in any run-of-the-mill OpenStep application is down around the one percent region. I posted a long letter to this effect to one of the UNIX rags a few years back (trying to rebut the dissemination of FUD at the time) and they refused to publish it. I will attempt to dig up the statistics I gathered at the time that demonstrated how little *PostScript* programming is required in OpenStep applications. * You don't hear from many GX programmers because there * aren't that many contributing to comp.sys.mac.advocacy * or comp.sys.next.advocacy or comp.sys.next.programmer. Well, what *are* they contributing to? You sure aren't contributing to much. Where's you're OpenStep GX AppKit layer? When will you start on it? Will you ever finish it, given you're spending your life telling us how lousy DPS is compared to the paper tiger of GX? ........ Henry ============================================================= Henry McGilton | Trilithon Software, and, Boulevardier, Java Composer | Pacific Research and Trading -----------------------------+------------------------------- mailto:henry@trilithon.com | http://www.trilithon.com =============================================================
From: Timothy J Luoma <nospam+yes-this-is-a-valid-address@luomat.peak.org> Newsgroups: comp.sys.next.programmer Subject: Re: dirname command Date: Tue, 14 Apr 1998 18:17:32 -0400 Organization: @Home Network Message-ID: <Pine.NXT.3.96.980414181605.6101A-100000@luomat> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII To: a2663376@athena.rrz.uni-koeln.de Someone gave me this: #! /bin/sh PATH=/bin:/usr/bin expr \ ${1-.}'/' : '\(/\)[^/]*//*$' \ \| ${1-.}'/' : '\(.*[^/]\)//*[^/][^/]*//*$' \ \| . I think the GNU dirname is also part of the sh-utils package: ftp://next-ftp.peak.org/pub/next/apps/unix/GNU_sh-utils.1.16.NIHS.b.tar.gz ftp://next-ftp.peak.org/pub/next/apps/unix/GNU_sh-utils.1.16.README TjL
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 15 Apr 1998 01:02:00 -0700 Organization: Primenet Services for the Internet Message-ID: <B159B890-32953@206.165.43.108> References: <3534431D.AD9C2AA6@trilithon.com> nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Henry McGilton <henry@trilithon.com> said: > * Mike Paquette works for Apple. > And works bloody hard, too. I am sure that he does. And your point? The GX team worked hard for APple also. Some still do in other graphics projects, mostly in the QTML, I believe. > * Other people program using DPS > very few people "program using DPS". This is one of the standard > items of mythology promulgated by the anti-OpenStep faction. Are you putting me in that faction? Why? Because I would prefer to see a different graphics architecture? Whether or not I have any valid points to make about DPS vs GX, I seldom, if ever, knock Objective C and while I'm not convinced that the Rhapsody Yellow Box framework will be "perfect," I'm quite willing to concede that a universally avaible OOP framework can have MAJOR advantages over Apple's Toolbox-style APIs. The > amount of *PostScript* programming necessary in any run-of-the-mill > OpenStep application is down around the one percent region. I > posted a long letter to this effect to one of the UNIX rags > a few years back (trying to rebut the dissemination of FUD at > the time) and they refused to publish it. I will attempt to > dig up the statistics I gathered at the time that demonstrated > how little *PostScript* programming is required in OpenStep > applications. > Directly, you mean. I understand that the DPS engine can render things faster than GX in many ways. I accept that. I just wanna have the ability to do in Rhapsody the kinds of things that I can with GX. > * You don't hear from many GX programmers because there > * aren't that many contributing to comp.sys.mac.advocacy > * or comp.sys.next.advocacy or comp.sys.next.programmer. > Well, what *are* they contributing to? Making 4-star apps for Macintosh, like Ready, Set, Go! or LIghtningDraw Pro, or Creator2. Currently, the only way to add a QuickTime Vector Graphics track to a QuickTime movie is via LightningDraw, and RSG is/was shipped with every Mac sold in China, if you don't think that any of those are strategic apps > You sure aren't contributing to much. Where's you're OpenStep GX AppKit > layer? > When will you start on it? Will you ever finish it, given you're > spending your life telling us how lousy DPS is compared to the > paper tiger of GX? > Why are you defending DPS? If DPS is so seldom used in Rhapsody, why worry as long as Rhaposdy can produce the graphics that YOU want it to? Everyone accuses me of being a nut about GX. Fine. But the only ones who would bother arguing with a nut over something that he is obsessed with are other nuts, I'll bet... ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.nsc,comp.sys.nsc.32k Subject: cmsg cancel <6gv95f$hqp$1021@proxye2.nycap.rr.com> Control: cancel <6gv95f$hqp$1021@proxye2.nycap.rr.com> Date: 14 Apr 1998 09:20:18 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6gv95f$hqp$1021@proxye2.nycap.rr.com> Sender: kbenrmcaNICKI@BORED.COM Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: pete@ohm.york.ac.uk (-bat.) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 15 Apr 1998 10:25:15 GMT Organization: The University of York, UK Sender: pcf1@york.ac.uk Message-ID: <6h21ub$cfe$1@pump1.york.ac.uk> References: <B159B890-32953@206.165.43.108> "Lawson English" <english@primenet.com> writes: > Why are you defending DPS? If DPS is so seldom used in Rhapsody, why worry > as long as Rhaposdy can produce the graphics that YOU want it to? Everyone Because we don't have to change our software and learn something new to do it. Which is, as far as I can make out, what 90% of this comes down to. You're pissed off 'cos you know and like GX and they took it away. Fair enough, I'd be pissed off if they'd kept GX and taken away DPS. I don't think either of us really gives much of a monkeys about which is technocally better, - in then end they both colour the pixels don't they ? It's just a quiestion of having to re-learn things and change long acquired programming habits which irritates people. I can be smug because I don't have to - and I can appreciate that you are annoyed because you do. People don't like change - I'm still exasperated that the japanese changed the side on which they put the indicators on their cars as a sop to those bits of the world that want to drive on the right. And it bugs me every time I hit the windscreen wipers by accident and that I cant indicate and change gear anymore. But thats life I guess, and I'll unlearn the old way given time. Might even learn to enjoy front wheel drive too. Same with you, it's gonna be a pain, but if you work with DPS for a while you'll get used to it, and if there are really features you miss about GX then code up some nice objecty replacements and drop them into the misc kit. -bat.
From: andrew_abernathy@omnigroup.com Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.hardware.misc,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: letter to macnn about developer article Date: 15 Apr 1998 10:47:44 GMT Organization: Omni Development, Inc. Message-ID: <6h238g$e1p$1@gaea.omnigroup.com> References: <01bd67ec$731a7b20$2ef0bfa8@davidsul> "NeXT Newbie" <macghod@concentric.net> wrote: > >by Oliver Dueck > > > >Last week, Apple announced some drastic changes to its developer program. > Prices were changed, services shuffled. > > No, prices were increased with no new services offered, or services deleted Speak for yourself. The new programs saved me $50 as I don't need the software seeds. (I have access to them at work, but don't need them at home.) And once Rhapsody CR1 ships, it'll be $50 easier to get friends to try it as a development platform. --- andrew_abernathy@omnigroup.com - MIME mail ok
From: Eric Hermanson <eric@alum.mit.edu> Newsgroups: comp.sys.next.programmer,comp.sys.next.software Subject: 2-D Plotting for OpenStep? Date: Wed, 15 Apr 1998 04:12:03 -0700 Organization: Digital Universe Corporation Message-ID: <35349602.34D58F2A@alum.mit.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello, Can anyone recommend the top 3 commercial 2-D plotting packages for OpenStep? Ideally, the packages would ship with frameworks that developers could use to incorporate plotting capability into YellowBox and maybe even WebObjects applications? Thank You, Eric
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 15 Apr 1998 08:17:16 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6h1qec$os8$2@ns3.vrx.net> References: <6h0114$bnb$2@ns3.vrx.net> <B1598B35-5BAB4@206.165.43.216> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: english@primenet.com In <B1598B35-5BAB4@206.165.43.216> "Lawson English" claimed: > You talk to former GX team members in e-mail? Oh, you mean Dan Lipton. > > Mike Paquette of Team 1 contributes to these threads all the time. Point taken. > ??? Team 1 makes comments that I assert are biased. Team 2 CAN'T make > comments because it isn't around any more and I suggest that if they WERE > still around, that they would be biased also. Biased in what way though? Can it not be possible that the technical details are indeed meaningless in these cases? I think it is, I'm writing production code with it. > Mike Paquette works for Apple. But I haven't seen much of a "story" from him. Maury
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 15 Apr 1998 08:25:21 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6h1qth$os8$3@ns3.vrx.net> References: <3534431D.AD9C2AA6@trilithon.com> <B159B890-32953@206.165.43.108> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: english@primenet.com In <B159B890-32953@206.165.43.108> "Lawson English" claimed: > Directly, you mean. I understand that the DPS engine can render things > faster than GX in many ways. I accept that. I just wanna have the ability > to do in Rhapsody the kinds of things that I can with GX. This is still a forest for the trees issue Lawson. If the app you're developing needs 1000 lines of PS code (and mine, a graphics app, likely doesn't even have anything close to that) whereas 100 would do in GX, this nicety is completely washed away by the fact that simply to get the program running under MacOS takes thousands of lines, and maybe none at all on OS. For instance, my radial fills are fancier than Illustrators, yet that took THREE LINES of PS. > Making 4-star apps for Macintosh, like Ready, Set, Go! or LIghtningDraw > Pro, or Creator2. GlyphiX, when released this summer, will be the best diagramming app in the world, bar none. Maury
From: M Rassbach <mark@milestonerdl.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 15 Apr 1998 09:41:36 -0500 Organization: Inc.Net http://www.inc.net Message-ID: <3534C720.3BACAC4@milestonerdl.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> <352D4957.84AE1890@alexa.com> <SCOTT.98Apr10103112@slave.doubleu.com> <3533D6C7.95AECD96@milestonerdl.com> <6h0b4v$h5v$3@ns3.vrx.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Maury Markowitz wrote: > In <3533D6C7.95AECD96@milestonerdl.com> M Rassbach claimed: > > But the GnuStep/MKLinux route won't leave you shafted the way Apple just > > shafted the Newton Developers. > > No, it will just leave you in the forever-limbo of projects that never go > final, or even approach true usability. Really? Wow. Better not say that too loud, as the people who use GCC might actually believe your point of view. > GNUStep, from a user perspective, > does not exist. For the most part it's a mostly-complete FoundationKit and > an incomplete to not-even-really-working everything else. And, because you have source, you don't get shafted. I have YET to see anyone defend Apple that the way they treated the Newton Development community was 'fair', 'just', or 'morally correct'. > Not a single one of my apps will work under GNUStep because the libraries > are not complete. It would appear then that you have a reason to post whatever you feel will benifit Apple Computer, and hope that Apple will continue to develop the environment where your code can exist. Must be nice to have such blind loyality, hoping that Apple will contiune to provide the environment for your code to grow. The Newton Developers had that kind of loyality, and look where they are today. A system bug (-10061) that Apple gives lip-service to fixing. Not to mention, not only a lack of growth of new platform, but not even a medium to continue selling what they have. But if you are thinking of jumping on the Rhapsody bandwagon, look at how Apple treated the Newton Developers and ask yourself if you want to be treated that way.
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 15 Apr 1998 11:49:45 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6h26sp$4fp$1@ns3.vrx.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> <352D4957.84AE1890@alexa.com> <SCOTT.98Apr10103112@slave.doubleu.com> <3533D6C7.95AECD96@milestonerdl.com> <6h0b4v$h5v$3@ns3.vrx.net> <3534C720.3BACAC4@milestonerdl.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: mark@milestonerdl.com In <3534C720.3BACAC4@milestonerdl.com> M Rassbach claimed: [snip your headers, please!] > Really? Wow. Better not say that too loud, as the people who use GCC might > actually believe your point of view. Different project, different state of completeness, different aims, different activity. That's a silly and misleading comparison. > > does not exist. For the most part it's a mostly-complete FoundationKit and > > an incomplete to not-even-really-working everything else. > > And, because you have source, you don't get shafted. Well aside from the fact that you never get product you mean? > I have YET to see anyone defend Apple that the way they treated the Newton > Development community was 'fair', 'just', or 'morally correct'. And I won't try to either. But what does that have to do with GNUStep? > It would appear then that you have a reason to post whatever you feel will > benifit Apple Computer, and hope that Apple will continue to develop the > environment where your code can exist. Ok. > Must be nice to have such blind loyality, hoping that Apple will contiune to > provide the environment for your code to grow. And this is different than the blind loyalty in GNUStep exactly how? The fact that people CAN work on it doesn't mean they DO. And they AREN'T. > But if you are thinking of jumping on the Rhapsody bandwagon, look at how Apple > treated the Newton Developers and ask yourself if you want to be treated that > way. What does that have to do with anything? Personally I wish they would give me $10000 for ever line of OpenStep code I write, but what does THAT have to do with anything. OpenStep works. Now. GNUStep doesn't. Say what you will about Apple or free software, but none of that changes these FACTS. Maury
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer,comp.sys.next.software Subject: Re: 2-D Plotting for OpenStep? Date: 15 Apr 1998 13:11:35 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6h2po7$t30$1@crib.bevc.blacksburg.va.us> References: <35349602.34D58F2A@alum.mit.edu> In article <35349602.34D58F2A@alum.mit.edu>, Eric Hermanson <eric@alum.mit.edu> wrote: > Can anyone recommend the top 3 commercial 2-D plotting packages for OpenStep? > Ideally, the packages would ship with frameworks that developers could use to > incorporate plotting capability into YellowBox and maybe even WebObjects > applications? You probably want to look at the OpenGraph stuff on VVI-DCS's web page: http://www.vvi.com/
From: CatherineW@email.unc.edu Subject: You Just Can't Lose Newsgroups: comp.sys.next.programmer Organization: UNC Message-ID: <35343ad4.0@193.15.242.210> Date: 15 Apr 98 04:43:00 GMT So, do you want to make money? If you have ONLY US$6 or a bit more(6 stamps), you can do it!!! ******************************************************* PLEASE READ ON ABOUT HOW TO BE SUCCESSFUL ******************************************************* 1. IS THIS REALLY LEGAL?? I called a lawyer first. The lawyer was a little skeptical that I would actually make any money but he said it WAS LEGAL if I wanted to try it. I told him it sounded a lot like a chain letter but the details of the system (SEE BELOW) actually made it a legitimate legal business. 2. Would the Post Office be ok with this....I called them: 1-800-725-2161 and they confirmed THIS IS ABSOLUTELY LEGAL! (See Title 18,h sections1302 NS 1341 of Postal Lottery Laws). This clarifies the program of collecting names and addresses for a mailing list. 3. Is this moral? Well, everyone who sends me a buck has a good chance of getting A LOT of money ... a much better chance than buying a lottery ticket!!! ============ HOW IT WORKS ============ Mail the 6 envelopes to the following addresses: STEP 1: Get 6 separate pieces of paper and write the following on each piece of paper: PLEASE ADD ME TO YOUR MAILING LIST. $1 US DOLLAR PROCESSING FEE IN ENCLOSED. (THIS IS KEY AS THIS IS WHAT MAKES IT LEGAL SINCE YOU ARE PAYING FOR AND LATER OFFERING A SERVICE). Now get 6 $1.00 bills and place ONE inside EACH of the 6 pieces of paper so the bill will not be seen through the envelope to prevent theft/robbery (Add one chemical paper for example;this will turn your bill invisible and the letter still light). Then, place one paper in each of the 6 envelopes and seal them. You should now have 6 sealed envelopes, each with a piece of paper stating the above phrase and a U.S. $1.00 bill. #1 Ronald Stevens 20 Lordly court Kings Park NY 11754 USA #2 J. Turley P.O. Box 50604 Knoxville, TN 37950-0604 USA #3 K. Agler 109 Fraternity Ct. Chapel Hill, NC 27516 USA #4 R Sauter 3500 Crescent Ct Flower Mound, TX 75028 USA #5 Catherine Watkins P.O. Box 67 Chapel Hill, NC 27514-0067 USA #6 Ryan 1201 Holleybank Dr. Matthews, NC 28105 USA STEP 2: Now take the #1 name off the list that you see above, move the other names up (6 becomes 5, 5 becomes 4, etc...) and add YOUR Name as number 6 on the list. (If you want to remain anonymous, put a nickname, but the address MUST be correct. It, of course, MUST contain your country, state/district/area, zip code, etc!!! You wouldn't want your money to fly away, would you?!?!). STEP 3: Now post your amended article to at least 200 newsgroups. Remember, 200 postings is just a guideline. The more you post, the more money you make! NOTE: IN MANY NEWSGROUPS THERE ARE PEOPLE THAT DELETE THIS KIND OF MESSAGES I RECOMEND YOU TO POST 1 OR EVEN 2 TIMES A WEEK TO 300 NEWSGROUPS SO OTHER PEOPLE CAN SEE YOUR MESSAGES.THIS IS A WAY TO INCREASE THE POSSIBILITIES FOR YOU TO GET MORE MONEY!! ------------------------------------------------------------------------ The TOP 200 newsgroups can be found at ---> www.op.net/usenet-stats.html *** BOTS *** Bots are small computer programs on a usenet server. 1) Bots look for certain characters in the "Subject:" field of your newsgroup posting. 2) Bots also look for "multiple postings". 3) If a Bot discovers any of the above, it will delete your posting. 4) Then send you a nasty e-mail. ***OUTSMART THE BOTS*** 1) You will make a lot MORE money if you outsmart the BOTS. 2) Post your message only ONCE. 3) Do NOT use characters such as (! $ % + # & * @ ?) in the "Subject" field. ------------------------------------------------------------------------ > ------------------------------------------------------------------------ > **REMEMBER, THE MORE NEWSGROUPS YOU POST IN, THE MORE MONEY YOU > WILL MAKE!! BUT YOU HAVE TO POST A MINIMUM OF 200** > That's it! You will begin receiving money from around the world within > day's! You may eventually want to rent a P.O. Box due to the large > amount of mail you receive. If you wish to stay anonymous, you can > invent a name to use, as long as the postman will deliver it. > **JUST MAKE SURE ALL THE ADDRESSES ARE CORRECT.** >------------------------------------------------------------------------- ================= Now the WHY part: ================= Out of 200 postings, say I receive only 5 replies (a very low example). So then I made $5.00 with my name at #6 on the letter. Now, each of the 5 persons who just sent me $1.00 make the MINIMUM 200 postings, each with my name at #5 and only 5 persons respond to each of the original 5, that is another $25.00 for me, now those 25 each make 200 MINIMUM posts with my name at #4 and only 5 replies each, I will bring in an additional $125.00! Now, those 125 persons turn around and post the MINIMUM 200 with my name at #3 and only receive 5 replies each, I will make an additional $626.00! OK, now here is the fun part, each of those 625 persons post a MINIMUM 200 letters with my name at #2 and they each only receive 5 replies, that just made me $3,125.00!!! Those 3,125 persons will all deliver this message to 200 newsgroups with my name at #1 and if still 5 persons per 200 newsgroups react I will receive $15,625,00! With a original investment of only $6.00! AMAZING! And as I said 5 responses is actually VERY LOW! Average is probable 20 to 30! So lets put those figures at just 15 responses per person. Here is what you will make: at #6 $15.00 at #5 $225.00 at #4 $3,375.00 at #3 $50,625.00 at #2 $759,375.00#1 $11,390,625.00 When your name is no longer on the list, you just take the latest posting in the newsgroups, and send out another $6.00 to names on the list, putting your name at number 6 again. And start posting again. The thing to remember is, do you realize that thousands of people all over the world are joining the internet and reading these articles everyday, JUST LIKE YOU are now!! So can you afford $6.00 and see if it really works?? I think so... People have said, "what if the plan is played out and no one sends you the money? What are the chances of that happening when there are tons of new honest users and new honest people who are joining the internet and newsgroups everyday and are willing to give it a try? Estimates are at 20,000 to 50,000 new users, every day, with thousands of those joining the actual internet. Remember, play FAIRLY and HONESTLY and this will work. You just have to be honest. ** By the way, if you try to deceive people by posting the messages with your name in the list and not sending the money to the rest of the people already on the list, you will NOT get as much. Someone I talked to knew someone who did that and she only made practically nothing, and that's after seven or eight weeks! Then she sent the 6 $1.00 bills, people added her to their lists, and in 4-5 weeks she had over $10k. This is the fairest and most honest way I have ever seen to share the wealth of the world without costing anything but our time!!! You also may want to buy mailing and e-mail lists for future dollars. Make sure you print this article out RIGHT NOW! Also, try to keep a list of everyone that sends you money and always keep an eye on the newsgroups to make sure everyone is playing fairly. Remember, HONESTY IS THE BEST POLICY. You don't need to cheat the basic idea to make the money!! GOOD LUCK to all and please play fairly and reap the huge rewards from this, which is tons of extra CASH. Please remember to declare your extra income. Thanks once again... ________________________________________________________________________ --
From: "Timothy Cheng" <cheng@amug.org> Newsgroups: comp.sys.next.programmer Subject: Win a DIGITAL CAMERA / Provide feedback Date: Wed, 15 Apr 1998 15:48:02 -0700 Organization: AMUG Internet Services Message-ID: <6h3e7e$ika$1@ns2.amug.org> Participate in an online survey, and you will qualify to win a FREE QuickCam DIGITAL CAMERA of your choice! If you use diagramming or modeling applications, simply visit <http://www.strategicinsight.com/EWSurvey.htm>, give us your opinions, and register to win a QuickCam digital camera! It's quick (only 17 questions!), and your opinions will help a software development company to create better tools to meet your needs. All information received is for our data analysis only. Names will be held strictly confidential and will not be sold. We appreciate your help! Responses must be received before April 15, 1998. See survey for full contest rules and eligibility requirements.
From: Henry McGilton <henry@trilithon.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 15 Apr 1998 08:49:21 -0700 Organization: Trilithon Research and Trading Message-ID: <3534D701.D218AB2E@trilithon.com> References: <3534431D.AD9C2AA6@trilithon.com> <B159B890-32953@206.165.43.108> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lawson English wrote: * Henry McGilton <henry@trilithon.com> said: * * very few people "program using DPS". This * * is one of the standard items of mythology * * promulgated by the anti-OpenStep faction. * Are you putting me in that faction? No --- you're in the pro-GX faction. See below. * Why? Because I would prefer to see a different * graphics architecture? Whether or not I have * any valid points to make about DPS vs GX, I * seldom, if ever, knock Objective C and while I'm * not convinced that the Rhapsody Yellow Box * framework will be "perfect," I'm quite willing * to concede that a universally avaible OOP framework * can have MAJOR advantages over Apple's Toolbox-style APIs. There's nothing wrong at all with your wishing a different graphics architecture. That's why I keep encouraging you to go ahead and implement it. The OpenStep Appkit layer and all of the applications written on top thereof were written on the assumption that there's a DPS engine underneath. Such an assumption leads to implementing in a certain fashion. Had the underlying imaging model been X-Swindles, the programming would have been different (and probably wouldn't have been finished yet) and all the kits and apps would depend on that imaging model. As for the anti-OpenStep faction: When OpenStep was first mooted, a bloke from Harris Corp wrote an article on why OpenStep was doomed. One stated reason was that OpsnStep used DPS instead of X. He stated, in a kind of roundabout twisted logic, that DPS's problem was that only one imaging model was used, whereas with X you learned one imaging model (poorly defined) for screen and then you learned PostScript for printing. In his logic, learning two imaging models made X somehow "better". Another item of mythology, as I've pointed out, was that OpenStep programmers have to be PostScript experts to get anything done. I gathered the statistics and published them, and that didn't convince anybody. People who haven't been within a kilometre of OpenStep keep repeating the mythology of having to be expert PostScript hackers to get the work done, and it simply isn't true. You can't easily and in any reasonable timeframe rip out DPS and replace it with seomthing else like GX, however much you want to. You know, I argued this issue with you in December 1996, and you've continued blowing the GX can do more than DPS trumpet ever since. I pointed out, back then, that GX and DPS are at different levels of abstraction. So you could easily build a client-side application kit that implements the GX API, and it would use DPS calls to do the low-level rendering. So go to it --- let's have this studly appkit layer Real Soon Now. As the Yankees say: PUT UP OR SHUT UP. * Directly, you mean. I understand that the DPS engine * can render things faster than GX in many ways. I accept * that. I just wanna have the ability to do in Rhapsody * the kinds of things that I can with GX. Build your GX appkit layer, and at least one programmer will use it. * Why are you defending DPS? Actually, I'm not defending DPS per se. DPS is a tool that I use when appropriate. DPS is the tool that the Appkit layer is built upon. * If DPS is so seldom used in Rhapsody, That is *not* what I said --- cease and desist from twisting the words and the meaning. * why worry as long as Rhaposdy can produce the graphics * that YOU want it to? * Everyone accuses me of being a nut about GX. That's because you're long on opinions about how superior GX is to DPS, but woefully short on real-world application kits that us benighted heathens can use. Come on now, where's your GX AppKit layer? Show us all what a wizard you are. ........ Henry ============================================================= Henry McGilton | Trilithon Software, and, Boulevardier, Java Composer | Pacific Research and Trading -----------------------------+------------------------------- mailto:henry@trilithon.com | http://www.trilithon.com =============================================================
From: trev@sc.edu (Trevor Zion Bauknight) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Wed, 15 Apr 1998 13:07:09 -0400 Organization: Metaphysical Bunko Squad Message-ID: <trev-1504981307100001@nas-sa-p7.usc.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> <352D4957.84AE1890@alexa.com> <SCOTT.98Apr10103112@slave.doubleu.com> <3533D6C7.95AECD96@milestonerdl.com> <6h0b4v$h5v$3@ns3.vrx.net> <3534C720.3BACAC4@milestonerdl.com> In article <3534C720.3BACAC4@milestonerdl.com>, mark@milestonerdl.com wrote: > I have YET to see anyone defend Apple that the way they treated the Newton > Development community was 'fair', 'just', or 'morally correct'. Products get cancelled all the time. Apple should probably fix that one hideous outstanding bug (it there's anyone there who can these days), but that's where I see the "moral" responsibility ending. The Newton will continue to function, probably well past the time when Apple has a replacement product ready. I might not have cancelled a product without already having its replacement ready, but I'm not an expert on Apple's business end of things. > But if you are thinking of jumping on the Rhapsody bandwagon, look at how Apple > treated the Newton Developers and ask yourself if you want to be treated that > way. I see this as the difference between the future of Apple's central product, its OS, and the future of its incidental projects which were not doing well in the market. Big difference. Trev -- http://www.uscsu.sc.edu/~tzbaukni trev@sc.eduEATMOSPAM (unmunge)
From: "NeXT Newbie" <macghod@concentric.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.hardware.misc,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: letter to macnn about developer article Date: 15 Apr 1998 18:38:20 GMT Message-ID: <01bd689d$48e281c0$33f0bfa8@davidsul> References: <01bd67ec$731a7b20$2ef0bfa8@davidsul> <6h238g$e1p$1@gaea.omnigroup.com> You are talking about getting just the cds without the seeds for $200, I take it? I thought this was a option BEFORE this program was introduced as well? From what I have seen the $200 option wasnt just announced by apple, what was announced by apple was a $500 plan and the $3500 plan. Their was NO $200 plan introduced, I assumed it was their all along. I am under the impression the $200 program was their all along, you just didnt know about it. And the "free program" apple announced was also nothing new, it has been in place for quite some time > > No, prices were increased with no new services offered, or services deleted > > > Speak for yourself. The new programs saved me $50 as I don't need the > software seeds. (I have access to them at work, but don't need them at home.) > And once Rhapsody CR1 ships, it'll be $50 easier to get friends to try it as > a development platform.
From: dylan@loop.com Newsgroups: comp.sys.next.programmer Subject: webobjects intro Date: Wed, 15 Apr 1998 14:58:49 -0600 Organization: Deja News - The Leader in Internet Discussion Message-ID: <6h33ho$l0t$1@nnrp1.dejanews.com> Hi: I'm interested in learning how to program in Web Objects. I haven't any programming experience, however. Any suggestions? Please email and post. Thanks! Dylan MacDonald -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: CatherineW@icarus.com Subject: You Just Can't Lose Newsgroups: comp.sys.next.programmer Organization: ME Message-ID: <3534c142.0@193.15.242.210> Date: 15 Apr 98 14:16:34 GMT So, do you want to make money? If you have ONLY US$6 or a bit more(6 stamps), you can do it!!! ******************************************************* PLEASE READ ON ABOUT HOW TO BE SUCCESSFUL ******************************************************* 1. IS THIS REALLY LEGAL?? I called a lawyer first. The lawyer was a little skeptical that I would actually make any money but he said it WAS LEGAL if I wanted to try it. I told him it sounded a lot like a chain letter but the details of the system (SEE BELOW) actually made it a legitimate legal business. 2. Would the Post Office be ok with this....I called them: 1-800-725-2161 and they confirmed THIS IS ABSOLUTELY LEGAL! (See Title 18,h sections1302 NS 1341 of Postal Lottery Laws). This clarifies the program of collecting names and addresses for a mailing list. 3. Is this moral? Well, everyone who sends me a buck has a good chance of getting A LOT of money ... a much better chance than buying a lottery ticket!!! ============ HOW IT WORKS ============ Mail the 6 envelopes to the following addresses: STEP 1: Get 6 separate pieces of paper and write the following on each piece of paper: PLEASE ADD ME TO YOUR MAILING LIST. $1 US DOLLAR PROCESSING FEE IN ENCLOSED. (THIS IS KEY AS THIS IS WHAT MAKES IT LEGAL SINCE YOU ARE PAYING FOR AND LATER OFFERING A SERVICE). Now get 6 $1.00 bills and place ONE inside EACH of the 6 pieces of paper so the bill will not be seen through the envelope to prevent theft/robbery (Add colored paper for example;this will turn your bill invisible and the letter still light). Then, place one paper in each of the 6 envelopes and seal them. You should now have 6 sealed envelopes, each with a piece of paper stating the above phrase and a U.S. $1.00 bill. #1 Ronald Stevens 20 Lordly court Kings Park NY 11754 USA #2 J. Turley P.O. Box 50604 Knoxville, TN 37950-0604 USA #3 K. Agler 109 Fraternity Ct. Chapel Hill, NC 27516 USA #4 R Sauter 3500 Crescent Ct Flower Mound, TX 75028 USA #5 Catherine Watkins P.O. Box 67 Chapel Hill, NC 27514-0067 USA #6 Ryan 1201 Holleybank Dr. Matthews, NC 28105 USA STEP 2: Now take the #1 name off the list that you see above, move the other names up (6 becomes 5, 5 becomes 4, etc...) and add YOUR Name as number 6 on the list. (If you want to remain anonymous, put a nickname, but the address MUST be correct. It, of course, MUST contain your country, state/district/area, zip code, etc!!! You wouldn't want your money to fly away, would you?!?!). STEP 3: Now post your amended article to at least 200 newsgroups. Remember, 200 postings is just a guideline. The more you post, the more money you make! NOTE: IN MANY NEWSGROUPS THERE ARE PEOPLE THAT DELETE THIS KIND OF MESSAGES I RECOMEND YOU TO POST 1 OR EVEN 2 TIMES A WEEK TO 300 NEWSGROUPS SO OTHER PEOPLE CAN SEE YOUR MESSAGES.THIS IS A WAY TO INCREASE THE POSSIBILITIES FOR YOU TO GET MORE MONEY!! ------------------------------------------------------------------------ The TOP 200 newsgroups can be found at ---> www.op.net/usenet-stats.html *** BOTS *** Bots are small computer programs on a usenet server. 1) Bots look for certain characters in the "Subject:" field of your newsgroup posting. 2) Bots also look for "multiple postings". 3) If a Bot discovers any of the above, it will delete your posting. 4) Then send you a nasty e-mail. ***OUTSMART THE BOTS*** 1) You will make a lot MORE money if you outsmart the BOTS. 2) Post your message only ONCE. 3) Do NOT use characters such as (! $ % + # & * @ ?) in the "Subject" field. ------------------------------------------------------------------------ > ------------------------------------------------------------------------ > **REMEMBER, THE MORE NEWSGROUPS YOU POST IN, THE MORE MONEY YOU > WILL MAKE!! BUT YOU HAVE TO POST A MINIMUM OF 200** > That's it! You will begin receiving money from around the world within > day's! You may eventually want to rent a P.O. Box due to the large > amount of mail you receive. If you wish to stay anonymous, you can > invent a name to use, as long as the postman will deliver it. > **JUST MAKE SURE ALL THE ADDRESSES ARE CORRECT.** >------------------------------------------------------------------------- ================= Now the WHY part: ================= Out of 200 postings, say I receive only 5 replies (a very low example). So then I made $5.00 with my name at #6 on the letter. Now, each of the 5 persons who just sent me $1.00 make the MINIMUM 200 postings, each with my name at #5 and only 5 persons respond to each of the original 5, that is another $25.00 for me, now those 25 each make 200 MINIMUM posts with my name at #4 and only 5 replies each, I will bring in an additional $125.00! Now, those 125 persons turn around and post the MINIMUM 200 with my name at #3 and only receive 5 replies each, I will make an additional $626.00! OK, now here is the fun part, each of those 625 persons post a MINIMUM 200 letters with my name at #2 and they each only receive 5 replies, that just made me $3,125.00!!! Those 3,125 persons will all deliver this message to 200 newsgroups with my name at #1 and if still 5 persons per 200 newsgroups react I will receive $15,625,00! With a original investment of only $6.00! AMAZING! And as I said 5 responses is actually VERY LOW! Average is probable 20 to 30! So lets put those figures at just 15 responses per person. Here is what you will make: at #6 $15.00 at #5 $225.00 at #4 $3,375.00 at #3 $50,625.00 at #2 $759,375.00#1 $11,390,625.00 When your name is no longer on the list, you just take the latest posting in the newsgroups, and send out another $6.00 to names on the list, putting your name at number 6 again. And start posting again. The thing to remember is, do you realize that thousands of people all over the world are joining the internet and reading these articles everyday, JUST LIKE YOU are now!! So can you afford $6.00 and see if it really works?? I think so... People have said, "what if the plan is played out and no one sends you the money? What are the chances of that happening when there are tons of new honest users and new honest people who are joining the internet and newsgroups everyday and are willing to give it a try? Estimates are at 20,000 to 50,000 new users, every day, with thousands of those joining the actual internet. Remember, play FAIRLY and HONESTLY and this will work. You just have to be honest. ** By the way, if you try to deceive people by posting the messages with your name in the list and not sending the money to the rest of the people already on the list, you will NOT get as much. Someone I talked to knew someone who did that and she only made practically nothing, and that's after seven or eight weeks! Then she sent the 6 $1.00 bills, people added her to their lists, and in 4-5 weeks she had over $10k. This is the fairest and most honest way I have ever seen to share the wealth of the world without costing anything but our time!!! You also may want to buy mailing and e-mail lists for future dollars. Make sure you print this article out RIGHT NOW! Also, try to keep a list of everyone that sends you money and always keep an eye on the newsgroups to make sure everyone is playing fairly. Remember, HONESTY IS THE BEST POLICY. You don't need to cheat the basic idea to make the money!! GOOD LUCK to all and please play fairly and reap the huge rewards from this, which is tons of extra CASH. Please remember to declare your extra income. Thanks once again... ________________________________________________________________________ --
From: Matthew Cromer <matthew_cromer@iname.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 16 Apr 1998 12:03:20 GMT Organization: NETworthy Distribution: world Message-ID: <6h4s28$575@nrtphc11.bnr.ca> References: <6ghn2b$noq$1@usenet11.supernews.com> <6h46pj$m6$1@new-news.cc.brandeis.edu> <6h47vu$18b$1@new-news.cc.brandeis.edu> <B15B0A02-19BCA8@207.217.155.85> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <rmcassid-1404981222220001@dante.eng.uci.edu> , rmcassid@uci.edu writes: >The biggest problem with education in the US right now is a lack of >dollars. If I could teach for what I make now (which isn't very much) I'd >do it and I think a lot of other people would as well. But teaching isn't >very rewarding when you can't afford housing. Education in the US has a >lot of other problems as well, but dollars will solve some of the most >destructive ones. If money is the answer, then how come American education (much more expensive than any other country) lags so many other countries all of whom spend far less? Any why do we spend so much more today (in real dollars) and get so much less for our money? Matthew Cromer matthew_cromer@iname.com
From: Matthew Cromer <matthew_cromer@iname.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 16 Apr 1998 12:04:31 GMT Organization: NETworthy Distribution: world Message-ID: <6h4s4f$575@nrtphc11.bnr.ca> References: <6ghn2b$noq$1@usenet11.supernews.com> <6h46pj$m6$1@new-news.cc.brandeis.edu> <6h47vu$18b$1@new-news.cc.brandeis.edu> <B15B0A02-19BCA8@207.217.155.85> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <rmcassid-1404981222220001@dante.eng.uci.edu> , rmcassid@uci.edu writes: >Consider that some 30% of all public schools in our nations capitol were >deemed dangerous or unsuitable for occupation this past summer and >couldn't open. When was the last time you heard of a post office or fire >station being unsuitable for occupation? DC, of course, has much higher spending per student than any state or even city government. Matthew Cromer matthew_cromer@iname.com
From: Matthew Cromer <matthew_cromer@iname.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 16 Apr 1998 12:29:15 GMT Organization: NETworthy Distribution: world Message-ID: <6h4tir$6hi@nrtphc11.bnr.ca> References: <6ghn2b$noq$1@usenet11.supernews.com> <6h46pj$m6$1@new-news.cc.brandeis.edu> <6h47vu$18b$1@new-news.cc.brandeis.edu> <B15B0A02-19BCA8@207.217.155.85> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <gmgraves-1504981051040001@sf-usr1-56-184.dialup.slip.net> George Graves, gmgraves@slip.net writes: >1) Cost. The cost of everything has risen all out of proportion to income. > Everything is at least 10X more expensive than it was in 1960, yet > personal income and the tax structure have grown only by about 5X. > This results in a lower standard of living, not only for individuals but > for institutions like schools as well. Actually, the standard of living as measured by material consumption is increasing tremendously--more automobiles per home, bigger TV sets, more TV's, more electronic equipment, more computers, much bigger homes, more homeownership, etc. Incomes are up faster than inflation. And real prices for goods are declining. Matthew Cromer matthew_cromer@iname.com
From: M Rassbach <mark@milestonerdl.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: Thu, 16 Apr 1998 07:05:52 -0500 Organization: Inc.Net http://www.inc.net Message-ID: <3535F420.32659535@milestonerdl.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> <352D4957.84AE1890@alexa.com> <SCOTT.98Apr10103112@slave.doubleu.com> <3533D6C7.95AECD96@milestonerdl.com> <6h0b4v$h5v$3@ns3.vrx.net> <3534C720.3BACAC4@milestonerdl.com> <trev-1504981307100001@nas-sa-p7.usc.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Trevor Zion Bauknight wrote: > In article <3534C720.3BACAC4@milestonerdl.com>, mark@milestonerdl.com wrote: > > > I have YET to see anyone defend Apple that the way they treated the Newton > > Development community was 'fair', 'just', or 'morally correct'. > > Products get cancelled all the time. You bet products get canceled all the time. Not many vendors are your strategic partners. That makes the Apple newton cancellation of bit different. Especially when Apple was the only vendor of the product. > Apple should probably fix that one > hideous outstanding bug (it there's anyone there who can these days), but > that's where I see the "moral" responsibility ending. So telling your business partners not to worry you'll have newton's then canceling the project one week later is moral? > The Newton will > continue to function, probably well past the time when Apple has a > replacement product ready. I might not have cancelled a product without > already having its replacement ready, but I'm not an expert on Apple's > business end of things. > And how will apples new product address the market the message pad 2x00 addressed? What develop or in their right mind would now pick Apple? > > But if you are thinking of jumping on the Rhapsody bandwagon, look at > how Apple > > treated the Newton Developers and ask yourself if you want to be treated that > > way. > > I see this as the difference between the future of Apple's central > product, its OS, and the future of its incidental projects which were not > doing well in the market. Big difference. So you then are of the believe that past actions do not predict the future? I maintain that apples past actions with respect to the newton development predicts the future. One time the Macintosh didn't do well in the market. The Apple ][ line supported the Macintosh for a number of years. How soon you all forget history.
From: Matthew Cromer <matthew_cromer@iname.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 16 Apr 1998 12:21:12 GMT Organization: NETworthy Distribution: world Message-ID: <6h4t3o$63q@nrtphc11.bnr.ca> References: <6ghn2b$noq$1@usenet11.supernews.com> <6h46pj$m6$1@new-news.cc.brandeis.edu> <6h47vu$18b$1@new-news.cc.brandeis.edu> <B15B0A02-19BCA8@207.217.155.85> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <EG2Z.30$p5.231928@news.itd.umich.edu> , not@my.address.net writes: >Let's be honest about the level of funding for inner city public schools. >In a lot of cases the toilets are broken and they can't afford to fix them. >I'm not saying money is the answer. I'm just saying let's be honest about >what they've had to work with. You might be interested to know that the DC public schools, widely considered the worst in the nation, and many of them rat-traps, spends $9,000+ per student per year. Clearly the students are not seeing the benefits of that money--I expect that the admininistrators and their cronies are driving porsches and living in mansions. Private schools manage to do much more with much less money, homeschool works great too. Matthew Cromer matthew_cromer@iname.com
From: Andre-John Mas <ama@fabre.act.qc.ca> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: The Newton [was Re: Apple developer program] Date: Thu, 16 Apr 1998 09:28:15 -0500 Organization: Communications Accessibles Montreal, Quebec Canada Message-ID: <Pine.LNX.3.95.980416091640.2344C-100000@fabre.act.qc.ca> References: <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> <352D4957.84AE1890@alexa.com> <SCOTT.98Apr10103112@slave.doubleu.com> <3533D6C7.95AECD96@milestonerdl.com> <6h0b4v$h5v$3@ns3.vrx.net> <3534C720.3BACAC4@milestonerdl.com> <trev-1504981307100001@nas-sa-p7.usc.net> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII In-Reply-To: <trev-1504981307100001@nas-sa-p7.usc.net> > > But if you are thinking of jumping on the Rhapsody bandwagon, look at > how Apple > > treated the Newton Developers and ask yourself if you want to be treated that > > way. > It is true what Apple did to the Newton Project was not the best thing. I wonder if they had gone ahead in making Newton Inc. whether the Newton could have survived. The Newton was a great product, though it lost what it was trying to achieve: lost-cost PDA device. Instead it was becoming big (not even pocket sized) and expensive (more than $800). In spite of the profit Apple has made, like any company in the computer industry it needs keep lean and do what it is good at. What I would like to see is the technology in the Newton rolled into othe Apple products. One technology is the hand-writing recognition. I can't remember where I read it, though there is a company in the USA that has expressed their intrest in the Newton technology and would like to buy it to keep the Newton alive. I just hope that Apple lets them buy the technology or licences it the for a sensible fee. One thing that I saw the Newton being useful for is Scientific uses. If the Newton had had an adaptor at the top, it would have been great for adding scientific sensory equipment for in the field data collection (sure this sounds like something out of Star Trek, though so did the mobile phone). AJ
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.hardware.misc,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: letter to macnn about developer article Followup-To: comp.sys.mac.advocacy,comp.sys.next.advocacy Date: 16 Apr 1998 14:12:05 GMT Organization: P & L Systems Message-ID: <6h53jl$9sp$10@ironhorse.plsys.co.uk> References: <01bd67ec$731a7b20$2ef0bfa8@davidsul> <6h238g$e1p$1@gaea.omnigroup.com> <01bd689d$48e281c0$33f0bfa8@davidsul> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: macghod@concentric.net NOTE: FOLLOWUPS TO COMP.SYS.MAC.ADVOCACY,COMP.SYS.NEXT.ADVOCACY In <01bd689d$48e281c0$33f0bfa8@davidsul> Mr Sullivan of Santa Clara, CA, aka "NeXT Newbie" wrote: > You are talking about getting just the cds without the seeds for $200, I > take it? > I thought this was a option BEFORE this program was introduced as well? > From what I have seen the $200 option wasnt just announced by apple, what > was announced by apple was a $500 plan and the $3500 plan. Their was NO > $200 plan introduced, I assumed it was their all along. > *If* this is the case (my impression is that it isn't, but corrections welcome) then it is disingenuous of you to claim "prices were increased with no new services offered", since on this item price was not increased. mmalc.
From: "HECTOR" <Hector@cableinet.co.uk> Newsgroups: comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.pc.hardware Subject: TRY THIS AND BE AMAZED!!!!!!!!1 Date: Thu, 16 Apr 1998 16:59:16 +0100 Message-ID: <35362f99.0@ispc-news.cableinet.net> Organization: "Cable Internet (post doesn't reflect views of Cable Internet)" It is really possible! People claim to make almost $50K in just 4 weeks! It may be hard to believe, but read on . . . (For $6 and 6 stamps, it might be worth a try) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ A little while back, I was browsing these newsgroups, just like you are now, and came across an article similar to this that said you could make thousands of dollars within weeks with only an initial investment of $6.00! So I thought, "Yeah, right, this must be a scam!" But like most of us, I was curious. Like most of us, I kept reading. Anyway, it said that if you send $1.00 to each of the 6 names and addresses stated in the article, you could make thousands in a very short period of time. You then place your own name and address at the bottom of the list at #6, and post the article to at least 200 newsgroups. (There are about 22,000.) or e-mail them to friends, or e-mailing lists... No catch, that was it. Even though the investment was a measly $6, I had three questions that needed to be answered before I could get involved in this sort of thing. 1. IS THIS REALLY LEGAL? I called a lawyer first. The lawyer was a little skeptical that I would actually make any money but he said it WAS LEGAL if I wanted to try it. I told him it sounded a lot like a chain letter but the details of the system (SEE BELOW) actually made it a legitimate legal business. 2. Would the Post Office be ok with this...? I called them: 1-800-725-2161 and they confirmed THIS IS ABSOLUTELY LEGAL! (See Title 18,h sections 1302 NS 1341 of Postal Lottery Laws). This clarifies the program of collecting names and addresses for a mailing list. 3. Is this moral? Well, everyone who sends me a buck has a good chance of getting A LOT of money ... a much better chance than buying a lottery ticket! So, having these questions answered, I invested EXACTLY $7.92 ... six $1.00 bills and six 32 cent postage stamps ... and boy am I glad I did! Within 7 days, I started getting money in the mail! I was shocked! I still figured it would end soon, and didn't give it another thought. But the money just continued coming in. In my first week, I made about $20.00 to $30.00 dollars. By the end of the second week I had a mad total of $1,000.00! In the third week I had over $10,000.00 and it was still growing. This is now my fourth week and I have made a total of just over $42,000.00 and it's still coming in..... It's certainly worth $6.00 and 6 stamps! So now I'm reposting this so I can make even more money! The *ONLY* thing stopping *ANYONE* from enriching their own bank account is pure laziness! It took me all of 5 MINUTES to print this out, follow the directions, and begin posting to newsgroups. It took me a mere 45 minutes to post to over 200 newsgroups. And for this GRAND TOTAL investment of $ 7.92 (US) and under ONE HOUR of my time, I have reaped an incredible amount of money -- like nothing I've ever even heard of anywhere before! 'Nuff said! Let me tell you how this works, and most importantly, why it works. Also, make sure you print a copy of this article now, so you can get the information off of it when you need it. The process is very simple and consists of THREE easy steps. ============ HOW IT WORKS ============ STEP 1: ------ Get 6 separate pieces of paper and write the following on each piece of paper: PLEASE ADD ME TO YOUR MAILING LIST. $1 US DOLLAR PROCESSING FEE IS ENCLOSED. Your name and address. (THIS IS KEY AS THIS IS WHAT MAKES IT LEGAL SINCE YOU ARE PAYING FOR AND LATER OFFERING A SERVICE). Now get 6 $1.00 bills and place ONE inside EACH of the 6 pieces of paper so the bill will not be seen through the envelope to prevent theft/robbery. Then, place one paper in each of the 6 envelopes and seal them. You should now have 6 sealed envelopes, each with a piece of paper stating the above phrase and an U.S. $1.00 bill. Mail the 6 envelopes to the following addresses: 1 - K. Uppermann 2330 Vehicle Dr. #148 Rancho Cordova, CA 95670 2- Occupant P.O. Box 5258 Quincy, Il. 62305-5258 3- M Afandi M Saman No. 9, Lorong 8, Taman Mewah 08000 Sungai Petani Kedah, Malaysia 4- J Wilhelm 3507 Del Lago Cir #305 Tampa FL 33614 5- Y.H. Sohn C.P.O. Box 7469, Chung-Ku, Seoul, Korea 100-674 6- Hector 108 Poolstock Wigan Lancs United Kingdom WN3 5EW STEP 2: Now take the #1 name off the list that you see above, move the other names up (6 becomes 5, 5 becomes 4, etc.) and add YOUR Name as number 6 on the list. (If you want to remain anonymous put a nickname, but the address MUST be correct. It, of course, MUST contain your country, state/district/area, zip code, etc! You wouldn't want your money to fly away, would you?). STEP 3: Now post your amended article to at least 200 newsgroups. Remember that 200 postings are just a guideline. The more you post, the more money you make! Don't know HOW to post in the news groups? Well do exactly the following: ------------------------------------------------------------------------ HOW TO POST TO NEWSGROUPS FAST WITH YOUR WEB BROWSER: The fastest way to post a newsletter: Highlight and COPY (Ctrl-C) the text of this posted message and PASTE (Ctrl-V) it into a plain text editor (as Wordpad) and save it. After you have made the necessary changes that are stated above, simply COPY (Ctrl-C) and PASTE (Ctrl-V) the text into the message composition window, after selecting a newsgroup, and post it! (Or you can attach the file, without writing anything to the message window.) ------------------------------------------------------------------------ If you have Netscape Navigator 3.0 do the following: 1. Click on any newsgroup like normal, then click on 'TO NEWS'. This will bring up a box to type a message in. 2. Leave the newsgroup box like it is, change the subject box to something flashy, something to catch the eye, as "$$$ NEED CASH $$$? READ HERE! $! $!$" Or "$$$! MAKE FAST CASH, YOU CAN'T LOSE! $$$". Or you can use my subject title. 3. Now click on 'ATTACHMENTS'. Then click on 'ATTACH FILE'. Find your file on your Hard Disk (the one you saved from the text editor). Once you find it, click on it and then click 'OPEN' and 'OK'. You should now see your file name in the attachments box. 4. Now click on 'SEND'/'POST'. You see? Now you just have 199 to go! (Don't worry, it's easy and quick once you get used to it.) NOTE: All the versions of Netscape Navigator's are similar to each other, so you'll have no problem to do this if you don't have Netscape Navigator 3.0. ------------------------------------------------------------------------ ! QUICK TIP! (For Netscape Navigator 3.x and above) You can post this message to many newsgroups at a time, by simply selecting a newsgroup near the top of the screen, hold down the SHIFT, and then select a newsgroup near the bottom of the screen. All of the newsgroups in/between will be selected. After that, you follow/do the basic steps, stated below at this letter, except of step #1. You can go to the page stated below in this letter and click on a newsgroup to open up the newsgroups window. Once you've done this, in the same window go to 'OPTIONS', and then mark 'SHOW ALL NEWSGROUPS' and 'SHOW ALL MESSAGES'. Now you can see all the newsgroups and you can apply easier the above tip. ------------------------------------------------------------------------ If you have MS Internet Explorer do the following: 1. Go to the newsgroups and press 'POST AN ARTICLE'. To the new window type your headline in the subject area and then click in the large window below. There either PASTE your letter (which it's been copied from the text editor), or attach the file which contains it. 2. Then click on 'SEND' or 'OK'. NOTE: All versions of MS Internet Explorer are similar to each other, so you won't have any problem doing this. GENERAL NOTES ON POSTING: A nice page where you'll find all the newsgroups if you want help is http://www.liszt.com/ (When you go to the home page, click on the link 'Newsgroup Directory'). But I don't think you'll have any problem posting because it's very easy once you've found the newsgroups. All these web browsers are similar. It doesn't matter which one you have. (But it makes it very easy if you have Netscape Navigator 3.0 or later. You may download it from the Internet if you don't have it.) You just have to remember the basic steps, stated below. BASIC STEPS FOR POSTING: 1. Find a newsgroup and you click on it. 2. You click on 'POST AN/NEW ARTICLE' or 'TO NEWS' or anything else similar to these. 3. You type your flashy headline in the subject box. 4. Now, either you attach the file containing your amended letter, or you PASTE the letter. (You have to COPY it from the text editor, of course, from before.) 5. Finally, you click on 'SEND' or 'POST' or 'OK', whatever is there. ------------------------------------------------------------------------ **REMEMBER, THE MORE NEWSGROUPS YOU POST IN, THE MORE MONEY YOU WILL MAKE! BUT YOU HAVE TO POST A MINIMUM OF 200** That's it! You will begin receiving money from around the world within days! You may eventually want to rent a P.O.Box due to the large amount of mail you receive. If you wish to stay anonymous, you can invent a name to use, as long as the postman will deliver it. **JUST MAKE SURE ALL THE ADDRESSES ARE CORRECT. ** ------------------------------------------------------------------------- ANOTHER EXCELLENT PROGRAM: "AutoPoster" http://www.spck.se/bulk/ MULTIPLE POSTING ON HUNDEREDS OF NEWSGROUPS AT ONCE . The worlds most effective AutoPoster to Internets newsgroups! A high speed Newsgroup Auto Poster/Newsreader for Windows 95 Autopost articles to more than 27,000 newsgroups Super fast - up to 12,000 posting per hour on a single PC Easy to run Built in Special Functions such as: Scramble Random Fake Sender Rotate Unlimited Postings Group Postings Force Hits ------------------------------------------------------------------------- ================= Now the WHY part: ================= Out of 200 postings; say I receive only 5 replies (a very low example). So then I made $5.00 with my name at #6 on the letter. Now, each of the 5 persons who just sent me $1.00 make the MINIMUM 200 postings, each with my name at #5 and only 5 persons respond to each of the original 5, that is another $25.00 for me, now those 25 each make 200 MINIMUM posts with my name at #4 and only 5 replies each, I will bring in an additional $125.00! Now, those 125 persons turn around and post the MINIMUM 200 with my name at #3 and only receive 5 replies each, I will make an additional $626.00! OK, now here is the fun part, each of those 625 persons post a MINIMUM 200 letters with my name at #2 and they each only receive 5 replies, that just made me $3,125.00! Those 3,125 persons will all deliver this message to 200 newsgroups with my name at #1 and if still 5 persons per 200 newsgroups react I will receive $15,625,00! With an original investment of only $6.00! AMAZING! And as I said 5 responses is actually VERY LOW! Average are probable 20 to 30! So lets put those figures at just 15 responses per person. Here is what you will make: at #6 $15.00 at #5 $225.00 at #4 $3,375.00 at #3 $50,625.00 at #2 $759,375.00 at #1 $11,390,625.00 When your name is no longer on the list, you just take the latest posting in the newsgroups, and send out another $6.00 to names on the list, putting your name at number 6 again. And start posting again. The thing to remember is, do you realize that thousands of people all over the world are joining the internet and reading these articles everyday, JUST LIKE YOU are now! So can you afford $6.00 and see if it really works? I think so... People have said, "what if the plan is played out and no one sends you the money? So what! What are the chances of that happening when there are tons of new honest users and new honest people who are joining the internet and newsgroups everyday and are willing to give it a try? Estimates are at 20,000 to 50,000 new users, every day, with thousands of those joining the actual Internet. Remember, play FAIRLY and HONESTLY and this will work. You just have to be honest. By the way, if you try to deceive people by posting the messages with your name in the list and not sending the money to the rest of the people already on the list, you will NOT get as much. Someone I talked to knew someone who did that and he only made about $150.00, and that's after seven or eight weeks! Then he sent the 6 $1.00 bills, people added him to their lists, and in 4-5 weeks he had over $10k. This is the fairest and most honest way I have ever seen to share the wealth of the world without costing anything but our time! You also may want to buy mailing and e-mail lists for future dollars. Make sure you print this article out RIGHT NOW, also. Try to keep a list of everyone that sends you money and always keep an eye on the newsgroups to make sure everyone is playing fairly. Remember that HONESTY IS THE BEST POLICY. You don't need to cheat the basic idea to make the money! GOOD LUCK to all and please play fairly and reap the huge rewards from this, which is tons of extra CASH. Please remember to declare your extra income. Thanks once again... ===================================================================== LEGAL? ? ? (Comments from Bob Novak who started this new version.) "People have asked me if this is really legal. Well, it is! You are using the Internet to advertise you business. What is that business? You are assembling a mailing list of people who are interested in home based computer and online business and methods of generating income at home. Remember that people send you a small fee to be added to your mailing list. It is legal. What will you do with your list of thousands of names? Compile all of them into a database and sell them as "Mailing Lists" on the internet in a similar manner, if you wish, and make more money. How do you think you get all the junk mail that you do? Credit card companies, mail order, Utilities, anyone you deal with through the mail can sell your name and address on a mailing list, unless you ask them not to, in addition to there regular business, So, why not do the same with the list you collect. You can find more info about "Mailing Lists" on the internet using any search engine..." So, build your mailing list, keep good accounts, declare the income and pay your taxes. By doing this you prove your business intentions. Keep an eye on the newsgroups and when the cash has stopped coming (that means your name is no longer on the list), you just take the latest posting at the newsgroups, send another $6.00 to the names stated on the list, make your corrections (put your name at #6) and start posting again. ===================================================================== NOTES: *1. In some countries, the export of the country's exchange is illegal. But you can get the license to do this from the post office, explaining the above statements (that you have an online business, etc. You may have to pay an extra tax, but that's OK, the amount of the incoming money is HUGE! And as I said, a few countries have that restriction. *2. You may want to buy mailing and e-mail lists for future dollars. (Or Database or Spreadsheet software.) *3. If you're really not sure or still think this can't be for real, please print a copy of this article and pass it along to someone who really needs the money, and see what happens. *4. You will start getting responses within 1-2 weeks, it depends.
From: ttnnbfla@news.net Newsgroups: comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin Subject: Do you want ALL 30,000 + Newsgroups ? Date: 16 Apr 1998 18:46:39 GMT Organization: Strictly Uniform Message-ID: <6h5jmf$gm3$1724@nclient5-gui.server.virgin.net> If your ISP censors your newsgroup access (eg you are not getting at least 30,000 newsgroups) visit http://207.120.227.53/usenet or our fast mirror at http://199.93.4.135/homegarden/web/usenet/index.htm
From: ttnnbfla@news.net Newsgroups: comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin Subject: Uncensored News Group Access Date: 16 Apr 1998 18:46:41 GMT Organization: Strictly Uniform Message-ID: <6h5jmh$gm3$1725@nclient5-gui.server.virgin.net> If your ISP censors your newsgroup access (eg you are not getting at least 30,000 newsgroups) visit http://207.120.227.53/usenet or our fast mirror at http://199.93.4.135/homegarden/web/usenet/index.htm
From: ttnnbfla@news.net Newsgroups: comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin Subject: 30,000 Newsgroups including 'banned' ones available Date: 16 Apr 1998 18:46:37 GMT Organization: Strictly Uniform Message-ID: <6h5jmd$gm3$1723@nclient5-gui.server.virgin.net> If your ISP censors your newsgroup access (eg you are not getting at least 30,000 newsgroups) visit http://207.120.227.53/usenet or our fast mirror at http://199.93.4.135/homegarden/web/usenet/index.htm
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: The Newton [was Re: Apple developer program] Date: 16 Apr 1998 12:26:00 -0700 Organization: Primenet Services for the Internet Message-ID: <B15BAA68-17DA9@206.165.43.169> References: <Pine.LNX.3.95.980416091640.2344C-100000@fabre.act.qc.ca> To: "Andre-John Mas" <ama@fabre.act.qc.ca> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Andre-John Mas <ama@fabre.act.qc.ca> said: > > I can't remember where I read it, though there is a company in the USA > that has expressed their intrest in the Newton technology and would like > to buy it to keep the Newton alive. I just hope that Apple lets them buy > the technology or licences it the for a sensible fee. > That bid is pretty much dead in the water. Apparently, Apple has asked them to pay $50 million just to *license* the NewtonOS, not purchase it. ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: han9kin@secssa.samsung.co.kr (Suh Sang-Hyuk) Newsgroups: comp.sys.next.programmer Subject: Stopping current RunLoop? Date: 16 Apr 1998 19:53:18 GMT Organization: DACOM Internet Service Message-ID: <6h5nje$t75$1@newstmp1.bora.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi. I'm writing a server program using socket. First, I made a socket and accept in background. When a client connected, my server program will make a thread to service. In this thread, after initializing itself, receiving client's request with this code. [[NSRunLoop currentRunLoop] run]; If the client send a request, and server thread will receive it, and response it. But, at this time, if response ends, How can I stop currentRunLoop's running? My code is structured as following: <thread started> <initializing itself> <set observer self to NSNotificationCenter to be notified by socket's receiving> [[NSRunLoop currentRunLoop] run]; // then, reponse to client's request! <If response ends, this line must be referenced!!> <I want to stop NSRunLoop's running to instruct to this line!> <release some variables, and deallocing self> In implementation of timeout, the same technic is required. I cannot stop the RunLoop! to stop to execute the next lines. and terminating the thread safely! How can I do it? I need your help! Thanks in advance.
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <35343ad4.0@193.15.242.210> Control: cancel <35343ad4.0@193.15.242.210> Date: 15 Apr 1998 21:39:48 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.35343ad4.0@193.15.242.210> Sender: CatherineW@email.unc.edu Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Greg_Anderson@afs.com (Gregory H. Anderson) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 16 Apr 1998 21:28:40 GMT Organization: Anderson Financial Systems Inc. Message-ID: <6h5t68$3mb@shelob.afs.com> References: <SCOTT.98Apr15232615@slave.doubleu.com> Scott Hess writes > Draw is a object drawing program. If a drawing program only needs %.5 > PS code to run under OpenStep, imagine how much a non-drawing program > will need! Better get that Red Book, eh? In the past, I have posted similar statistics for PasteUp (page layout) and WriteUp (word processor). Some people just don't want to believe. -- Gregory H. Anderson | "We're in the land of the blind, Visionary Ophthalmologist | selling working eyeballs, and they Anderson Financial Systems | balk at the choice of color." -- Tony greg@afs.com (NeXTmail OK) | Lovell, on Mac user reactions to NeXT
From: Robert Forsyth <bobbyf@forsee.tcp.co.uk> Newsgroups: comp.soft-sys.nextstep,comp.sys.mac.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Newby needs help Date: Thu, 16 Apr 1998 21:50:35 +0100 Organization: Total Connectivity Providers - Maximising the Internet Message-ID: <35366F1B.98331849@forsee.tcp.co.uk> References: <01bc46e5$6b4880a0$24f0bfa8@davidsul> <6god24$lq4$1@news.rt66.com> <01bd6585$58835660$0b0ba8c0@woohoo> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Earl Malmrose wrote: > > Jake Garcia <jakeg@rt66.com> wrote in article > <6god24$lq4$1@news.rt66.com>... > > Doesn't the standard Windows 95 filesystem limit file/dir names to 14 > chars? > > I could be wrong... > > You're wrong. Standard Win95 filesystem supports filenames whose complete > file and path length is up to around 260 characters. Yes, but, doesn't Openstep only see the MSDOS 8.3 name. Isn't ppp on the Openstep CD-ROM. If you download to Windows 95, move/copy the files to openstep ( e.g. /tmp for the rest of this paragraph). Move/Rename the files in the /tmp so that they have the extension that they would have had and has been mangled by Window 95 is you wish to decompress in Workspace. Use gnutar instead of tar; move of these files will be gzip-ed tar archives and will have an extension such as .tar.gz, .gtz, etc., Anyway, use: gnutar xvzf <archivename> the 'z' will gunzip the tar, before extracting.
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Organization: Is a sign of weakness Message-ID: <SCOTT.98Apr15231250@slave.doubleu.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> <352D4957.84AE1890@alexa.com> <SCOTT.98Apr10103112@slave.doubleu.com> <3533D6C7.95AECD96@milestonerdl.com> <6h0b4v$h5v$3@ns3.vrx.net> <3534C720.3BACAC4@milestonerdl.com> In-reply-to: M Rassbach's message of Wed, 15 Apr 1998 09:41:36 -0500 Date: 16 Apr 98 20:34:12 GMT In article <3534C720.3BACAC4@milestonerdl.com>, M Rassbach <mark@milestonerdl.com> writes: But if you are thinking of jumping on the Rhapsody bandwagon, look at how Apple treated the Newton Developers and ask yourself if you want to be treated that way. Didn't this start when you took issue with my suggestion that GNUStep isn't really an alternative to Rhapsody in many very fundamental ways? Well, then, what do you think the odds are on GNUStep if Rhapsody _doesn't_ ship, or Apple kills it off? There are a lot of orphan GUI toolkits out there. For better or worse, Rhapsody is the center of the OpenStep universe. If Rhapsody does well, GNUStep will do alright. If Rhapsody digs a hole, GNUStep won't be far behind. [Actually, if Rhapsody digs a hole, I'd have to question the point of making an OpenStep compliant library. I'd take the good things, and start off in my own direction, rather than slavishly making the same mistakes.] -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Organization: Is a sign of weakness Message-ID: <SCOTT.98Apr15232615@slave.doubleu.com> References: <6h0114$bnb$2@ns3.vrx.net> <B1598B35-5BAB4@206.165.43.216> <3534431D.AD9C2AA6@trilithon.com> In-reply-to: Henry McGilton's message of Tue, 14 Apr 1998 22:18:21 -0700 Date: 16 Apr 98 20:34:13 GMT In article <3534431D.AD9C2AA6@trilithon.com>, Henry McGilton <henry@trilithon.com> writes: Lawson English wrote: * Other people program using DPS very few people "program using DPS". This is one of the standard items of mythology promulgated by the anti-OpenStep faction. The amount of *PostScript* programming necessary in any run-of-the-mill OpenStep application is down around the one percent region. I posted a long letter to this effect to one of the UNIX rags a few years back (trying to rebut the dissemination of FUD at the time) and they refused to publish it. I will attempt to dig up the statistics I gathered at the time that demonstrated how little *PostScript* programming is required in OpenStep applications. Well, some quick stats: Objc/C code PSwrap code Stuart 26klines 26 lines TickleServices 22klines 0 lines <private> 75klines 16 lines By the way, the last is what I'm currently working on for a client, and about 20klines of it is in essence a custom version of an Edit-level word processor. [Put another way, it's not leveraging the Text object, it's doing the font and line layout from the ground up. Didn't use the Text object because there's still need for a Windows3.1 version of the UI.] Of course, there's another good indicator: Objc/C code PSwrap code Draw 15klines 75 lines Draw is a object drawing program. If a drawing program only needs %.5 PS code to run under OpenStep, imagine how much a non-drawing program will need! Better get that Red Book, eh? Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6h3e7e$ika$1@ns2.amug.org> Control: cancel <6h3e7e$ika$1@ns2.amug.org> Date: 15 Apr 1998 23:52:28 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6h3e7e$ika$1@ns2.amug.org> Sender: "Timothy Cheng" <cheng@amug.org> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: don.brown@cesoft.com (Donald Brown) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: The Newton [was Re: Apple developer program] Date: Thu, 16 Apr 1998 22:25:27 -0500 Organization: CE Software Message-ID: <MPG.fa0879291c2e1d49898b7@news.supernews.com> References: <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> <352D4957.84AE1890@alexa.com> <SCOTT.98Apr10103112@slave.doubleu.com> <3533D6C7.95AECD96@milestonerdl.com> <6h0b4v$h5v$3@ns3.vrx.net> <3534C720.3BACAC4@milestonerdl.com> <trev-1504981307100001@nas-sa-p7.usc.net> <Pine.LNX.3.95.980416091640.2344C-100000@fabre.act.qc.ca> In article <Pine.LNX.3.95.980416091640.2344C-100000@fabre.act.qc.ca>, ama@fabre.act.qc.ca says... > > > But if you are thinking of jumping on the Rhapsody bandwagon, look at > > how Apple > > > treated the Newton Developers and ask yourself if you want to be treated that > > > way. > > > > It is true what Apple did to the Newton Project was not the best thing. I > wonder if they had gone ahead in making Newton Inc. whether the Newton > could have survived. The Newton was a great product, though it lost what > it was trying to achieve: lost-cost PDA device. Instead it was becoming > big (not even pocket sized) and expensive (more than $800). > The point that was made, that you may not have heard, was that only weeks before Newton was killed, Jobs sent a message to the Newton developers saying that not spinning the Newton out was a good thing and that Apple was so confident of it and particularly the eMate and that life was wonderful and keep the faith baby! There are two scarey possibilities. (A) It was known at the time that the Newton was dead, Job lied through his teeth, KNEW developers would soon learn he lied through his teeth, but didn't care. (B) Within a couple of weeks, Apple went from go-go-go on the Newton to tossing it out with the dishwater. I'm not sure which speaks worse for Apple management. Donald
From: nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) Newsgroups: comp.sys.next.programmer Subject: suppressing warnings from cc ? Date: 17 Apr 1998 02:42:26 GMT Organization: none Message-ID: <6h6fii$45o$2@ha2.rdc1.nj.home.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit If these are, as they appear, harmless, is there a way to suppress them? (or are they something that I should look into?) In file included from make.h:49, from main.c:19: /NextDeveloper/Headers/bsd/sys/types.h:119: warning: useless keyword or type name in empty declaration /NextDeveloper/Headers/bsd/sys/types.h:119: warning: empty declaration /NextDeveloper/Headers/bsd/sys/types.h:120: warning: useless keyword or type name in empty declaration /NextDeveloper/Headers/bsd/sys/types.h:120: warning: empty declaration /NextDeveloper/Headers/bsd/sys/types.h:136: warning: useless keyword or type name in empty declaration /NextDeveloper/Headers/bsd/sys/types.h:136: warning: empty declaration -- [Yes that is a valid mail address, until it gets spammed] Unofficial @Home FAQ: http://members.home.com/faqs/
From: sal@panix3.panix.com (Salvatore Denaro) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: The Newton [was Re: Apple developer program] Date: 17 Apr 1998 04:22:03 GMT Organization: PANIX Public Access Internet and Unix, NYC Message-ID: <slrn6jdm7b.3u2.sal@panix3.panix.com> References: <Pine.LNX.3.95.980416091640.2344C-100000@fabre.act.qc.ca> <B15BAA68-17DA9@206.165.43.169> On 16 Apr 1998 12:26:00 -0700, Lawson English <english@primenet.com> wrote: >That bid is pretty much dead in the water. Apparently, Apple has asked them >to pay $50 million just to *license* the NewtonOS, not purchase it. With the Newton teams gone, and most of the remaing ones sent to other areas, could Apple even fulfill a request like that? I'd be willing to give 5 to 1 odds that the Newt is dead forever. -- sal@panix.com Salvatore Denaro "The reality of the software business today is that if you find something that can make you ridiculously rich, then that's something that Microsoft is going to take away from you." -- Max Metral
From: "Ziya Oz" <ziyaoz@earthlink.net> Newsgroups: comp.sys.next.programmer Subject: Re: WebObjects 4.0? Date: Fri, 17 Apr 1998 05:18:41 -0400 Organization: BM Message-ID: <6h7632$bnr@argentina.earthlink.net> References: <352C2138.4835AB9F@alum.mit.edu> Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit In article <352C2138.4835AB9F@alum.mit.edu>, Eric Hermanson <eric@alum.mit.edu> wrote: > Does anyone know the approximate timeframe for the release of WebObjects 4.0 (most interested in the NT platform)? Septtemberish.
From: holger@object-factory.REMOVETHIS.com (Holger Hoffstaette) Newsgroups: comp.sys.next.programmer Subject: Re: Stopping current RunLoop? Date: 17 Apr 1998 08:45:11 GMT Organization: the unstoppable code machine Message-ID: <6h74qn$bhi$1@leonie.object-factory.com> References: <6h5nje$t75$1@newstmp1.bora.net> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Cc: han9kin@secssa.samsung.co.kr (also sent via email) Suh Sang-Hyuk wrote: > Hi. > I'm writing a server program using socket. > First, I made a socket and accept in background. > When a client connected, my server program will make a thread to service. > In this thread, after initializing itself, receiving client's request with > this code. > [[NSRunLoop currentRunLoop] run]; > If the client send a request, and server thread will receive it, and > response it. > But, at this time, if response ends, How can I stop currentRunLoop's running? > ... The following might help: --snip-- >From: ckane@next.com (Christopher Kane)>Newsgroups: comp.sys.next.programmer >Date: 8 Sep 1997 01:05:17 GMT >Subject: Re: Stopping NSRunLoops In article <340C94E7.6BE2@ergotech.com> Jim Redman <jim@ergotech.com> writes: > I feel that this should be a fairly dumb question, but just how do you > stop an NSRunLoop that has been started with a -run? Is it just in the > documentation and I missed it, or is there something clever about this? -run will keep iterating on an NSRunLoop as long as the run loop has sources of input, i.e. timers and ports. "ports" includes other things in Foundation too, which do background activities, such as reading from a file handle in the background or getting the exit notification for a spawned NSTask, which register ports with the appropriate run loop to trigger their notification in the invoking thread. Note that D.O. also works via ports. If you have control of all input sources (i.e. know all NSPorts that you've explicitly registered with the run loop, and know all D.O. connections, and know all timers, NSFileHandles, NSTasks, ...), you can invalidate or release these objects, which will remove their registered ports from the run loop as a side effect. Send -invalidate to all explicit NSPorts and NSConnections, remove timers explicitly from the run loop or cause them to stop otherwise, release retained NSFileHandles and NSTasks, etc. It might be simpler to run the run loop yourself and not use -run. Use -runUntilDate: or -runMode:beforeDate: in a loop which checks some external condition for when the thread should terminate, like: while (condition) { if (![[NSRunLoop currentRunLoop] runMode:NSDefaultRunLoopMode beforeDate:[NSDate dateWithTimeIntervalSinceNow:1.0]]) break; } This is essentially what -run is doing, except it uses [NSDate distantFuture]. Here, if -runMode:beforeDate: either return NO immediately, if the run loop has no input sources, or will block for up to 1 second waiting for something to happen (or less if something does happen). The condition could be some boolean flag with the thread knows about, for example. You don't want to make the condition too expensive to evaluate. Christopher Kane Application Frameworks Apple Computer, Inc. --snip-- good luck! Holger -- Holger Hoffstaette - holger"at"object-factory.com Object Factory GmbH - http://www.object-factory.com/
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: The Newton [was Re: Apple developer program] Date: 17 Apr 1998 01:23:00 -0700 Organization: Primenet Services for the Internet Message-ID: <B15C60A6-226FF@206.165.43.169> References: <slrn6jdm7b.3u2.sal@panix3.panix.com> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit > On 16 Apr 1998 12:26:00 -0700, Lawson English <english@primenet.com> > wrote: > >That bid is pretty much dead in the water. Apparently, Apple has asked > them > >to pay $50 million just to *license* the NewtonOS, not purchase it. > > With the Newton teams gone, and most of the remaing ones sent to other > areas, could Apple even fulfill a request like that? > > I'd be willing to give 5 to 1 odds that the Newt is dead forever. > You're likely right, but I'd give those same 5 to 1 odds that a goodly number of the original team members would quit their current jobs in a heartbeat if they knew that good money was sitting behind the new NewtonOS company (Planet Computing?) and would work VERY hard to make NewtonOS work as a non-Apple technology. ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: Thom McDonald <t.a.mcd@ix.netcom.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: The Newton Date: Fri, 17 Apr 1998 06:02:25 +0000 Organization: Median Productions Message-ID: <3536F06A.D23EA427@ix.netcom.com> References: <slrn6jdm7b.3u2.sal@panix3.panix.com> <B15C60A6-226FF@206.165.43.169> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Newton isn't entirely dead. It should be fairly obvious that Apple wants to retain Newton technologies for incorporating into a 'Lite' Mac OS which will be a Newton/PalmPilot/Windows CE-like platform. The good things about Newton, like touch-screen operation and hand-writing recognition will make nice additions to Mac OS. Of course this wouldn't be a totally public stance because if they got a really big offer, maybe they wouldn't want to turn it down. Tom t.a.mcd@ix.netcom.com
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <3534c142.0@193.15.242.210> Control: cancel <3534c142.0@193.15.242.210> Date: 16 Apr 1998 11:52:56 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.3534c142.0@193.15.242.210> Sender: CatherineW@icarus.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) Newsgroups: comp.sys.next.programmer Subject: GNU find 4.1 vm_allocate errors Date: 17 Apr 1998 12:35:18 GMT Organization: none Message-ID: <6h7ia6$fj9$3@ha2.rdc1.nj.home.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Using regular cc from NS3.3 I can build GNU find 4.1 but when it tries to do a long listing including several subdirs, it gives this error: memory allocation error: vm_allocate failed find: WHATEVERPATH: Not enough memory I built it on PEAK last night with no problem (BSDi machine) so this has to be something NeXT-specific (posix would be my guess). All else failing does anyone have a working GNU find for Intel/m68k? I deleted my old binary after 4.1 compiled successfully (DOOH!) TjL -- [Yes that is a valid mail address, until it gets spammed] Unofficial @Home FAQ: http://members.home.com/faqs/
From: Žoloft@hegel1.cs.chalmers.se.cs.chalmers.se (Olof Torgersson) Newsgroups: comp.sys.next.programmer Subject: Deploying on Windows95 Date: 17 Apr 1998 14:09:54 GMT Organization: Chalmers University of Technology Message-ID: <6h7nri$ipi$1@nyheter.chalmers.se> Hi When I try to run my applications developed on OpenStep Enterprise 4.2 using Windows/NT on Windows95 nothing works. I have tested three different applications and all three of them died before anything was shown on the screen. TextEdit and ther other demo apps run fine so the OpenStep installation seems ok. Also, my applications work on Windows/NT and OpenStep/Mach. Any ideas on what might be wrong? Olof Torgersson oloft@cs.chalmers.se
From: merx@pc.chemie.tu-darmstadt.de (Hendrik Merx) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: The Newton [was Re: Apple developer program] Date: Fri, 17 Apr 1998 19:09:45 +0200 Organization: Darmstadt University of Technology Message-ID: <merx-1704981909450001@komtur.pc.chemie.tu-darmstadt.de> References: <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <gmgraves-0804981131450001@sf-pm5-10-74.dialup.slip.net> <6ggmk2$jdl$1@news.digifix.com> <352D4957.84AE1890@alexa.com> <SCOTT.98Apr10103112@slave.doubleu.com> <3533D6C7.95AECD96@milestonerdl.com> <6h0b4v$h5v$3@ns3.vrx.net> <3534C720.3BACAC4@milestonerdl.com> <trev-1504981307100001@nas-sa-p7.usc.net> <Pine.LNX.3.95.980416091640.2344C-100000@fabre.act.qc.ca> In article <Pine.LNX.3.95.980416091640.2344C-100000@fabre.act.qc.ca>, Andre-John Mas <ama@fabre.act.qc.ca> wrote: > One thing that I saw the Newton being useful for is Scientific uses. If > the Newton had had an adaptor at the top, it would have been great for > adding scientific sensory equipment for in the field data collection (sure > this sounds like something out of Star Trek, though so did the mobile > phone). It rather sounds like the old HP-41 ;-). Hendrik Hendrik Merx Darmstadt University of Technology merx@pc.chemie.tu-darmstadt.de http://www.pc.chemie.tu-darmstadt.de/authors/merx/
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin Subject: cmsg cancel <6h5jmh$gm3$1725@nclient5-gui.server.virgin.net> Control: cancel <6h5jmh$gm3$1725@nclient5-gui.server.virgin.net> Date: 16 Apr 1998 18:51:26 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6h5jmh$gm3$1725@nclient5-gui.server.virgin.net> Sender: ttnnbfla@news.net Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin Subject: cmsg cancel <6h5jmf$gm3$1724@nclient5-gui.server.virgin.net> Control: cancel <6h5jmf$gm3$1724@nclient5-gui.server.virgin.net> Date: 16 Apr 1998 18:51:19 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6h5jmf$gm3$1724@nclient5-gui.server.virgin.net> Sender: ttnnbfla@news.net Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin Subject: cmsg cancel <6h5jmd$gm3$1723@nclient5-gui.server.virgin.net> Control: cancel <6h5jmd$gm3$1723@nclient5-gui.server.virgin.net> Date: 16 Apr 1998 18:51:18 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6h5jmd$gm3$1723@nclient5-gui.server.virgin.net> Sender: ttnnbfla@news.net Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Message-ID: <CrEETNm1HZY$@cc.usu.edu> From: "Howard R. Cole" <edx@cc.usu.edu> Date: 17 Apr 98 09:34:59 MDT References: <SCOTT.98Apr15232615@slave.doubleu.com> <6h5t68$3mb@shelob.afs.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit In <6h5t68$3mb@shelob.afs.com> Gregory H. Anderson wrote: > Scott Hess writes > > Draw is a object drawing program. If a drawing program only needs %.5 > > PS code to run under OpenStep, imagine how much a non-drawing program > > will need! Better get that Red Book, eh? > > In the past, I have posted similar statistics for PasteUp (page layout) > and WriteUp (word processor). Some people just don't want to believe. On the other hand, I've written a satellite-tracking/sensor-modeling utility (Rendezvous.app) which draws maps, 3D views as seen from the field-of-view of the satellite. trajectories of comets/asteroids through the solar system, and displays of current sensor focal plane array pixel state. This required beaucoup custom postscript code. About 10% of the sixty thousand or so total lines is postscript for drawing those custom views. Of course my application is a little out of the mainstream for most OpenStep applications (boy is that an understatement), and your point that the vast majority of OpenStep apps will require very little or no postscript coding is generally true. - HRC -
From: "Josef Wiesehoefer" <J.Wiesehoefer@kiel.netsurf.de> Newsgroups: comp.sys.next.programmer Subject: YellowBox - Database capabilities Date: Fri, 17 Apr 1998 20:01:59 +0200 Organization: CLS Internet Services GmbH Message-ID: <6h85br$g5c$1@freeside.cls.net> Mime-Version: 1.0 Content-Type: text/plain; charset="ISO-8859-1" Content-Transfer-Encoding: 7bit I send this message again, because I think it didn't reach the server. I am going to develop a bibliographical database for my university. Because we have Macs and Windows PCs, the application has to support both operating systems. My bugdet is very low, so I cannot afford any expensive database solutions. My question: Is some kind of database functionality built into the Yellow box or do I have to use the Enterprise Objects? I do not need relational capabilities, but the database should be able to manage 10,000s of records. Can anybody help me?? Thomas Wiesehoefer I'm sorry for my bad English.
From: Matthew Cromer <matthew_cromer@iname.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 17 Apr 1998 19:53:08 GMT Organization: NETworthy Distribution: world Message-ID: <6h8bv4$66a@nrtphc11.bnr.ca> References: <6glhks$ros$1@news01.deltanet.com> <merx-1704981909450001@komtur.pc.chemie.tu-darmstadt.de> <6h838o$57p$1@ironhorse.plsys.co.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <B15BAE84-274F0@206.165.43.169> Lawson English, english@primenet.com writes: >And the format for Macintosh OS volumes and files didn't change >signficantly during that time, either, so of course it wasn't a major >impediment. I'm pretty sure that he didn't create Stuffit during the >MFS=>HFS transition, and that was the last major change for the MacOS >until HFS+. > HFS vs HFS+ will not disrupt applications that operate on a file basis with the filesystem. The only notable incompatibilities are with programs that bypass the MacOS file system. Matthew Cromer matthew_cromer@iname.com
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 17 Apr 1998 14:13:00 -0700 Organization: Primenet Services for the Internet Message-ID: <B15D1508-574C5@206.165.43.62> References: <6h8bv4$66a@nrtphc11.bnr.ca> nntp://news.primenet.com/comp.sys.mac.oop.powerplant, nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.advocacy, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Matthew Cromer <matthew_cromer@iname.com> said: > > > HFS vs HFS+ will not disrupt applications that operate on a file basis > with the filesystem. > > The only notable incompatibilities are with programs that bypass the > MacOS file system. I rather suspect that developers of an application that compresses files need to have access to the most up-to-date version of hte OS, especially when the file-format radically changes. ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 17 Apr 1998 21:10:32 GMT Organization: Idiom Communications Message-ID: <6h8gg8$99h$1@news.idiom.com> References: <6h0114$bnb$2@ns3.vrx.net> <B1598B35-5BAB4@206.165.43.216> <3534431D.AD9C2AA6@trilithon.com> <SCOTT.98Apr15232615@slave.doubleu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: scott@doubleu.com Scott Hess may or may not have said: -> Draw is a object drawing program. If a drawing program only needs %.5 -> PS code to run under OpenStep, imagine how much a non-drawing program -> will need! Better get that Red Book, eh? Something that's even more telling than that: wc printPackage.ps 193 1037 8160 printPackage.ps wc windowPackage.ps 1962 5384 35806 windowPackage.ps So, all the ps code that implements the windowing system was done in 1,962 lines. All the ps code for printing took up only 193 lines. BTW, I haven't run the diffs, but AFAIK, neither of these files has been altered since around the time of NeXTStep 1.0. -jcr
From: LaserPointers.com Newsgroups: comp.sys.next.programmer Date: Fri, 17 Apr 1998 12:15:16 PDT Subject: Check it out! Organization: Email Platinum v.3.1b Message-ID: <3537d48e.0@bonaparte.pixi.com> Visit http://www.LaserPointers.com Check out our site for the smallest, least expensive pointers on the market! For only $19.50 plus shipping and handling, you can own one! It measures only 2inches by 9mm!, yet it can send a brilliant red dot up to 1200 feet, even in the brightest of rooms! To place you order, visit our home page at: http://www.LaserPointers.com LaserPointer.com staff
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: YellowBox - Database capabilities Date: 17 Apr 1998 20:38:37 GMT Organization: Digital Fix Development Message-ID: <6h8ekd$591$1@news.digifix.com> References: <6h85br$g5c$1@freeside.cls.net> In-Reply-To: <6h85br$g5c$1@freeside.cls.net> On 04/17/98, "Josef Wiesehoefer" wrote: >I send this message again, because I think it didn't reach the server. > >I am going to develop a bibliographical database for my university. Because >we have Macs and Windows PCs, the application has to support both operating >systems. My bugdet is very low, so I cannot afford any expensive database >solutions. >My question: Is some kind of database functionality built into the Yellow >box or do I have to use the Enterprise Objects? I do not need relational >capabilities, but the database should be able to manage 10,000s of records. My guess is that the best, free, alternative would be Rhapsody using AIAT... this assumes that AIAT is in the first release of Rhapsody.. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: YellowBox - Database capabilities Date: 17 Apr 1998 17:29:28 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6h8hjo$1up$1@crib.bevc.blacksburg.va.us> References: <6h85br$g5c$1@freeside.cls.net> <6h8ekd$591$1@news.digifix.com> In article <6h8ekd$591$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: > >I am going to develop a bibliographical database for my university. > My guess is that the best, free, alternative would be Rhapsody > using AIAT... this assumes that AIAT is in the first release of > Rhapsody.. Hmm, it never occurred to me to use a full-text index of ASCII flatfiles as a read-only database alternative.. I wonder how performance would compare. Two questions: Does anyone have guesses about the likelihood of AIAT making it into the first release of Rhapsody is? (I know they're promised it eventually..) I'm looking for a really good, free full-text indexing solution under Rhapsody. And also, does anyone know how it compares in features/performance/size to alternatives like freeWAIS, or quasi-text indexes like Glimpse? Does it do regular expressions? I didn't see answers to those questions on Apple's site. (I suppose I should cross-post this to a Mac newsgroup, but I don't know which one would be appropriate.)
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: Finding gamma of the display Newsgroups: comp.sys.next.programmer Message-ID: <3537f2b7.0@news.depaul.edu> Date: 18 Apr 98 00:24:23 GMT Is there a way to programmatically determine the gamma adjustment on a given output device? I believe it's typically a fixed value on OpenStep, but adjustable on Rhapsody/PPC. (Anyone know what the value is on OpenStep?) I'm wondering because of discrepancies between the RGB values for 'Netscape colors' across platforms. I took a Fractal Design Painter color set of Netscape colors, and converted it to an OpenStep NSColorList. When compared to the NeXT-provided set of Netscape colors, they didn't match up (in color or RGB values). I'm assuming that the NeXT Netscape color set was created under OpenStep's gamma settings, while Painter's version was created under Mac or Windows gamma settings. Similar-looking colors would then have different RGB values.
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 17 Apr 1998 22:11:03 GMT Organization: MiscKit Development Message-ID: <6h8k1n$m43$1@news.xmission.com> References: <6h0114$bnb$2@ns3.vrx.net> <B1598B35-5BAB4@206.165.43.216> <3534431D.AD9C2AA6@trilithon.com> <SCOTT.98Apr15232615@slave.doubleu.com> <6h8gg8$99h$1@news.idiom.com> jcr.remove@this.phrase.idiom.com (John C. Randolph) wrote: > Scott Hess may or may not have said: > -> Draw is a object drawing program. If a drawing program only needs %.5 > -> PS code to run under OpenStep, imagine how much a non-drawing program > -> will need! Better get that Red Book, eh? > > Something that's even more telling than that: > > wc printPackage.ps > 193 1037 8160 printPackage.ps > wc windowPackage.ps > 1962 5384 35806 windowPackage.ps > > So, all the ps code that implements the windowing system was done in 1,962 > lines. All the ps code for printing took up only 193 lines. That's a good reason to pick up the red book, too. Become a PS guru and you, too, can hack the window server code. You can change to point to focus instead of click to focus (get thee behind me, Satan) and other wacky tricks. BTW, there's also some PSWrap code in the ApplicationKit, but there's no way to count that without having the sources. I doubt that it is all that much of the code, percentage wise, though. In fact, it is probably miniscule. To some degree representative, percentage wise, would probably be the NeXT 3DKit, a rather graphically intensive kit by its own right. It has 8800 lines, of which four are in a PSWrap. The rest is Objective-C. > BTW, I haven't run the diffs, but AFAIK, neither of these files has been > altered since around the time of NeXTStep 1.0. They have, but not by much. Look at the modification dates on the files: NeXTSTEP 3.3: /usr/lib/NextStep 2 -r--r--r-- 1 root wheel 1991 Jan 15 1993 nlpPrintPackage.ps 8 -r--r--r-- 1 root wheel 8074 Mar 17 1993 printPackage.ps 2 -r--r--r-- 1 root wheel 1431 Feb 24 1992 printing.ps 35 -r--r--r-- 1 root wheel 35706 Oct 20 1994 windowPackage.ps OPENSTEP 4.2: /usr/lib/NextStep 2 -r--r--r-- 1 root wheel 1991 Jan 15 1993 nlpPrintPackage.ps 8 -r--r--r-- 1 root wheel 8074 Mar 17 1993 printPackage.ps 2 -r--r--r-- 1 root daemon 1431 Feb 24 1992 printing.ps /NextLibrary/Frameworks/AppKit.framework/Resources 8 -r--r--r-- 1 root wheel 8160 Oct 17 1996 printPackage.ps 35 -r--r--r-- 1 root wheel 35806 Mar 28 1997 windowPackage.ps The latter two were moved from the original location to become part of the framework, and it looks like minor modifications were made to things...like the copyright date. :-) % diff /usr/lib/NextStep/printPackage.ps /NextLibrary/Frameworks/AppKit.framework/Resources/printPackage.ps 2,4c2,7 < % NeXT Printing Package < % Version: 3.1 < % Copyright: 1988, NeXT, Inc. --- > % > % printPackage.ps > % Application Kit > % Copyright (c) 1988-1996, NeXT Software, Inc. > % All rights reserved. > % 14a18,19 > 0 0 defineuserobject > 1 null defineuserobject If I check out the differences between the 3.3 and 4.2 windowPackage.ps I get this: 128c128 < /version 2 def --- > /version 4 def 580a581 > /Courier findfont setfont 1057c1058 < 1 index -16 bitshift --- > 1 index -16 bitshift 255 and 1106c1107,1108 < dup 0 ne --- > activeApp //CDKitVersion /currentContextData winexec > 4 lt 1 index 0 ne and Looks like they fixed a few minor bugs. -- Later, -Don Yacktman don@misckit.com <a href="http://www.misckit.com/don.html">My home page</a>
From: nobody@nowhere33.yet Newsgroups: comp.sys.newton.programmer,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.palmtops Subject: Gorgeous teenager wants you to fuck her asshole until it bleeds! 95978 Date: Friday, 17 Apr 1998 22:57:05 -0600 Organization: <no organization> Distribution: World Message-ID: <17049822.5705@nowhere33.yet> Check out our free assortment of MASSIVE HARDCORE XXXTREME PORN PICS FREE!!! http://209.50.232.37/harris/ 2k0qxK73(GBi6`jcw$=b
From: systems@caribsurf.com Newsgroups: comp.sys.newton.programmer,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.palmtops Subject: ^&* HEATHER LOCKLER from the MERLOSE PLASE is in swin-suit. *&^ 98278 Date: Friday, 17 Apr 1998 23:01:36 -0600 Organization: <no organization> Distribution: World Message-ID: <17049823.0136@caribsurf.com> http://xxx-18.com/xxx/ dTl%``JJy.Dt<rBaua]R
From: nobody@nowhere33.yet Newsgroups: comp.sys.newton.programmer,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.palmtops Subject: Can you believe a pussy could strech so far? 44332 Date: Friday, 17 Apr 1998 23:16:09 -0600 Organization: <no organization> Distribution: World Message-ID: <17049823.1609@nowhere33.yet> Free Asian Pics All Fetishes 500 Images for FREE viewing (and saving) No fees whatsoever! Check it out http://sexplosion.com/windex/ 4$<N00ssIWmDeBk1E1-"
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: YellowBox - Database capabilities Date: 18 Apr 1998 04:26:06 GMT Organization: Digital Fix Development Message-ID: <6h9a0u$ed2$1@news.digifix.com> References: <6h85br$g5c$1@freeside.cls.net> <6h8ekd$591$1@news.digifix.com> <6h8hjo$1up$1@crib.bevc.blacksburg.va.us> In-Reply-To: <6h8hjo$1up$1@crib.bevc.blacksburg.va.us> On 04/17/98, Nathan Urban wrote: >In article <6h8ekd$591$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: > >> >I am going to develop a bibliographical database for my university. > >> My guess is that the best, free, alternative would be Rhapsody >> using AIAT... this assumes that AIAT is in the first release of >> Rhapsody.. > >Hmm, it never occurred to me to use a full-text index of ASCII flatfiles >as a read-only database alternative.. I wonder how performance would >compare. > Well, having read the AIAT stuff a long time ago, I seem to recall that it also allows you to have structured data as well.. Anyone with actual experience? >Two questions: Does anyone have guesses about the likelihood of >AIAT making it into the first release of Rhapsody is? (I know they're >promised it eventually..) I'm looking for a really good, free full-text >indexing solution under Rhapsody. And also, does anyone know how it >compares in features/performance/size to alternatives like freeWAIS, >or quasi-text indexes like Glimpse? Does it do regular expressions? >I didn't see answers to those questions on Apple's site. (I suppose I >should cross-post this to a Mac newsgroup, but I don't know which one >would be appropriate.) > The AIAT information is on Apple's site. There are links from here http://developer.apple.com/sdk/index.html -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: nobody@nowhere33.yet Newsgroups: comp.sys.newton.programmer,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.palmtops Subject: FUCKING and SUCKING! This gal has her PUSSY and her MOUTH full! 7314 Date: Saturday, 18 Apr 1998 00:02:58 -0600 Organization: <no organization> Distribution: World Message-ID: <18049800.0258@nowhere33.yet> 400 FREE PICS 48 FREE 5 MINUTE MPEG MOVIES WANT TO DOWNLOAD INSTED OF VIEW, WE HAVE A 3 GIG PILE OF PICS TO DOWNLOAD... TAKE ALL YOU WANT AND VIEW WHENEVER NEVER A CHARGE http://209.50.232.37/frank/ 341:ymyTji,27)SmaE&*
From: wmurphy@fox.nstn.ca Newsgroups: comp.sys.newton.programmer,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.palmtops Subject: SOME LIKE IT DRIPPING WETT !!! @@@@ 73185 Date: Saturday, 18 Apr 1998 13:12:55 -0600 Organization: <no organization> Distribution: World Message-ID: <18049813.1255@fox.nstn.ca> Here are some pics of my College Days!!! http://xxx-18.com/xxxhardcore/ 8Rs<]SAw./,4thtOed&-
From: nobody@nowhere33.yet Newsgroups: comp.sys.newton.programmer,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.palmtops Subject: Can you believe a pussy could strech so far? 86330 Date: Saturday, 18 Apr 1998 13:16:06 -0600 Organization: <no organization> Distribution: World Message-ID: <18049813.1606@nowhere33.yet> Free Asian Pics All Fetishes 500 Images for FREE viewing (and saving) No fees whatsoever! Check it out http://sexplosion.com/windex/ YJasVU??n$9i2g8VkWRH
From: nobody@nowhere33.yet Newsgroups: comp.sys.newton.programmer,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.palmtops Subject: Hot face with a hard cock fucking it good! 57530 Date: Saturday, 18 Apr 1998 13:17:00 -0600 Organization: <no organization> Distribution: World Message-ID: <18049813.1700@nowhere33.yet> RED RAW PUSSY CLOSEUPS UNBELIEVABLY HIGH REZ PICS THE BEST PART...ALL FREE http://209.50.232.37/harris/ @0HY<;%&TcxPqNw=Q=8.
From: sys0p.bbs@bbs.me.ncku.edu.tw Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k Subject: Little GANG BANG Action HERE !! 11941 Message-ID: <18049815.2552@bbs.me.ncku.edu.tw> Date: Saturday, 18 Apr 1998 15:25:52 -0600 Organization: <no organization> Distribution: World My Nympho Wife Needs another Man, can you help? http://xxx-18.com/wilks/ u(\1q3:>0[4hMF2UE5M^
From: q.bbs@bbs.scu.edu.tw Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k Subject: HOT oriental chick takes a black cock!! 15169 Message-ID: <18049815.0900@bbs.scu.edu.tw> Date: Saturday, 18 Apr 1998 15:09:00 -0600 Organization: <no organization> Distribution: World Here are some pics of my College Days!!! http://xxx-18.com/wilks/ W+>4SNuAk^Wk/ImXh8/a
From: u0710@ms6.hinet.net Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k Subject: GORGEOUS BLONDE GERMAN WITH THE NICEST TITS IN THE world 16521 Message-ID: <18049815.4316@ms6.hinet.net> Date: Saturday, 18 Apr 1998 15:43:16 -0600 Organization: <no organization> Distribution: World Hardcore ASS Screwing!!! Tons of Anal Action. http://xxx-18.com/wilks/ X,?5T8^CU_Xl1JWYj90c
From: valdez@columbian.coffee.net Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k Subject: Gay Male Sex Pics 29385 Message-ID: <18049815.5904@columbian.coffee.net> Date: Saturday, 18 Apr 1998 15:59:04 -0600 Organization: <no organization> Distribution: World 4-18-98 to 4-22-98...MASTERBATION MARATHON!!! Come see it live!!! Participants Welcome!!! http://xxx-18.com/wilks/ +8k)'CIN@k,5sTX"S[r+
From: howardk@iswest.com (Howard Knight) Organization: Internet Specialties West, Inc. Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k Subject: cmsg cancel <18049815.0900@bbs.scu.edu.tw> Date: 18 Apr 1998 10:20:45 GMT Control: cancel <18049815.0900@bbs.scu.edu.tw> Message-ID: <cancel.18049815.0900@bbs.scu.edu.tw> Sender: q.bbs@bbs.scu.edu.tw Spam cancelled. Autocancel spam type: spam
From: howardk@iswest.com (Howard Knight) Organization: Internet Specialties West, Inc. Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k Subject: cmsg cancel <18049815.2552@bbs.me.ncku.edu.tw> Date: 18 Apr 1998 10:23:08 GMT Control: cancel <18049815.2552@bbs.me.ncku.edu.tw> Message-ID: <cancel.18049815.2552@bbs.me.ncku.edu.tw> Sender: sys0p.bbs@bbs.me.ncku.edu.tw Spam cancelled. Autocancel spam type: spam
From: howardk@iswest.com (Howard Knight) Organization: Internet Specialties West, Inc. Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k Subject: cmsg cancel <18049815.4316@ms6.hinet.net> Date: 18 Apr 1998 10:26:29 GMT Control: cancel <18049815.4316@ms6.hinet.net> Message-ID: <cancel.18049815.4316@ms6.hinet.net> Sender: u0710@ms6.hinet.net Spam cancelled. Autocancel spam type: spam
Newsgroups: comp.sys.next.programmer From: dfevans@bbcr.uwaterloo.ca (David Evans) Subject: Re: Finding gamma of the display Message-ID: <892918938.222793@globe.uwaterloo.ca> Sender: news@watserv3.uwaterloo.ca Organization: University of Waterloo References: <3537f2b7.0@news.depaul.edu> Cache-Post-Path: globe.uwaterloo.ca!unknown@bcr11.uwaterloo.ca Date: Sat, 18 Apr 1998 17:04:21 GMT In article <3537f2b7.0@news.depaul.edu>, Jonathan W Hendry <jhendry@shrike.depaul.edu> wrote: >Is there a way to programmatically determine the gamma adjustment >on a given output device? > >I believe it's typically a fixed value on OpenStep, but adjustable >on Rhapsody/PPC. (Anyone know what the value is on OpenStep?) > I know that there's an operator called something like "settransferfunction" that sets the gamma; there's likely an analogue that retrieves it. Brian Willoughby (sp?) is the resident gamma expert (well, aside from Mike Paquette, of course!) so hopefully he'll chime in. Alas I don't have the developer docs installed on my machine at the moment so I can't help much beyond this. -- David Evans (NeXTMail/MIME OK) dfevans@bbcr.uwaterloo.ca Computer/Synth Junkie http://bbcr.uwaterloo.ca/~dfevans/ University of Waterloo "Default is the value selected by the composer Ontario, Canada overridden by your command." - Roland TR-707 Manual
Newsgroups: comp.sys.next.programmer From: dfevans@bbcr.uwaterloo.ca (David Evans) Subject: Re: Finding gamma of the display Message-ID: <892928475.749663@globe.uwaterloo.ca> Sender: news@watserv3.uwaterloo.ca Organization: University of Waterloo References: <3537f2b7.0@news.depaul.edu> <892918938.222793@globe.uwaterloo.ca> Cache-Post-Path: globe.uwaterloo.ca!unknown@bcr11.uwaterloo.ca Date: Sat, 18 Apr 1998 19:43:19 GMT In article <892918938.222793@globe.uwaterloo.ca>, David Evans <dfevans@bbcr.uwaterloo.ca> wrote: > I know that there's an operator called something like "settransferfunction" >that sets the gamma; there's likely an analogue that retrieves it. OK. I re-installed the developer docs (I've only had this 2GB disk for six months...) and found "settransfer" and "currenttransfer". Hmmm. Let's see... gallifrey:/Users/dfevans> pft Connection to PostScript established. currenttransfer pstack {} = --nostringval-- currentcolortransfer pstack {} {} {} {} so this basically tells us bugger-all. Guess I wasn't of much help! -- David Evans (NeXTMail/MIME OK) dfevans@bbcr.uwaterloo.ca Computer/Synth Junkie http://bbcr.uwaterloo.ca/~dfevans/ University of Waterloo "Default is the value selected by the composer Ontario, Canada overridden by your command." - Roland TR-707 Manual
From: raceend@letterbox.com (Raceend) Newsgroups: comp.sys.next.programmer Subject: Ziptool - programmers wanted! Date: Sat, 18 Apr 1998 20:36:53 GMT Organization: Customer of Globe Internet NV Message-ID: <3538f835.5298587@nntp.z09.glo.be> Hi Are there programmers who want to make the following simple win 95 program for me? The meaning is to zip multiple files to apart zipfiles. I have hundred of fonts and I want them all in apart zipfiles to make them ready for downloading. How it should work: - I select a source path; - I select some filenames; - I select a target path; - I press a button ("zip apart"); All the selected files of the source path are zipped apart with the same filename, to the target path. E.g. c:\windows\fonts\ times.ttf nuptial.ttf scrawl.ttf mystical.ttf "Zip Apart" e:\fonts\ times.zip nuptial.zip scrawl.zip mystical.zip It should be useful if you can filter the files from the source path. I'm mean that you can type e.g. *.ttf or a*.* or a??a????.* etc. The found files are shown in an apart box. Now I can select them to zip them apart. It would be great too if I can add a textfile to every zipfile. Many thanx in advance! PS: Your name will be added to the acknowledgements list which you can find at: http://titan.glo.be/~gd33771/barclaey.html ----------------------------------------------- Bart Claeys Fontasia International http://titan.glo.be/~gd33771/fontasia.html -----------------------------------------------
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <3537d48e.0@bonaparte.pixi.com> Control: cancel <3537d48e.0@bonaparte.pixi.com> Date: 17 Apr 1998 22:15:44 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.3537d48e.0@bonaparte.pixi.com> Sender: LaserPointers.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: HOT XXXXXX Newsgroups: comp.sys.next.programmer Date: Fri, 17 Apr 1998 20:19:03 PDT Subject: VISIT NOW HOT XXXXX Organization: Email Platinum v.3.1b Message-ID: <35380bf0.1@news.codenet.net> no tricks fees or bullshit just visit once if you dont think there is enough free pics dont come back. PICK FROM 25 CATAGORIES CONTAINING OVER 50,000 FREE PICS!!!!!! OVER 18 (21 IN SOME PLACES) JUST CLICK HERE ONCE AND SEE 4 YOURSELF!!!!!!! http://omega.hostings.net/teens.html http://omega.hostings.net/celebs.html no pop up, trick banner bullshit
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <5697892353623@digifix.com> Date: 19 Apr 1998 03:50:14 GMT Organization: Digital Fix Development Message-ID: <23877892958423@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1993. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: sysrq.bbs@bbs.hwh.edu.tw Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k Subject: YESSS THAT'S IT BABY CUM ALL OVER MY WET HOT PUSSY 51212 Date: Sunday, 19 Apr 1998 17:15:01 -0600 Organization: <no organization> Distribution: World Message-ID: <19049817.1501@bbs.hwh.edu.tw> http://cunts.dyn.ml.org/buttbangers/ U;/;oD+GMRDnG#`YEiXH
From: 1079@1079.com Newsgroups: comp.sys.next.programmer Subject: WIN $100 CASH www.myED.com Date: Sat, 18 Apr 1998 22:41:32 PDT Organization: 1462 Message-ID: <6hc21j$ing$54@camel21.mindspring.com> FREE WEBSITES TO THE PUBLIC BUILD IT WITH OUR WEBSITE BUILDER AND WIN $100.00! BEST SITE OUT OF FIRST 50 WINS http://www.myED.com Thank You, The Everything Directory
From: u@canguess.com Newsgroups: comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.oric,comp.sys.palmtops Subject: ^&* HEATHER LOCKLER from the MERLOSE PLASE is in swin-suit. *&^ 79597 Date: Sunday, 19 Apr 1998 14:07:27 -0600 Organization: <no organization> Distribution: World Message-ID: <19049814.0727@canguess.com> http://xxx-18.com/testy2/ Xd?UTpv"l>XL0jn9iq0B
From: 564@564.com Newsgroups: comp.sys.next.programmer Subject: Business Website $9.95/mo www.myED.com Date: Sat, 18 Apr 1998 22:50:14 PDT Organization: 1101 Message-ID: <6hc2dt$7i0$54@camel25.mindspring.com> Business Websites created online. $9.95/mo Free listings NAICS Codes World Time Currency Exchange Rates Business Profiles http://www.myED.com Thank you, The Everything Directory
From: nobody@nowhere33.yet Newsgroups: comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.oric,comp.sys.palmtops Subject: Annie loves to masturbate in her dorm room... 32960 Date: Sunday, 19 Apr 1998 00:24:42 -0600 Organization: <no organization> Distribution: World Message-ID: <19049800.2442@nowhere33.yet> Check out our free assortment of MASSIVE HARDCORE XXXTREME PORN PICS FREE!!! http://hotteens.dyn.ml.org/frank/ .;o,+FLQCn/"`AEh@H_q
Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k Subject: cmsg cancel <19049817.1501@bbs.hwh.edu.tw> ignore no reply Control: cancel <19049817.1501@bbs.hwh.edu.tw> Message-ID: <cancel.19049817.1501@bbs.hwh.edu.tw> Date: Sun, 19 Apr 1998 08:34:03 +0000 Sender: sysrq.bbs@bbs.hwh.edu.tw From: andrew@erlenstar.demon.co.uk Organization: Annihilator v0.3 Spam (EMP) cancelled - type=NXBOT
Newsgroups: comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.oric,comp.sys.palmtops Subject: cmsg cancel <19049814.0727@canguess.com> ignore no reply Control: cancel <19049814.0727@canguess.com> Message-ID: <cancel.19049814.0727@canguess.com> Date: Sun, 19 Apr 1998 08:33:55 +0000 Sender: u@canguess.com From: andrew@erlenstar.demon.co.uk Organization: Annihilator v0.3 Spam (EMP) cancelled - type=NXBOT
From: synnoveah@earthlink.net Newsgroups: comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.oric,comp.sys.palmtops Subject: CLICK HERE!!! 78902 Date: Sunday, 19 Apr 1998 16:53:01 -0600 Organization: <no organization> Distribution: World Message-ID: <19049816.5301@earthlink.net> Here are some pics of my College Days!!! http://hotmama.dyn.ml.org/eddy/analfuckpics/ Wd?USouzl>XK0jn8iq/A
From: nobody@nowhere33.yet Newsgroups: comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.oric,comp.sys.palmtops Subject: God Damn!! A Father Sticks his tongue in his daughter's ass hole!! 27049 Date: Sunday, 19 Apr 1998 01:50:53 -0600 Organization: <no organization> Distribution: World Message-ID: <19049801.5053@nowhere33.yet> 400 FREE PICS 48 FREE 5 MINUTE MPEG MOVIES WANT TO DOWNLOAD INSTED OF VIEW, WE HAVE A 3 GIG PILE OF PICS TO DOWNLOAD... TAKE ALL YOU WANT AND VIEW WHENEVER NEVER A CHARGE http://buttsluts.dyn.ml.org/frank/ a5I>^Y'Lvhbv:SxcsB:l
From: nobody@nowhere33.yet Newsgroups: comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.oric,comp.sys.palmtops Subject: MERRY CUM-MAS!! 35144 Date: Sunday, 19 Apr 1998 02:08:55 -0600 Organization: <no organization> Distribution: World Message-ID: <19049802.0855@nowhere33.yet> INTENSE HARDCORE IMAGES OF SLUTTY LESBIANS LICKING CUNTS AND ASSHOLES RAW RAW RAW http://tightpussy.dyn.ml.org/michelle/ i=PEe`.S$pi$AZ&jzJAs
From: stefan.boehringer_REMOVE_ME_@ruhr-uni-bochum.de (Stefan Boehringer) Newsgroups: comp.sys.next.programmer Subject: NSData and VM-poking on mach Date: 19 Apr 1998 12:27:32 GMT Organization: Ruhr-Universitaet Bochum, Rechenzentrum Message-ID: <6hcqjk$t6h$1@sun579.rz.ruhr-uni-bochum.de> I was hoping that NSMutableData would restrict itself to writing out dirty pages when being instantiated as mappedFile on calling writeToFile:atomically: (though I must admit that the method syntax suggests otherwise). In contrast writeToFile:atomically: truncates the file to 0 and rewrites it entirely. Obviously this obliterates any strategy to change small amount of a NSData which is mapped to a file and afterwards using writeToFile:atomically: to nail them down. Now my next hope was to have a simple category implementation to mimic the behaviour longed for. Since I can easily find out the vm-memory region involved (by calling the bytes method) and afterwards scrutinizing it for dirty pages. However I could not find an obvious way to do just that. Does anybody know how one could tell a dirty page from a non-dirty one? No mach-function seems to exist. Thank you. Best wishes, Stefan
From: sal@panix3.panix.com (Salvatore Denaro) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: The Newton Date: 19 Apr 1998 15:23:36 GMT Organization: PANIX Public Access Internet and Unix, NYC Message-ID: <slrn6jk5no.rln.sal@panix3.panix.com> References: <slrn6jdm7b.3u2.sal@panix3.panix.com> <B15C60A6-226FF@206.165.43.169> <3536F06A.D23EA427@ix.netcom.com> On Fri, 17 Apr 1998 06:02:25 +0000, Thom McDonald <t.a.mcd@ix.netcom.com> wrote: >Newton isn't entirely dead. The name, the OS and the product line are all dead. >It should be fairly obvious that Apple wants to retain Newton technologies for >incorporating into a 'Lite' Mac OS which will be a Newton/PalmPilot/Windows >CE-like platform. The good things about Newton, like touch-screen operation >and hand-writing recognition will make nice additions to Mac OS. It won't be a Newton. >Of course this wouldn't be a totally public stance because if they got a >really big offer, maybe they wouldn't want to turn it down. I don't think they could sell it even if they wanted. It's not like there is a folder called "Newton Technologies" that they can just drag to a floppy and sell. Much of what was "Newton" is now gone. They can't even fix an ongoing OS bug (the -10016 error) -- sal@panix.com Salvatore Denaro "The reality of the software business today is that if you find something that can make you ridiculously rich, then that's something that Microsoft is going to take away from you." -- Max Metral
From: u@canguess.com Newsgroups: comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.oric,comp.sys.palmtops Subject: ASIAN babe wants YOU to take care of here Date: Sunday, 19 Apr 1998 18:22:32 -0600 Organization: <no organization> Distribution: World Message-ID: <19049818.2232@canguess.com> http://xxx-18.com/hotbabe/
From: "cal" <zmendel@home.com> Newsgroups: comp.sys.next.programmer Subject: Where to purchase OpenStep? Date: Sun, 19 Apr 1998 12:57:13 -0400 Organization: @Home Network Message-ID: <6hdaal$quc$1@ha2.rdc1.nj.home.com> I'm just getting back into NeXTStep development since 1990. I tried looking on the Apple Enterprise site for buying OpenStep but couldn't find anything except Apple products. Where does one purchase OpenStep and how much does it cost for a developer version? Thanks. cal [anti-spam: remove z in address to reply]
From: "cal" <zmendel@home.com> Newsgroups: comp.sys.next.marketplace,comp.sys.next.hardware,comp.sys.next.programmer Subject: To ADB or not to ADB? That is the question. Date: Sun, 19 Apr 1998 14:38:15 -0400 Organization: @Home Network Message-ID: <6hdg83$coe$1@ha2.rdc1.nj.home.com> I'm buying a NeXTstation Turbo Color. Some have ADB. Some don't. I know what ADB is (I have a Mac). From a user or developer standpoint, what's the functional difference? Should I consider ADB over non-ADB? Does the ADB just have smarter cable routing to the monitor and allow other ADB devices to plug in? I'm going to be using it as a controller for some home automation projects. Thanks for any help. cal [anti-spam: remove "z" in address to reply]
From: nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) Newsgroups: comp.sys.next.marketplace,comp.sys.next.hardware,comp.sys.next.programmer Subject: Re: To ADB or not to ADB? That is the question. Followup-To: comp.sys.next.hardware Date: 19 Apr 1998 19:59:02 GMT Organization: none Message-ID: <6hdl26$1p1$6@ha2.rdc1.nj.home.com> References: <6hdg83$coe$1@ha2.rdc1.nj.home.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: mendel@home.com NOTE: FOLLOWUPS TO COMP.SYS.NEXT.HARDWARE In <6hdg83$coe$1@ha2.rdc1.nj.home.com> "cal" wrote: > From a user or developer standpoint, what's the functional difference? > Should I consider ADB over non-ADB? Does the ADB just have smarter cable > routing to the monitor and allow other ADB devices to plug in? I think the only real difference is from the user standpoint of "which do you prefer?" ADB obviously makes it a little easier to replace hardware that has failed. ADB's mouse was supposed to be a little more ergonomic, some folks didn't like the way it felt (I did, but never had one to use for very long). The real problem I had with the ADB was the COMMAND 'bar' rather than the two COMMAND keys that non-ADB has. I was always hitting it by accident (again, I never used one for all that long) and was more used to the COMMAND keys (one on each side of the space bar.... the ADB had the command-bar right under the space bar) When all is said and done, I still think it boils down to personal preference with a slight edge on ADB for being more easily replaced. TjL (another @Home user in your area) -- [Yes that is a valid mail address, until it gets spammed] Unofficial @Home FAQ: http://members.home.com/faqs/ Who decided we _wanted_ a Tarzan for the 90s?
From: Christian Jensen <cejensen@winternet.com> Newsgroups: comp.sys.next.programmer Subject: Re: Clocks Date: 19 Apr 1998 21:29:55 GMT Organization: StarNet Communications, Inc Sender: Christian Jensen <cejensen@winternet.com> Message-ID: <6hdqcj$333$1@blackice.winternet.com> New to programming, I am working on a simple clock project and have run into a puzzler: The Workspace Dock clock uses one big tiff (clockbits.tiff) from which it displays its various faces, text, and digits, rather than lots of little tiffs which is what I assumed it used. How is this done? I'm sure that the info for this is buried somewhere in the NeXTDev bookshelf, but searches on relevant keywords return a bewildering number of irrelevant hits. Pointers to meaningful documents, source, etc. would be *much* appreciated. Many thanks, --Chris ************************** Chris Jensen cejensen@winternet.com MIME, Sun, NeXTMail OK "Sacred cows make the best hamburger." --Mark Twain
Newsgroups: comp.sys.next.programmer Subject: -Easy money better than MLM! From: wvhyehflgreenstuff@email.com Organization: Success Pte Ltd Message-ID: <353a6ab6.0@news.cyberway.com.sg> Date: 19 Apr 98 21:20:54 GMT **************************************************************** * This Article was Posted By an unregistered version of: * * Newsgroup AutoPoster 95 * * Send email address for info! Fax: +46-31-470588 * **************************************************************** HONEST "PEOPLE HELPING PEOPLE" OPPORTUNITY 100 TIMES MORE EFFECTIVE THAN MLM! Could YOU use an extra $3,000, $5,000 or MAYBE $10,000 in the next 2 weeks? The Total Investment is only Five Dollars and less than One Hour of Work! THAT'S IT! Unless you want to do it again. COMPARED TO OTHER MLM'S: 1. It is so much easier to start. 2. You don't need to continually monitor or work it. 3. One time, very small initial investment. 4. Very high response rate. 5. Fantastic return on investment. 6. Perfectly legal. PLEASE TAKE THREE MINUTES TO FIND OUT HOW! THIS IS THE FASTEST, EASIEST PROGRAM you will ever do. Complete it in less than ONE HOUR and you will never forget the day you first received it. If you are doing other programs, by all means stay with them. The more the merrier! But, PLEASE READ ON! There are only THREE LEVELS. This three-level program is more realistic and much, much faster. Because it is so easy, the response rate for this program is VERY HIGH and VERY FAST. You will receive your rewards in about FOURTEEN DAYS. That's only TWO WEEKS - not three months. Then, buy those extra things you've been dreaming of. Cindy Allen tells how she ran this gift summation four times last year. The first time, she received $3,000 in cash in two weeks and then $7,000 in cash the next three times. When this letter is continued as it should be, EVERYONE PROFITS! Don't be afraid to make a gift to a stranger, it will come back to you TEN FOLD. Many of us have pet programs that we want to support, food, medicine or medical care for poor children is another. Maybe, you just need a new car, want to pay off some bills or take a much needed vacation. DO IT. IT'S YOUR TURN! HERE ARE THE SIMPLE DETAILS You e-mail just 20 copies (the more copies you send the more cash you make) of this program to people you personally know, to people like you who are interested in earning extra cash, and to people who send you e-mail about their programs. WHY? Because they are already believers and your program is BETTER AND FASTER. Even if you are already in a program, continue to stay with it. But, do yourself a favor and DO THIS ONE for the fast cash. RIGHT NOW! It is so simple and the cost is so little. Like going out for fast food or buying a couple of beers. JUST GIFT ONE PERSON A 5 DOLLAR BILL. THAT'S IT! THAT'S ALL! Follow these simple instructions and in as little as TWO WEEKS you could have at least $3,000, $5,000, or up to $10,000 in five dollar gifts. Why? Because many people WILL respond due to the "LOW" cost to get started and a VERY HIGH REWARDING POTENTIAL! So, get going and help each other! The government certainly isn't going to. 1. On a sheet of paper, clearly write down YOUR name and address, along with the statement "I GIVE THIS FIVE DOLLAR GIFT TO YOU OF MY OWN FREE WILL AND EXPECT NOTHING IN RETURN." Fold it around a FIVE DOLLAR BILL. Send it to the FIRST name on the list. ONLY THE FIRST PERSON ON THE LIST GETS YOUR NAME AND A FIVE DOLLAR GIFT. 2. Now, remove the first name and address from the list and move the other two names up. Then, add YOUR name and address to the third (#3) position. 3. Save your changes and then e-mail 20 copies or more of this letter. Remember, an excellent source of names are the people who send you other programs and the names listed on the letters that they send you. Do it right away. It's so easy! Don't mull it over. ONE HOUR! THAT'S IT! There are no mailing lists to buy or wait for. You can do it again and again with your regular group of Gifters. Why not? It pays to help others! Each time you receive a MLM offer, respond with THIS letter! Your name will climb to the number ONE position in a DIZZYING geometric rate. So, come on! The prospect of an easy $3,000, $5,000 to $10,000 in TWO WEEKS is worth a little experimentation, isn't it? It only takes less than One Hour of your time and a five dollar bill (cash). ACT FAST AND GET MONEY FAST! HONESTY AND INTEGRITY MAKE THIS PLAN WORK. COPY THE NAMES CAREFULLY AND SEND YOUR FIVE DOLLAR CASH GIFT TO THE FIRST NAME. SPECIAL NOTE: Please don't try to cheat the system. You will only be cheating yourself. PEOPLE HELPING OTHER PEOPLE! 1. MRA Consultants 761 Amsterdam Road Dept. #1 Mt. Laurel, NJ 08054-3201 2. Picasso Chen Blk 611 Ang Mo Kio Aveune 5 #11-2805 Singapore 560611 3. Simon Tan Blk 4 Hill View Ave #03-1090 Singapore 661004
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.next.programmer Subject: Re: Finding gamma of the display Date: 19 Apr 1998 22:24:58 GMT Organization: P & L Systems Message-ID: <6hdtjq$57p$5@ironhorse.plsys.co.uk> References: <3537f2b7.0@news.depaul.edu> <892918938.222793@globe.uwaterloo.ca> <892928475.749663@globe.uwaterloo.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: dfevans@bbcr.uwaterloo.ca In <892928475.749663@globe.uwaterloo.ca> David Evans wrote: > In article <892918938.222793@globe.uwaterloo.ca>, > David Evans <dfevans@bbcr.uwaterloo.ca> wrote: > > I know that there's an operator called something like "settransferfunction" > >that sets the gamma; there's likely an analogue that retrieves it. > [...] > so this basically tells us bugger-all. > Guess I wasn't of much help! > Oh, so close... :-) Date: Sun, 19 Apr 1998 05:54:19 -0700 (PDT) Sender: rhapsody-dev@omnigroup.com From: mmalcolm crawford <Malcolm_Crawford@plsys.co.uk> To: Multiple recipients of list <rhapsody-dev@omnigroup.com> Subject: Re: Finding gamma of the display device? Use all code herein described at your own risk! Jon wrote: > Anyone know if there's a way in Rhapsody to determine the > gamma value for a display? > What an interesting question; I hadn't really considered this before, so thanks to a bit of help from a friend here in getting the right values back (I can never remember how to get things off the stack <sigh>)... You can sort of, if you make a couple of assumptions. The framebuffer transfer is a function, apparently usually defined: { 1 gammaValue div exp } for each of R G and B (cf documentation below) If you accept this (and a quick experiment suggests this is reasonable), then you can get the gamma value for each of R G and B (they're usually the same) using the following pswrap: defineps cfbt( | float *r; float *g; float *b) 0 currentframebuffertransfer pop 0.5 exch exec log 0.5 log exch div 3 -1 roll 0.5 exch exec log 0.5 log exch div 3 -1 roll 0.5 exch exec log 0.5 log exch div 3 -1 roll b g r endps Note that you have to get the values from the globaldict, so your code would probably look something like this: [gammaView lockFocus]; PSsetshared(1); cfbt(&r, &g, &b); PSsetshared(0); [gammaView unlockFocus]; Armed with this knowledge it's trivially simple to do the "reverse" -- so I've now written a small OPENSTEP app which allows you to set the current gamma value for your machine, manipulating R G and B independently if you want. Seeing your workspace only in shades of green provides minutes of entertainment. Let me know if you're interested... Best wishes, mmalc. P & L Systems -- developers of Mesa http://www.plsys.co.uk/plsys/ Tel: +44 1494 432422 Fax: +44 1494 432478 Here's the relevent documentation (thanks to Mike Paquette): setframebuffertransfer redproc greenproc blueproc grayproc fbnum setframebuffertransfer - Warning: This operator should only be used for the development of screen-calibration products. Sets the framebuffer transfer functions in effect for the framebuffer indexed by fbnum. fbnum ranges from 0 to countframebuffers-1. The framebuffer transfer describes the relationship between the framebuffer values of the display, and the voltage produced to drive the monitor. The initial four operands define the transfer procedures: Monochrome devices use grayproc (but see the Note below), color devices use the others. The procedures must be allocated in shared virtual memory. In addition, the Window Server assumes that the framebuffer values are directly proportional to screen brightness. This is important for the accuracy of dithering, compositing, and similar calculations. The default transfer for NeXT Color Displays is { 1 2.2 div exp } bind dup dup {}
From: Wesley Horner <wesman@azrael.uoregon.edu> Newsgroups: comp.sys.next.programmer Subject: Re: Where to purchase OpenStep? Date: 19 Apr 1998 22:42:02 GMT Organization: University of Oregon, Eugene Message-ID: <6hdujq$dm2$1@pith.uoregon.edu> References: <6hdaal$quc$1@ha2.rdc1.nj.home.com> cal <zmendel@home.com> wrote: > I'm just getting back into NeXTStep development since 1990. I tried looking > on the Apple Enterprise site for buying OpenStep but couldn't find anything > except Apple products. Where does one purchase OpenStep and how much does > it cost for a developer version? I believe I saw it for sale at www.blackhole.com. I hope your a student because the developer version is pricy. You might be better off becoming an apple developer and programming for rhapsody. -- ~~~~wesman@gladstone.uoregon.edu~~~~~~~~~~NeXTMail OK!~~~~~~~~~~~~~~~~ Vax a vicious creature known to eat 110AC and quotes through its *DCL*. Vax are usually found in groups of Vaxen called clusters where they lay in wait to ravage thier prey known as users.
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.next.programmer Subject: Re: This doesn't bode well for Rhapsody Date: 19 Apr 1998 22:47:31 GMT Organization: P & L Systems Message-ID: <6hduu3$57p$7@ironhorse.plsys.co.uk> References: <3514423f.2211203@news.gatech.edu> <6f42kn$evk$1@news.idiom.com> <351629AB.45B94D20@null.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: null@null.com In <351629AB.45B94D20@null.com> null wrote: > > We had some biz discussions with apple about rhapsody, but they were > so incredibly cautious about targeting rhapsody for consumer apps at > the expense of macos that I doubted their resolve. > <yawn> FUD The question is now not whether or not to target Rhapsody, it is whether you want to move forward with Apple and target YellowBox, or stick in the past with an outdated and significantly less powerful API. > Apple seems to be "Newtonizing" Rhapsody. Otherwise known as "Lets not make > anything that might take away from MacOS sales" (formally sung to the tune > of "Lets not allow the Newton to compete for PowerBook sales") > In some respects, yes, and this is absolutely the right thing to do. What they're not doing, though is trashing Rhapsody like they did Newton. They're also doing what I hope they will also do with the Newton, namely bringing as many of the good technologies (esp YellowBox) across to MacOS as fast as possible. > The below little rumor is also troubling.... > Only if you don't have a clue. > the Knife is reminded of the technology housecleaning reportedly underway at > Adobe. Specifically, some clever souls suggest that Adobe may ditch further > development of Display PostScript. This bit of downsizing could complicate > matters for Rhapsody, since Apple's new wave OS makes heavy use of the > space-age display techology > No it won't -- DPS was co-developed by NeXT; Apple doesn't need Adobe to continue development. Best wishes, mmalc.
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.next.programmer Subject: Re: Finding gamma of the display Date: 19 Apr 1998 22:38:32 GMT Organization: P & L Systems Message-ID: <6hdud8$57p$6@ironhorse.plsys.co.uk> References: <3537f2b7.0@news.depaul.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: jhendry@shrike.depaul.edu In <3537f2b7.0@news.depaul.edu> Jonathan W Hendry wrote: > I believe it's typically a fixed value on OpenStep, but adjustable > on Rhapsody/PPC. (Anyone know what the value is on OpenStep?) > OpenStep is usually 2.2 > I'm wondering because of discrepancies between the RGB values for > 'Netscape colors' across platforms. I took a Fractal Design Painter > color set of Netscape colors, and converted it to an OpenStep > NSColorList. When compared to the NeXT-provided set of Netscape > colors, they didn't match up (in color or RGB values). > > I'm assuming that the NeXT Netscape color set was created under > OpenStep's gamma settings, while Painter's version was created > under Mac or Windows gamma settings. Similar-looking colors > would then have different RGB values. > Hmm, the RGB colours should be constant whatever, for the Netscape 216 color palette under 8 bit mode all colors must have an R, G, or B value of 0, 51, 102, 153, 204, or 255. It's then up to the OS to determine how this will be displayed... Which "NeXT-provided set of Netscape colors", by the way?! You mean the ones from http://www.plsys.co.uk/~malcolm/NEXTSTEP/WWW/ ? The list there was kindly provided by someone else (Don?) -- I wish I could remember who as I ought to credit them appropriately. Best wishes, mmalc.
From: nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) Newsgroups: comp.sys.next.programmer Subject: Re: Where to purchase OpenStep? Date: 19 Apr 1998 23:37:24 GMT Organization: none Message-ID: <6he1rk$1k3$4@ha2.rdc1.nj.home.com> References: <6hdaal$quc$1@ha2.rdc1.nj.home.com> <6hdujq$dm2$1@pith.uoregon.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit In <6hdujq$dm2$1@pith.uoregon.edu> Wesley Horner wrote: > cal <zmendel@home.com> wrote: > > I'm just getting back into NeXTStep development since 1990. I tried looking > > on the Apple Enterprise site for buying OpenStep but couldn't find anything > > except Apple products. Where does one purchase OpenStep and how much does > > it cost for a developer version? > > I believe I saw it for sale at www.blackhole.com. I think he meant http://www.blackholeinc.com/ TjL -- [Yes that is a valid mail address, until it gets spammed] Unofficial @Home FAQ: http://members.home.com/faqs/ Who decided we _wanted_ a Tarzan for the 90s?
From: tjw@penguin.omnigroup.walrus.com Newsgroups: comp.sys.next.programmer Subject: Re: NSData and VM-poking on mach Date: 19 Apr 1998 18:08:10 -0700 Organization: Omni Development, Inc. Message-ID: <6he75q$8h8$1@gaea.omnigroup.com> References: <6hcqjk$t6h$1@sun579.rz.ruhr-uni-bochum.de> I've whipped up an example of part of this problem (keeping track of the modified pages -- not writing them back out) and placed it at: ftp://ftp.omnigroup.com/pub/software/ExampleCode/OFPageMonitor.tar.gz This will hopefully get you started down the path you need to take. -tim stefan.boehringer_REMOVE_ME_@ruhr-uni-bochum.de (Stefan Boehringer) writes: [...] >Since I can easily find out the vm-memory region involved (by calling the >bytes method) and afterwards scrutinizing it for dirty pages. However I could >not find an obvious way to do just that. Does anybody know how one could tell >a dirty page from a non-dirty one? No mach-function seems to exist. [...] -- Remove the animals from my email address to respond via email. Hey spam kings ... parse these email addresses! postmaster@localhost postmaster@127.0.0.1 root@localhost root@127.0.0.1
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <35380bf0.1@news.codenet.net> Control: cancel <35380bf0.1@news.codenet.net> Date: 19 Apr 1998 03:13:07 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.35380bf0.1@news.codenet.net> Sender: HOT XXXXXX Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6hc21j$ing$54@camel21.mindspring.com> Control: cancel <6hc21j$ing$54@camel21.mindspring.com> Date: 19 Apr 1998 05:28:24 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6hc21j$ing$54@camel21.mindspring.com> Sender: 1079@1079.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6hc2dt$7i0$54@camel25.mindspring.com> Control: cancel <6hc2dt$7i0$54@camel25.mindspring.com> Date: 19 Apr 1998 05:37:04 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6hc2dt$7i0$54@camel25.mindspring.com> Sender: 564@564.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: klui@cup.hp.com (Ken Lui) Newsgroups: comp.sys.next.marketplace,comp.sys.next.hardware,comp.sys.next.programmer Subject: Re: To ADB or not to ADB? That is the question. Date: 20 Apr 1998 08:30:25 GMT Organization: Hewlett-Packard Company Message-ID: <6hf131$2hp$2@ocean.cup.hp.com> References: <6hdg83$coe$1@ha2.rdc1.nj.home.com> In article <6hdg83$coe$1@ha2.rdc1.nj.home.com>, cal <zmendel@home.com> wrote: > From a user or developer standpoint, what's the >functional difference? Should I consider ADB over non-ADB? The ADB keyboard, while it looks cooler than than the non-ADB NeXT keyboard, lacks the nice feel of the non-ADB keyboard, and doesn't have the multi- colored NeXT logo. Anybody ever tried transplanting parts of the non-ADB keyboard to an ADB version? I know the row with the space bar won't work, and the caps lock functionality would be problematic. Ken -- Ken Lui 19111 Pruneridge Avenue M/S 44UR klui@cup.hp.com Cupertino, CA 95014-0795 USA Information Solutions & Services 1.408.447.3230 FAX 1.408.447.1053 Views within this message may not be those of the Hewlett-Packard Company
From: pemmerik@solair1.inter.NL.net (P.J.L.van Emmerik) Newsgroups: comp.sys.next.programmer Subject: Lookong for InterfaceViewer or something like it. Date: Mon, 20 Apr 1998 05:59:44 GMT Organization: NLnet Message-ID: <6heo81$6k5$1@news.NL.net> Becouse of the overwhelming silence on my previous request on the whereabouts of developer of InterfaceViewer here a retry: I want to document the interface of a new application we made. A tool like InterfaceViewer would be a tremendous help. For thos who do not know: InterfaceViewer reads a .nib file and creates a schematic of all the links between objects in the .nib, both the graphic (buttons views etc.) and the nongraphic objects (delegeates etc). We are still using NEXTSTEP 3.3 Who can help me out? Please Email to: emmerik@hpb.holec-projects.nl P.J.L. van Emmerik Holec Projects B.V. Email: emmerik@hpb.holec-projects.nl PO.BOX 565, 7550 AN Hengelo pemmerik@solair1.inter.NL.net The Netherlands Phone: +31 74 2558 688 --
From: rdelucca@yahoo.com (Robert) Newsgroups: comp.sys.next.marketplace,comp.sys.next.hardware,comp.sys.next.programmer Subject: Re: To ADB or not to ADB? That is the question. Date: Mon, 20 Apr 1998 07:40:29 -0500 Organization: Slim Message-ID: <rdelucca-2004980740290001@async251-29.async.duke.edu> References: <6hdg83$coe$1@ha2.rdc1.nj.home.com> In article <6hdg83$coe$1@ha2.rdc1.nj.home.com>, "cal" <zmendel@home.com> wrote: > I'm buying a NeXTstation Turbo Color. Some have ADB. Some don't. I know > what ADB is (I have a Mac). From a user or developer standpoint, what's the > functional difference? Should I consider ADB over non-ADB? Does the ADB > just have smarter cable routing to the monitor and allow other ADB devices > to plug in? > > I'm going to be using it as a controller for some home automation projects. > > Thanks for any help. > cal > [anti-spam: remove "z" in address to reply] I had both types of slab. Far as I know, the ADB allows you to use other manufacturers ADB devices. The ADB keyboard and mouse (Next's) were superior, in my opinion, the mouse especially. As for cabling, I frankly don't recall (use only powerbooks now, alas). Robert
From: Andre-John Mas <ama@fabre.act.qc.ca> Newsgroups: comp.sys.next.programmer Subject: OpenStep on top of Linux? Date: Mon, 20 Apr 1998 09:21:23 -0500 Organization: Communications Accessibles Montreal, Quebec Canada Message-ID: <Pine.LNX.3.95.980420091603.21479B-100000@fabre.act.qc.ca> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII If I understand the OpenStep OS, then it is actualy an OS ontop of an OS, the current bottom layer OS being Mach. Everyone would love an OS that is controlled by everyone, the only OS that currentlt fits this description in Linux. So with the GNUstep project do you think that it would be possible to create GNUstep to run on top of Linux and still be compatible with OpenStep. I wonder if a version of Rhapsody running on top of MkLinux or any other version of Linux would be a workable technology. AJ
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.oric,comp.sys.palmtops Subject: cmsg cancel <19049818.2232@canguess.com> Control: cancel <19049818.2232@canguess.com> Date: 19 Apr 1998 16:43:40 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.19049818.2232@canguess.com> Sender: u@canguess.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: "J Barbosa" <jbarbosa@ix.netcom.com> Newsgroups: comp.sys.next.programmer Subject: OPENSTEP Developers in the Chicago Area Date: Mon, 20 Apr 1998 14:11:26 -0500 Organization: Netcom Message-ID: <6hg6l8$2hs@sjx-ixn6.ix.netcom.com> We're looking for a couple of OPENSTEP developers in Chicago for a mission-critical EOF based application. Does anyone have any suggestions? Consulting/contract programmers are fine. Jorge
From: tom@basil.icce.dev.rug.null.nl (Tom Hageman) Newsgroups: comp.sys.next.programmer Subject: Re: Clocks Date: 20 Apr 1998 22:23:00 GMT Organization: Warty Wolfs Sender: news@basil.icce.dev.rug.null.nl (NEWS pusher) Message-ID: <ErqC1q.8uG@basil.icce.dev.rug.null.nl> References: <6hdqcj$333$1@blackice.winternet.com> Christian Jensen <cejensen@winternet.com> wrote: > > New to programming, I am working on a simple clock project and have run into > a puzzler: The Workspace Dock clock uses one big tiff (clockbits.tiff) from > which it displays its various faces, text, and digits, rather than lots of > little tiffs which is what I assumed it used. How is this done? > > I'm sure that the info for this is buried somewhere in the NeXTDev bookshelf, > but searches on relevant keywords return a bewildering number of irrelevant > hits. Pointers to meaningful documents, source, etc. would be *much* > appreciated. Check out MiscClockView in the MiscKit (in Palettes/MiscClockPalette/MiscClockView/ for NEXTSTEP-based MiscKit-1.10.0, and Temp/MiscClockView/ for OPENSTEP-based MiscKit-2.0.5.) [Hmmm. Seems that NXImage's "sub-image" concept in NEXTSTEP didn't make it into OPENSTEP's NSImage...] For more info about the MiscKit, see http://www.misckit.com/. -- __/__/__/__/ Tom Hageman <tom@basil.icce.dev.rug.null.nl> [NeXTmail/Mime OK] __/ __/_/ IC Group <tom@dev.icgroup.rug.nl> (work) __/__/__/ __/ _/_/ Confused? You won't be after the NeXT episode.
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: OpenStep on top of Linux? Date: 20 Apr 1998 14:49:12 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6hfn98$qhk$1@ns3.vrx.net> References: <Pine.LNX.3.95.980420091603.21479B-100000@fabre.act.qc.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: ama@fabre.act.qc.ca In <Pine.LNX.3.95.980420091603.21479B-100000@fabre.act.qc.ca> Andre-John Mas claimed: > If I understand the OpenStep OS, then it is actualy an OS ontop > of an OS, the current bottom layer OS being Mach. Mach in of itself is not an OS, it's an OS controller. > Everyone would > love an OS that is controlled by everyone, the only OS that > currentlt fits this description in Linux. There are any number of OS's that fit this description, FreeBSD, MK-Lites, OSF/1, GNU Hurd etc. > project do you think that it would be possible to create GNUstep > to run on top of Linux and still be compatible with OpenStep. GNUStep development seems to be mostly dead. As does (sadly) the Hurd. > I wonder if a version of Rhapsody running on top of MkLinux or > any other version of Linux would be a workable technology. Certainly, but why? Although it would certainly have some "see if we can do it" appeal, I don't see any specific commercial reason for doing so. A more important and practical solution would be to make sure there is a very high degree of compatibility between Linux and Rhapsody. Maury
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <353a6ab6.0@news.cyberway.com.sg> Control: cancel <353a6ab6.0@news.cyberway.com.sg> Date: 19 Apr 1998 21:47:11 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.353a6ab6.0@news.cyberway.com.sg> Sender: wvhyehflgreenstuff@email.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: sub images (was) Clocks Date: 21 Apr 1998 00:31:39 GMT Organization: Idiom Communications Message-ID: <6hgpdb$d8e$1@news.idiom.com> References: <6hdqcj$333$1@blackice.winternet.com> <ErqC1q.8uG@basil.icce.dev.rug.null.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: tom@basil.icce.dev.rug.null.nl Tom Hageman may or may not have said: -> Check out MiscClockView in the MiscKit (in -> Palettes/MiscClockPalette/MiscClockView/ for NEXTSTEP-based MiscKit-1.10.0, and -> Temp/MiscClockView/ for OPENSTEP-based MiscKit-2.0.5.) -> -> [Hmmm. Seems that NXImage's "sub-image" concept in NEXTSTEP didn't make it -> into OPENSTEP's NSImage...] No, it sure didn't and that's quite unfortunate. I've got some code that I need to hand over to Don that implements a sub-image class. You create sub images by sending an NS image a message like [myImage subImage: subImageRect] where subImageRect has to fit within the bounds of the receiver. -jcr
From: fedor@vnet.net (Adam Fedor) Newsgroups: comp.sys.next.programmer Subject: Re: OpenStep on top of Linux? Date: 21 Apr 1998 01:05:38 GMT Organization: Vnet Internet Access, Inc. Message-ID: <6hgrd2$jpe$1@ralph.vnet.net> References: <Pine.LNX.3.95.980420091603.21479B-100000@fabre.act.qc.ca> <6hfn98$qhk$1@ns3.vrx.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit In <6hfn98$qhk$1@ns3.vrx.net> Maury Markowitz wrote: > > GNUStep development seems to be mostly dead. As does (sadly) the Hurd. > Well, as in real life, very slow movement can often be mistaken for death. Actually, GNUstep is still moving forward (I'm actually starting to have a hard time keeping up with submissions). I doubt Hurd is dead either. It's just more complex than anyone first realized. GNustep has the same interface as Rhapsody, but it is not compatible in the sense that you could compile an application on Rhapsody and run it on GNUstep. -- Adam Fedor. Digital Optics Co. | Those who can't do, simulate. fedor@doc.com (MIME) | fedor@vnet.net (MIME, NeXT) |
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Followup-To: comp.sys.mac.advocacy,comp.sys.next.advocacy Date: 21 Apr 1998 07:29:45 GMT Organization: P & L Systems Message-ID: <6hhht9$57p$15@ironhorse.plsys.co.uk> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6gecns$f9p$1@news01.deltanet.com> <1d767ho.1tixfqy1dqgm03N@hobbit1.injep.fr> <6gg6af$fit$1@anvil.BLaCKSMITH.com> <6ggabn$rts$2@news01.deltanet.com> <1d76snt.17ecjre1ykt2xdN@ppgutkneco.lirmm.fr> <1998040920073876968@dialup-202.def.oleane.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: Le_Jax@iName.com In <1998040920073876968@dialup-202.def.oleane.com> Jacques Foucry wrote: > Olivier Gutknecht <gutkneco+news@lirmm.fr> wrote (Ýcrivait)€: > > > Agreed. The only explanation is that Apple is not interested anymore in > > small/independant development groups. > > Perhaps, but without small/independant developer, no "Super Clock", no > "Stickies", no new Control Strip... > And the downside...? FWIW, Stickies for Rhapsody was pretty much our Friday afternoon project for the folks at Apple Cork... Best wishes, mmalc.
From: pemmerik@solair1.inter.NL.net (P.J.L.van Emmerik) Newsgroups: comp.sys.next.programmer Subject: Looking for reverse engineering tools for OBJ-C and NEXT3.3 Date: Tue, 21 Apr 1998 08:06:00 GMT Organization: NLnet Message-ID: <6hhk1f$s1g$1@news.NL.net> For documenting puposes i am lookong for reversed engineering tools. We want to be able to create object diagrams etc. from the code. For C++ there are tools that can do the trick, but i do not know any tool that can do it for Objective-C. Any helpfull pointers? Pleace E-mail to: emmerik@hpb.holec-projects.nl P.J.L. van Emmerik Holec Projects B.V. Email: emmerik@hpb.holec-projects.nl PO.BOX 565, 7550 AN Hengelo pemmerik@solair1.inter.NL.net The Netherlands Phone: +31 74 2558 688 --
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: NSImage oddness Date: 21 Apr 1998 09:15:59 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6hho4f$e3g$4@ns3.vrx.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I'm placing pictures inside a view that's flipped. This is easy enough normally, you simply do a PSTranslate(1,-1) first and then do your draw (sorry, wrapped in a Gsave/Grestore). Well it doesn't seem to be that easy actually, because when I do this the rect it draws into seems to be incorrect. More specifically I send out a stream of draws as the "container" for the image is resized, any fast moves on the resize results in the image being drawn into what appears to be an old rect. I know what you're thinking, but no, the rect is correct. If I remove the translate it works fine. In fact it's odder than that, if I do ANY ps code before (say a rotate) this problem occurs. If I remove it, it works fine. I should point out I'm using the imageRep class, not image. There's one other odd symptom that might have some bearing. If I _do_ remove any PS code from the code prior to drawing it works perfectly - however the image is upside down (obviously) but more interestingly the image draw happens really slowly - ie it's following the rect correctly but draw times are maybe twice as slow. Has anyone seen effects like this? I _think_ that is has something to do with the rect being cached somewhere (that I don't know) in the NS code with the image being drawn at some later time and getting the old value. However I can't for the life of me figure out why removing the "flipping" translate call would have such an effect. Maury
From: mark.s.frank.remove_this_part@tek.com Newsgroups: comp.sys.next.programmer Subject: Help with iostream.h on Enterprise Date: 21 Apr 1998 15:48:37 GMT Organization: Tektronix, Inc. Distribution: World Message-ID: <6hif4l$opr$2@bvadm.bv.tek.com> Has anyone been able to include iostream.h in C++ files on Openstep Enterprise for NT? When I try to do this, the compiler complains that it can't find useoldio.h. Any help would be appreciated. - Thanks, Mark
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: OpenStep on top of Linux? Date: 21 Apr 1998 09:01:05 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6hhn8h$e3g$2@ns3.vrx.net> References: <Pine.LNX.3.95.980420091603.21479B-100000@fabre.act.qc.ca> <6hfn98$qhk$1@ns3.vrx.net> <6hgrd2$jpe$1@ralph.vnet.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: fedor@vnet.net In <6hgrd2$jpe$1@ralph.vnet.net> Adam Fedor claimed: > Well, as in real life, very slow movement can often be mistaken for death. > Actually, GNUstep is still moving forward (I'm actually starting to have a > hard time keeping up with submissions). Yes, but the level of development seems to be isolated entirely to the FoundationKit level, the PB, IB and AK layers have not been majority updated in a long time (two years for IB and counting) > I doubt Hurd is dead either. It's > just more complex than anyone first realized. For sure, but that's the problem as I see it. I think the Hurt is a VERY important step forward in the Unix world, perhaps one of the single most important general "research" projects underway. Yet I see little to suggest that the current team has any real hope of releasing product at any reasonable future date. That's not a shame, I think it's potentially damaging. Say what you will about OSF, they do get product out the door, and I think this is one effort that should have at least some level of OSF "real" backing. > GNustep has the same interface as Rhapsody, but it is not compatible in the > sense that you could compile an application on Rhapsody and run it on > GNUstep. GNUStep actually has the API of the _public_ side of OpenStep - whereas most OpenStep apps use a number of additional "goodies" that are not strictly part of the standard. I have fears that porting efforts (if it ever "ships") will be entirely non-trivial, notably with the updates OS is getting in Rhapsody. Some will take this as "downer" on GNU or GNUStep. Far from it, I think nothing would help Rhapsody more than a good GNUStep. However it seems to me that GNU just isn't set up for really large projects like Hurd or GNUStep, and that's making release times far too long for my liking. Maury
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: sub images (was) Clocks Date: 21 Apr 1998 09:03:20 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6hhnco$e3g$3@ns3.vrx.net> References: <6hdqcj$333$1@blackice.winternet.com> <ErqC1q.8uG@basil.icce.dev.rug.null.nl> <6hgpdb$d8e$1@news.idiom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: jcr.remove@this.phrase.idiom.com In <6hgpdb$d8e$1@news.idiom.com> John C. Randolph claimed: > No, it sure didn't and that's quite unfortunate. I've got some code that I > need to hand over to Don that implements a sub-image class. You create sub > images by sending an NS image a message > like [myImage subImage: subImageRect] where subImageRect has to fit within > the bounds of the receiver. It's hard to be sure, but are you referring to placing an image inside the rect of another view? Why not do a... [[[myImage imageReps] objectAtIndex:0] drawRect:subImageRect]; Is that the same thing? Maury
Date: 21 Apr 1998 15:04:22 GMT From: clewis@ferret.ocunix.on.ca (Chris Lewis) Sender: rbriston@skylinc.net Message-ID: <cancel.18049816.1827@skylinc.net> Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k Subject: cmsg cancel <18049816.1827@skylinc.net> Control: cancel <18049816.1827@skylinc.net> XXX18 spam cancelled by clewis@ferret.ocunix.on.ca Total spams this type to date: Total this spam type for this user: Total this spam type for this user today: Originating site: www.axon.se Complaint addresses:
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: sub images (was) Clocks Date: 21 Apr 1998 17:46:27 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6him1j$5gq$6@ns3.vrx.net> References: <6hdqcj$333$1@blackice.winternet.com> <ErqC1q.8uG@basil.icce.dev.rug.null.nl> <6hgpdb$d8e$1@news.idiom.com> <6hhnco$e3g$3@ns3.vrx.net> <6hivc5$3i8$1@news.digifix.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: sanguish@digifix.com In <6hivc5$3i8$1@news.digifix.com> Scott Anguish claimed: > You used to be able to specify that the source of an image was > actually part of a larger image. Ahhh, I get it, you have this big image which a whole bunch of little images in it. Maury
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: sub images (was) Clocks Date: 21 Apr 1998 20:25:41 GMT Organization: Digital Fix Development Message-ID: <6hivc5$3i8$1@news.digifix.com> References: <6hdqcj$333$1@blackice.winternet.com> <ErqC1q.8uG@basil.icce.dev.rug.null.nl> <6hgpdb$d8e$1@news.idiom.com> <6hhnco$e3g$3@ns3.vrx.net> In-Reply-To: <6hhnco$e3g$3@ns3.vrx.net> On 04/21/98, Maury Markowitz wrote: >In <6hgpdb$d8e$1@news.idiom.com> John C. Randolph claimed: >> No, it sure didn't and that's quite unfortunate. I've got some code that I >> need to hand over to Don that implements a sub-image class. You create sub >> images by sending an NS image a message >> like [myImage subImage: subImageRect] where subImageRect has to fit within >> the bounds of the receiver. > > It's hard to be sure, but are you referring to placing an image inside the >rect of another view? Why not do a... > >[[[myImage imageReps] objectAtIndex:0] drawRect:subImageRect]; > > Is that the same thing? > Nope.. You used to be able to specify that the source of an image was actually part of a larger image. Have a look at the MiscClockPalette in the MiscClock.subproj to see an example of this. In the old days this was a nice way to reduce the number of image files that you had to deal with.. Especially important when you were providing them in a palette and the developer had to remember to put them into the app itself. Mind you, with Bundle support, this would be much easier now, and the MiscClock should be re-written to support separate image files.. The original author is just too damn lazy.. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
Date: 21 Apr 1998 19:12:52 GMT From: clewis@ferret.ocunix.on.ca (Chris Lewis) Sender: u1000233@email.sjsu.edu Message-ID: <cancel.18049816.3058@email.sjsu.edu> Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k Subject: cmsg cancel <18049816.3058@email.sjsu.edu> Control: cancel <18049816.3058@email.sjsu.edu> XXX18 spam cancelled by clewis@ferret.ocunix.on.ca Total spams this type to date: Total this spam type for this user: Total this spam type for this user today: Originating site: www.axon.se Complaint addresses:
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.next.programmer Subject: Re: sub images (was) Clocks Date: 22 Apr 1998 00:14:47 GMT Organization: MiscKit Development Message-ID: <6hjcpn$ru7$3@news.xmission.com> References: <6hdqcj$333$1@blackice.winternet.com> <ErqC1q.8uG@basil.icce.dev.rug.null.nl> <6hgpdb$d8e$1@news.idiom.com> <6hhnco$e3g$3@ns3.vrx.net> <6hivc5$3i8$1@news.digifix.com> sanguish@digifix.com (Scott Anguish) wrote: > Have a look at the MiscClockPalette in the MiscClock.subproj > [...]. > Mind you, with Bundle support, this would be much easier now, > and the MiscClock should be re-written to support separate image > files.. > > The original author is just too damn lazy.. But at least three other authors aren't. There will be a new OPENSTEP version in the next release of the OPENSTEP MiscKit. The hardest part will be folding it all together to make sure that none of the functionality from any of the submissions is lost... -- Later, -Don Yacktman don@misckit.com <a href="http://www.misckit.com/don.html">My home page</a>
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: sub images (was) Clocks Date: 22 Apr 1998 03:36:49 GMT Organization: Digital Fix Development Message-ID: <6hjokh$btl$1@news.digifix.com> References: <6hdqcj$333$1@blackice.winternet.com> <ErqC1q.8uG@basil.icce.dev.rug.null.nl> <6hgpdb$d8e$1@news.idiom.com> <6hhnco$e3g$3@ns3.vrx.net> <6hivc5$3i8$1@news.digifix.com> <6hjcpn$ru7$3@news.xmission.com> In-Reply-To: <6hjcpn$ru7$3@news.xmission.com> On 04/21/98, Don Yacktman wrote: >sanguish@digifix.com (Scott Anguish) wrote: >> Have a look at the MiscClockPalette in the MiscClock.subproj >> [...]. >> Mind you, with Bundle support, this would be much easier now, >> and the MiscClock should be re-written to support separate image >> files.. >> >> The original author is just too damn lazy.. > >But at least three other authors aren't. Well thats good... :-) (BTW, for those who didn't realize, I'm the original perp on that palette.. so I was calling myself lazy..) >There will be a new OPENSTEP >version in the next release of the OPENSTEP MiscKit. The hardest part will >be folding it all together to make sure that none of the functionality from >any of the submissions is lost... > A big hand for Don who clearly spends huge amounts of time on the MiscKit.... -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
Newsgroups: comp.sys.next.programmer Subject: Re: OpenStep on top of Linux? References: <Pine.LNX.3.95.980420091603.21479B-100000@fabre.act.qc.ca> <6hfn98$qhk$1@ns3.vrx.net> <6hgrd2$jpe$1@ralph.vnet.net> <6hhn8h$e3g$2@ns3.vrx.net> From: richard@brainstorm.co.uk. (Richard Frith-Macdonald) Message-ID: <353cd195.0@nnrp1.news.uk.psi.net> Date: 21 Apr 98 17:04:21 GMT maury@remove_this.istar.ca (Maury Markowitz) wrote: >In <6hgrd2$jpe$1@ralph.vnet.net> Adam Fedor claimed: >> Well, as in real life, very slow movement can often be mistaken for death. >> Actually, GNUstep is still moving forward (I'm actually starting to have a >> hard time keeping up with submissions). > > Yes, but the level of development seems to be isolated entirely to the >FoundationKit level, the PB, IB and AK layers have not been majority updated >in a long time (two years for IB and counting) I think that things are likely to start moving in the next six months - everything has really been waiting for display ghostscript, but that is almost here now (it took ages to raise the money to pay for its development). With DGS in place, gui development can pick up, and it's already nearly good enough to start development on some apps. GNUstep now has a makefiles package, which does the most important part of PBs job. Perhaps, if anyone actually wanted to start on IB, we would begin to get some real progress on applications - that's part of what's holding me back - I don't really want to code UI stuff by hand. <SNIP> >> GNustep has the same interface as Rhapsody, but it is not compatible in the >> sense that you could compile an application on Rhapsody and run it on >> GNUstep. > > GNUStep actually has the API of the _public_ side of OpenStep - whereas >most OpenStep apps use a number of additional "goodies" that are not strictly >part of the standard. I have fears that porting efforts (if it ever "ships") >will be entirely non-trivial, notably with the updates OS is getting in >Rhapsody. Well GNUstep is 'officially' coded to the OpenStep spec. In practice, it's coded to the Rhapsody documentation, with many classes that aren't in the OpenStep spec. So porting may not be as much of a problem as you seem to think. > Some will take this as "downer" on GNU or GNUStep. Far from it, I think >nothing would help Rhapsody more than a good GNUStep. However it seems to me >that GNU just isn't set up for really large projects like Hurd or GNUStep, >and that's making release times far too long for my liking. Well for sure GNU isn't set up for it - it's voluntary after all. Perhpas you would like to help?
From: Ken MacLeod <ken@bitsko.slc.ut.us> Newsgroups: comp.sys.next.programmer Subject: Re: OpenStep on top of Linux? Date: 22 Apr 1998 08:17:26 -0500 Organization: PSINet Message-ID: <m3zphe2dy1.fsf@biff.bitsko.slc.ut.us> References: <Pine.LNX.3.95.980420091603.21479B-100000@fabre.act.qc.ca> <6hfn98$qhk$1@ns3.vrx.net> <6hgrd2$jpe$1@ralph.vnet.net> <6hhn8h$e3g$2@ns3.vrx.net> <353cd195.0@nnrp1.news.uk.psi.net> richard@brainstorm.co.uk. (Richard Frith-Macdonald) writes: > Perhaps, if anyone actually wanted to start on IB, we would begin to > get some real progress on applications - that's part of what's > holding me back - I don't really want to code UI stuff by hand. I don't know if this would do what you want, but on the Casbah mailing list we've discussed using XML style and layout engines for GUIs. My preference is for DSSSL-style layout which is based heavily on TeX. We would not only get a layout engine for the rich text widget but a layout engine for the GUI as well. Before an IB app becomes available you could use XML directly. <http://www.ntlug.org/casbah/> -- Ken MacLeod ken@bitsko.slc.ut.us
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: OpenStep on top of Linux? Date: 22 Apr 1998 09:09:05 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6hkc3h$hvs$1@ns3.vrx.net> References: <Pine.LNX.3.95.980420091603.21479B-100000@fabre.act.qc.ca> <6hfn98$qhk$1@ns3.vrx.net> <6hgrd2$jpe$1@ralph.vnet.net> <6hhn8h$e3g$2@ns3.vrx.net> <353cd195.0@nnrp1.news.uk.psi.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: richard@brainstorm.co.uk. In <353cd195.0@nnrp1.news.uk.psi.net> Richard Frith-Macdonald claimed: > I think that things are likely to start moving in the next six months - > everything has really been waiting for display ghostscript, but that is almost > here now (it took ages to raise the money to pay for its development). Well that's good. > With DGS in place, gui development can pick up, and it's already nearly good > enough to start development on some apps. But certianly not in the same fashion as under OpenStep. IB and PB are basically non-functional. > Perhaps, if anyone actually wanted to start on IB, we would begin to get some > real progress on applications - that's part of what's holding me back - I don't > really want to code UI stuff by hand. For sure. > Well for sure GNU isn't set up for it - it's voluntary after all. Perhpas you > would like to help? Sorry, I'm not that good (yet). Maury
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: NSImage oddness Date: 22 Apr 1998 13:58:34 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6hkt2a$id$1@ns3.vrx.net> References: <6hho4f$e3g$4@ns3.vrx.net> <Ers3F8.E59@basil.icce.dev.rug.null.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: tom@basil.icce.dev.rug.null.nl In <Ers3F8.E59@basil.icce.dev.rug.null.nl> Tom Hageman claimed: > Doesn't this just shift the coordinate system one point to the right and one > point down? (or up, depending on the flippedness of the view). Or am I totally > flipped out? Thonk!! Make that PSscale. :-) > Did you try NSImage's -setFlipped: ? It doesn't have one does it? Oh sorry, we're using ImageRep's. Maury
From: Timothy J Luoma <nospam+yes-this-is-a-valid-address@luomat.peak.org> Newsgroups: comp.sys.next.programmer Subject: Re: Adding workspace functionality Date: Wed, 22 Apr 1998 13:21:56 -0400 Organization: @Home Network Message-ID: <Pine.NXT.3.96.980422131604.24330B-100000@luomat> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII To: jonathan klein <jklein@ivy.hampshire.edu> In-Reply-To: <353e2507.0@192.33.12.30> With LaunchBar you can make any app active, and there's a windowPackage.ps patch which lets you use the command+arrow keys to change active windows from among the visible windows. With this combination I can get to any window of any app I want without using the mouse at all. The patches can be found at http://www.peak.org/~luomat/ (as I say there, not for the weak of heart, but for programmer types it should not be a problem. As for your problem of the service not being active, I think the only solution is the one which LaunchBar takes, namely commandeering a certain key combination (in LaunchBar's case, command-space) to bring you app forward. There is some PS code in a file inside the app w rapper for LaunchBar which might show you how to do the same for alt-tab or whatever you want to take over (command-tab would be my suggestion) TjL
From: ADULTS ONLY!!!!!!!!!! Newsgroups: comp.sys.next.programmer Date: Sat, 18 Apr 1998 22:37:50 PDT Subject: 4" CLITS LICKED BY HOT TEENS AND CELEBS! Organization: Email Platinum v.3.1b Message-ID: <35397e04.1@news.codenet.net> 3 OF THE BEST NO BULLSHIT SITES AROUND!!!!!!!!!!!!1 JUST CLICK HERE ONCE AND SEE 4 YOURSELF!!!!!!! http://pics.xxxhosting.net http://pics.xxxhosting.net/teens.html -101 new school age teens every day!! http://pics.xxxhosting.net/celebs.html 69 new celeb pics daily http://pics.xxxhosting.net/asian.html 1000 asian pics daily no bull shit here just free xxx picss!!
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 22 Apr 1998 15:24:58 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6hlg6a$i2l$1@crib.bevc.blacksburg.va.us> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6h0b4v$h5v$3@ns3.vrx.net> <3534C720.3BACAC4@milestonerdl.com> <SCOTT.98Apr15231250@slave.doubleu.com> In article <SCOTT.98Apr15231250@slave.doubleu.com>, scott@doubleu.com (Scott Hess) wrote: > Well, then, what do you think the odds are on GNUStep if Rhapsody > _doesn't_ ship, or Apple kills it off? There are a lot of orphan GUI > toolkits out there. For better or worse, Rhapsody is the center of > the OpenStep universe. If Rhapsody does well, GNUStep will do > alright. If Rhapsody digs a hole, GNUStep won't be far behind. I think that the GNUstep people would disagree.. the project was started precisely _because_ it looked like NEXTSTEP was dying, and they wanted an alternative to live on. I think that the status of Rhapsody is not going to affect GNUstep very much unless/until GNUstep truly becomes mostly/completely source-code compatible with Rhapsody, in which case it will get a big boost. But there were sufficiently many people who were excited about bringing OpenStep to other platforms in a free way despite NEXTSTEP's impending demise to get the project started in the first place. [Newsgroups: line trimmed.]
From: Alex Blakemore <alex@genoa.com> Newsgroups: comp.sys.next.programmer Subject: How can you store nibs in CVS on NT? Date: 21 Apr 1998 06:00:14 GMT Organization: Genoa Software Systems Message-ID: <6hhclf$77q@saturn.genoa.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Has anyone successfully used CVS to store nibs on NT? I've ported the wrap, unwrap scripts to NT and they work by themselves, and I've modified the cvswrappers file to invoke them for nibs. But checking binary files into RCS seems to corrupt them. Its as if the -k 'b' option is totally ignored. It would be helpful to know if anyone has suceeded in using CVS on NT, if only to know its worthwhile to keep plugging away at this. Gawd, I miss Unix :-( Thanks for any help. -- Alex Blakemore alex@genoa.com NeXT, MIME and ASCII mail accepted
Message-ID: <353E39A2.710A4FE2@denalisites.com> From: Paul Holman <pablos@denalisites.com> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Comma Delimited Textfile to NSDictionary Content-Type: text/plain; charset=us-ascii; x-mac-type="54455854"; x-mac-creator="4D4F5353" Content-Transfer-Encoding: 7bit Date: Wed, 22 Apr 1998 18:41:51 GMT NNTP-Posting-Date: Wed, 22 Apr 1998 10:41:51 AKDT Organization: InternetAlaska OK, I'm going to expose myself as a beginner here (too many years of drag & drop spoiled me). I need some help figuring out a simple task: I have a comma delimited text file, i.e.: "Steve","Jobs","Pixar" "Avie","Tevanian","Apple" . . . Which I want to turn into a couple NSDictionaries. I would like to be able to create a separate one for all the "Pixar" entries and one for the "Apple" entries. Anyway, to get me started a pointer on how to convert my data from comma delimited to an NSDictionary would be a big help. Thanks.
From: jrudd@cygnus.com (John Rudd) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 22 Apr 1998 21:01:03 GMT Organization: Cygnus Solutions Message-ID: <6hllqf$cg8$2@cronkite.cygnus.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6h0b4v$h5v$3@ns3.vrx.net> <3534C720.3BACAC4@milestonerdl.com> <SCOTT.98Apr15231250@slave.doubleu.com> <6hlg6a$i2l$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: nurban@crib.bevc.blacksburg.va.us In <6hlg6a$i2l$1@crib.bevc.blacksburg.va.us> Nathan Urban wrote: > In article <SCOTT.98Apr15231250@slave.doubleu.com>, scott@doubleu.com (Scott Hess) wrote: > > > Well, then, what do you think the odds are on GNUStep if Rhapsody > > _doesn't_ ship, or Apple kills it off? There are a lot of orphan GUI > > toolkits out there. For better or worse, Rhapsody is the center of > > the OpenStep universe. If Rhapsody does well, GNUStep will do > > alright. If Rhapsody digs a hole, GNUStep won't be far behind. > > I think that the GNUstep people would disagree.. the project was > started precisely _because_ it looked like NEXTSTEP was dying, and they > wanted an alternative to live on. I think that the status of Rhapsody is > not going to affect GNUstep very much unless/until GNUstep truly becomes > mostly/completely source-code compatible with Rhapsody, in which case it > will get a big boost. But there were sufficiently many people who were > excited about bringing OpenStep to other platforms in a free way despite > NEXTSTEP's impending demise to get the project started in the first place. > Indeed. If it wasn't for Rhapsody, I would have started spending more time doing whatever I could to help the Gnustep project in early 97. Whether it was something as simple as monetary donations (which aren't feasable for me right now cuz now I'm trying to buy a house), testing, or whatever. However, as interested as I am in free software and a free Openstep, I just don't have time to spend on keeping Openstep alive via helping Gnustep. It is less of a priority, in my eyes, because Openstep has a new lease on life at Apple. If that hadn't been the case (or if that ceases to be the case), I would probably consider Gnustep a more justifiable place to spend my resources. -- John "kzin" Rudd jrudd@cygnus.com http://www.cygnus.com/~jrudd Intel: Putting \"I want a pair of Daisy Eagle semi-auto paintball pistols the backward in \in shoulder rigs. Who cares if you win the game as long backward compatible\as you can John Woo as you dive over obstacles?" - anon
From: macghod@concentric.net Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.programmer.codewarrior Subject: mac out in the cold? Date: 23 Apr 1998 16:17:39 GMT Message-ID: <6hnpj3$spq$4@newsfep3.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit OPENSTEP. The best thing since sliced bread. With just some tweaking with interface builder (Note, I am not a next programmer, this is just what I have heard) your app will not only run under RHapsody, but also windows 95(!!!), OPENSTEP, (and maybe winnt? Or just OPENSTEP for winnt?). BUT!, isnt their a operating system missing? Yes their is, macos! Openstep apps dont run on the mac! (Good lord, I just know someone will email me and say, duh! havent you heard of the blue box?) People often mention, why would windows developers want to continue like they are, when with openstep, they can put out apps that run on windows, openstep, and rhapsody! Well, why will mac developers want to program for the mac, when to program for the mac they have to do a seperate app? Program on rhapsody only, and you get windows, OPENSTEP, rhapsody. If you want your software to run on people macs who are using os 8.1 and lower (will 8.2 run rhapsody apps?) you have to do a whole other application!
From: Sebastien Berthet <sberthet@ina.fr> Newsgroups: comp.sys.next.programmer Subject: how can one get a PICT file from a PS NSImage ? Date: Thu, 23 Apr 1998 18:26:39 +0200 Organization: INA Message-ID: <353F6BBF.2F0C98E@ina.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello, On YB/NT, I'm trying to render the PS content of an NSImageView to a "Quickdrawable" format. First, I tried with NSBitmapImageRep initialized with initWithFocusedViewRect. But the resulting bitplanes are not compatible with Quickdraw so that I cannot make any magic cast. Second, I wanted to use NSPICTImageRep. Unfortunately, you can't init a NSPICTImageRep with anything but PICT datas. If I'd manage to create a PICT file from a NSImageView, I would easely import it to Quickdraw. So, how can one get a PICT file from a PS NSImage ? Is there any pointers or sample code ? thanks -- Sebastien Berthet
Subject: CVS pserver error Newsgroups: comp.sys.next.programmer From: charles@drbs.com (Charles C. Hocker) Message-ID: <353ea806.0@news11.kcdata.com> Date: 23 Apr 1998 02:31:34 -0600 Hello, I was wondering if someone could tell me what I am doing wrong in setting up CVS using a client/server configuration using a pserver. Here is what I did to both the client and the server: 1) I edited my /etc/services file and added: cvspserver 2401/tcp 2) I edited my /etc/inetd.conf file and added: cvspserver stream tcp nowait root /usr/local/bin/cvs cvs -b /usr/local/bin --allow-root=/LocalDeveloper/CVS 3) I created a $CVSROOT/CVSROOT/passwd file that contained my userid and password that I copied from NetInfo on the server. I wrote a small Perl script to verify that I had the correct password (ie the Perl script generated the same password). The passwd file contain one entry: foo:0Qosaw9G39ZvA The un-encrypted password is "hello." 4) While trying to track down this error, I added the following entries into the NetInfo database in both the client and server: /services/cvspserver name=cvspserver port=2401 protocol=tcp This information is the same that was added to the /etc/services file. 5) I also created a $CVSROOT/CVSROOT/config file on my server with the following two lines: RCSBIN=/usr/local/bin SystemAuth=no 6) I set my $CVSROOT variable in my client machine to : :pserver:foo@bar.foobar.com:/LocalDeveloper/CVS in both my .cshrc and my .login using the appropate syntax. When I try to logon the following occurs: mark> cvs login (Logging in to foo@bar.foobar.com) CVS password: cvs [login aborted]: unrecognized auth response from bar.foobar.com: Usage: cvs [cvs-options] command [command-options-and-arguments] mark> Any suggestions? Charles -- --------------------------------------------------------------- Charles C. Hocker charles@intrawebioe.com ASCII, MIME & NeXTmail american@aztec.asu.edu "Food is Power. We use it to change behavior. Some May Call that Bribery. We Do Not Apologize." Catherine Bertini, executive director, UN World Food Program, Beijing, China, UN 4th World Conference on Women, Sept. 1995. ---------------------------------------------------------------
From: Curtis Crowson <curtis_crowson@emory.org> Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.programmer.codewarrior Subject: Re: mac out in the cold? Date: Thu, 23 Apr 1998 12:52:30 -0400 Organization: Emory University Message-ID: <353F71CD.97640F8E@emory.org> References: <6hnpj3$spq$4@newsfep3.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Apple had this thought also. They are planning (correct me if I'm wrong ;-) to release a version of the API's that run on top of the MacOS. macghod@concentric.net wrote: > OPENSTEP. The best thing since sliced bread. With just some tweaking with > interface builder (Note, I am not a next programmer, this is just what I have > heard) your app will not only run under RHapsody, but also windows 95(!!!), > OPENSTEP, (and maybe winnt? Or just OPENSTEP for winnt?). BUT!, isnt their > a operating system missing? Yes their is, macos! Openstep apps dont run on > the mac! (Good lord, I just know someone will email me and say, duh! havent > you heard of the blue box?) > > People often mention, why would windows developers want to continue like they > are, when with openstep, they can put out apps that run on windows, openstep, > and rhapsody! Well, why will mac developers want to program for the mac, > when to program for the mac they have to do a seperate app? Program on > rhapsody only, and you get windows, OPENSTEP, rhapsody. If you want your > software to run on people macs who are using os 8.1 and lower (will 8.2 run > rhapsody apps?) you have to do a whole other application!
Newsgroups: comp.sys.next.programmer From: dfevans@bbcr.uwaterloo.ca (David Evans) Subject: Re: Adding workspace functionality Message-ID: <893298368.812859@globe.uwaterloo.ca> Sender: news@watserv3.uwaterloo.ca Organization: University of Waterloo References: <353e2507.0@192.33.12.30> Cache-Post-Path: globe.uwaterloo.ca!unknown@bcr11.uwaterloo.ca Date: Thu, 23 Apr 1998 02:28:27 GMT In article <353e2507.0@192.33.12.30>, jonathan klein <jklein@ivy.hampshire.edu> wrote: > I wrote an app to do this, and tried placing it as a service. >It works great when the service is active, but of course, the service >isn't always active!! > Frank and Timothy have given good descriptions about beating up on the DPS input handler and such. You may also want to look into the LaunchPaths Workspace default, which allows you to start apps at login time. -- David Evans (NeXTMail/MIME OK) dfevans@bbcr.uwaterloo.ca Computer/Synth Junkie http://bbcr.uwaterloo.ca/~dfevans/ University of Waterloo "Default is the value selected by the composer Ontario, Canada overridden by your command." - Roland TR-707 Manual
Newsgroups: comp.sys.next.programmer From: valerie@deneb (Moukdarath Valerie) Subject: EOF 2.2 with ODBC stored procedure Message-ID: <ErvLvy.6tG@x-lan.alienor.fr> Sender: news@x-lan.alienor.fr Organization: x&lan Date: Thu, 23 Apr 1998 16:59:57 GMT Hi, I have an application developed on OpenStep 4.2 runnig on Windows NT. I am using EOF 2.2 with ODBC Adaptor to connect to SQL AnyWhere. I try to use stored procedure with arguments (2) but the application allways crash. If I use stored procedure without argument it is OK. The problem is how can I use stored procedure with arguments. valerie.
Subject: Re: Apple developer program Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <joe.ragosta-0904980642200001@elk33.dol.net> <6gitpj$csn$1@interport.net> <joe.ragosta-0904980134360001@wil68.dol.net> In-Reply-To: <joe.ragosta-0904980134360001@wil68.dol.net> From: jnelson@jnelson.com Message-ID: <gEN%.411$ub.297766@news.internetMCI.com> Date: Thu, 23 Apr 1998 20:37:32 GMT NNTP-Posting-Date: Thu, 23 Apr 1998 16:37:32 EDT On 04/08/98, Joe Ragosta wrote: >Sure. They should encourage everyone. > >But how do you allocate limited resources? Apple has decided that >spending the money on TV ads or magazine ads is a better place to put >scarce resources than subsidizing hobbyists. Apple Computer *started* as a bunch of hobbiests in a garage. Ignoring them will be like cutting their own throats. -- Dark Hacker | e-mail hacker@computation.com Fortress Of Computation | Web http://www.computation.com/pub/hacker/ ____________________________________________________________________ "Life would be so much easier if only we had the source code."
From: huner@korax.net Newsgroups: comp.sys.next.programmer Subject: OPENSTEP MACH or OPENSTEP ENTERPRISE Date: Thu, 23 Apr 1998 16:17:52 -0600 Organization: Deja News - The Leader in Internet Discussion Message-ID: <6hob5v$ehu$1@nnrp1.dejanews.com> I am itching to get this monster dual PII machine. Thing is I am a OPENSTEP fan. I could install OPENSTEP MACH 4.2 and NT WS4.0 for dual boot. But, OPENSTEP MACH won't see the second PII, so I figure it's a kind of a waste. So, I'm thinking that I could get OPENSTEP ENTERPRISE for NT. That way I could still reap the power of OPENSTEP development and use the dual PII power of the machine. Am I nuts? Are people out there using OPENSTEP on NT? Are they happy developing on it? Are there problems with developing on OPENSTEP on NT? Is OPENSTEP on NT with a dual PII fast and stable? Please provide your input before I blow alot of money. Thanks, Ozgur Huner -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: macghod@concentric.net Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer,comp.sys.next.sysadmin Subject: Am I doing something wrong?? Followup-To: comp.sys.next.programmer Date: 23 Apr 1998 21:34:50 GMT Message-ID: <6hoc5q$86f$1@newsfep2.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NOTE: FOLLOWUPS TO COMP.SYS.NEXT.PROGRAMMER
From: macghod@concentric.net Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.programmer.codewarrior Subject: Re: mac out in the cold? Date: 23 Apr 1998 21:52:15 GMT Message-ID: <6hod6f$86f$3@newsfep2.sprintmail.com> References: <6hnpj3$spq$4@newsfep3.sprintmail.com> <6hni4v$453$1@ns3.vrx.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: maury@remove_this.istar.ca In <6hni4v$453$1@ns3.vrx.net> Maury Markowitz wrote: > In <6hnpj3$spq$4@newsfep3.sprintmail.com> macghod@concentric.net claimed: > > a operating system missing? Yes their is, macos! Openstep apps dont run > on > > the mac! (Good lord, I just know someone will email me and say, duh! > havent > > you heard of the blue box?) > > No, we'll say "duh! haven't you heard of YellowBox for MacOS?". You clipped out the part about os 8. So you are saying openstep apps will run on a g3 running os 8.1? Their are about 15 million macs out their that are running os 8.1 or lower, are you saying openstep will run on these?
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.programmer.codewarrior Subject: Re: mac out in the cold? Date: 23 Apr 1998 14:10:39 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6hni4v$453$1@ns3.vrx.net> References: <6hnpj3$spq$4@newsfep3.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: macghod@concentric.net In <6hnpj3$spq$4@newsfep3.sprintmail.com> macghod@concentric.net claimed: > a operating system missing? Yes their is, macos! Openstep apps dont run on > the mac! (Good lord, I just know someone will email me and say, duh! havent > you heard of the blue box?) No, we'll say "duh! haven't you heard of YellowBox for MacOS?". Apparently not. Maury
From: huner@korax.net Newsgroups: comp.sys.next.programmer Subject: WebObjects, Databases and Platforms Date: Thu, 23 Apr 1998 15:32:51 -0600 Organization: Deja News - The Leader in Internet Discussion Message-ID: <6ho8hj$amf$1@nnrp1.dejanews.com> To: huner@korax.net Hello to all OPENSTEP, Rhapsody, NT and WebObjects developers! I am requesting your input on developing for WebObjects on OPENSTEP or NT. Have people found any obstacles in developing WebObjects on NT? Is OPENSTEP a better choice? What combinations of database engines and WebObjects have been successful and what have been a nuisance? What are the deployment license costs for the the database systems you are using? And, for WebObjects, is the deployment license cost worth it? I am planning a project which will require bringing together heterogeneous data on multiple database servers and use WebObjects to deploy apps on an intranet. If you have any experiences, good or bad, on these issues please share. Thanks, Ozgur Huner -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 23 Apr 1998 08:10:11 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6hmt13$i8o$1@ns3.vrx.net> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <6h0b4v$h5v$3@ns3.vrx.net> <3534C720.3BACAC4@milestonerdl.com> <SCOTT.98Apr15231250@slave.doubleu.com> <6hlg6a$i2l$1@crib.bevc.blacksburg.va.us> <6hllqf$cg8$2@cronkite.cygnus.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: jrudd@cygnus.com In <6hllqf$cg8$2@cronkite.cygnus.com> John Rudd claimed: > However, as interested as I am in free software and a free Openstep, I just > don't have time to spend on keeping Openstep alive via helping Gnustep. It > is less of a priority, in my eyes, because Openstep has a new lease on life > at Apple. If that hadn't been the case (or if that ceases to be the case), I > would probably consider Gnustep a more justifiable place to spend my > resources. Then let us all hope that they pull it off. In terms of the tech presented so far, I'm terribly happy with what I've seen so far. Maury
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody Graphics Questions Date: 24 Apr 1998 03:26:56 GMT Organization: Digital Fix Development Message-ID: <6hp0q0$b59$1@news.digifix.com> References: <zegelin-2404981257030001@dialup35.canb.ispsys.net> In-Reply-To: <zegelin-2404981257030001@dialup35.canb.ispsys.net> On 04/23/98, Peter wrote: >Hi! > >Ive been interested in Rhapsody for sometime but havent actually been able >to try it (havent even seen it ). However, Id be grateful if someone would >answer the following graphics questions for me: > >€ Are thick lines drawn about the center of their path rather than below >and to the right as with MacOs? > They are drawn centered on the path. >€ Are sloped lines drawn at different widths depending on slope so that >they dont look thicker than vertical & horizontal lines of the same width? > The widths are the consistent. >€ Are lines and curves automatically antialiased as they're drawn and if >not is there an easy way of doing it? > They are not automatically anti-aliased.. however, the Rhapsody windowserver does support anti-aliasing... so it is possible for a program to turn that on. >€ To do graphics in Rhapsody do you have to learn postscript as well? Or >is it just advisable? You don't need to learn Postscript. As far as it being advisable... I'd learn it as I need to when I come across options that aren't covered by classes (most is covered by classes) > >€ Is Display Postscript slow? The only reason I ask is that displaying pdf >files on an LC is painfully slow and Im assuming (wrongly?) that Acrobat >Reader uses postscript somehow. > On an LC???? Rhapsody isn't going to run on an LC... DPS isn't slow. >Regions: >€ Are regions and their traps (DiffRgn,UnionRgn etc and especially >BitMapToRgn) fully supported? Is their usage much the same as on MacOs? A >pointer to info on regions & Rhaposdy would be great! > There are no regions on Rhapsody.. You have define paths, which you can then hit-test... >Any help would be much appreciated. > >Thanks in advance > >Peter > -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: "Zico" <ZicoKnows@hotmail.com> Newsgroups: comp.sys.next.programmer Subject: Re: Am I doing something wrong?? Date: Thu, 23 Apr 1998 18:42:03 -0400 Organization: All USENET -- http://www.Supernews.com Message-ID: <6hogi4$63f$1@supernews.com> References: <6hoc5q$86f$1@newsfep2.sprintmail.com> x-no-archive: yes macghod@concentric.net wrote: >NOTE: FOLLOWUPS TO COMP.SYS.NEXT.PROGRAMMER Yes. :-) Z
From: zegelin@actonline.com.au (Peter) Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody Graphics Questions Date: 24 Apr 1998 04:01:49 GMT Organization: CyberOne Pty Ltd Message-ID: <zegelin-2404981402300001@dialup35.canb.ispsys.net> References: <zegelin-2404981257030001@dialup35.canb.ispsys.net> <6hp0q0$b59$1@news.digifix.com> Cache-Post-Path: genie.cyberone.com.au!unknown@dialup35.canb.ispsys.net Dont worry I wasnt going to try to run it on an LC :-) Its just the difference in speed displaying pdf files and most other things is quite large. I only really started hanging out for a new machine after trying to read a 600 page pdf file on an LC with a 12" screen. Its bad. BTW thanks for the info. You must be up pretty late! Regards Peter
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <35397e04.1@news.codenet.net> Control: cancel <35397e04.1@news.codenet.net> Date: 23 Apr 1998 09:07:20 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.35397e04.1@news.codenet.net> Sender: ADULTS ONLY!!!!!!!!!! Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: gierkeNOSPAM@delaware.infi.net (Patrick William Gierke) Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.programmer.codewarrior Subject: Re: mac out in the cold? Date: Thu, 23 Apr 1998 19:23:25 -0500 Organization: Irish, and damn proud of it! Message-ID: <gierkeNOSPAM-2304981923270001@pm1-24.ile.infi.net> References: <6hnpj3$spq$4@newsfep3.sprintmail.com> <353F71CD.97640F8E@emory.org> In article <353F71CD.97640F8E@emory.org>, Curtis Crowson <curtis_crowson@emory.org> wrote: > Apple had this thought also. They are planning (correct me if I'm wrong ;-) to > release a version of the API's that run on top of the MacOS. > > macghod@concentric.net wrote: > > > OPENSTEP. The best thing since sliced bread. With just some tweaking with > > interface builder (Note, I am not a next programmer, this is just what I have > > heard) your app will not only run under RHapsody, but also windows 95(!!!), > > OPENSTEP, (and maybe winnt? Or just OPENSTEP for winnt?). BUT!, isnt their > > a operating system missing? Yes their is, macos! Openstep apps dont run on > > the mac! (Good lord, I just know someone will email me and say, duh! havent > > you heard of the blue box?) > > > > People often mention, why would windows developers want to continue like they > > are, when with openstep, they can put out apps that run on windows, openstep, > > and rhapsody! Well, why will mac developers want to program for the mac, > > when to program for the mac they have to do a seperate app? Program on > > rhapsody only, and you get windows, OPENSTEP, rhapsody. If you want your > > software to run on people macs who are using os 8.1 and lower (will 8.2 run > > rhapsody apps?) you have to do a whole other application! Apple intends to release a set of shared libraries that will make up the "Yellow Box for MacOS/Windows". All you need is to install them, and you can run anything that is compiled for Rhapsody -- WARNING! The preceeding message consisted of statements that may or may not contain factual material. Use of these messages poses a serious threat to mental ignorance. Studies have proven that factual material that contains opinion expands the level of thinking experienced by the reader. Under no circumstances are we (the author) to assume any responsibility for gain of knowledge.
From: macghod@concentric.net Newsgroups: comp.sys.next.programmer Subject: Re: Am I doing something wrong?? Date: 24 Apr 1998 06:55:11 GMT Message-ID: <6hpd0f$h0l$1@newsfep1.sprintmail.com> References: <6hoc5q$86f$1@newsfep2.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: macghod@concentric.net I apologize for this post, I had a message in here, I do not know why it turned out blank? Anyways here was the original post, actually the gist was this: With several programs, I have been unable to run them because I have been unable to save the prefs! I had the resolution at 640 or 800, and the set buttons were bellow the bottom of the screen, and their was no scroll bar, thus no way to save the preferences!! This happened in pppmonitor (I didnt have a card that supported over 640, and then even in radical news (I set it to 32 bit color at 800/600) Now I have to ask if I am doing something wrong, cause that could be the case. It would seem hard to believe that programmers would do this on purpose?
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: sub images (was) Clocks Date: 23 Apr 1998 12:18:01 GMT Organization: Idiom Communications Message-ID: <6hnbhp$g3k$3@news.idiom.com> References: <6hdqcj$333$1@blackice.winternet.com> <ErqC1q.8uG@basil.icce.dev.rug.null.nl> <6hgpdb$d8e$1@news.idiom.com> <6hhnco$e3g$3@ns3.vrx.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: maury@remove_this.istar.ca Maury Markowitz may or may not have said: -> In <6hgpdb$d8e$1@news.idiom.com> John C. Randolph claimed: -> > No, it sure didn't and that's quite unfortunate. I've got some code that I -> > need to hand over to Don that implements a sub-image class. You create sub -> > images by sending an NS image a message -> > like [myImage subImage: subImageRect] where subImageRect has to fit within -> > the bounds of the receiver. -> -> It's hard to be sure, but are you referring to placing an image inside the -> rect of another view? Why not do a... -> -> [[[myImage imageReps] objectAtIndex:0] drawRect:subImageRect]; -> -> Is that the same thing? That's pretty much what my subImage code is doing under the hood, but it's very handy when (for example) I need to draw LED images, and I can stick them in an array, and tell them all to draw themselves into the clockview with a -makeObjectsPerform: call. (I use another trick to have them keep track of where to draw so that they don't collide.) Of course, the *right* way to do this is to turn the clockbits.tiff file into a bitmap font, but I haven't gotten that ambitious yet. -jcr
From: macghod@concentric.net Newsgroups: comp.sys.next.programmer Subject: Re: OpenStep on top of Linux? Date: 24 Apr 1998 09:30:11 GMT Message-ID: <6hpm33$3vj$4@newsfep3.sprintmail.com> References: <Pine.LNX.3.95.980420091603.21479B-100000@fabre.act.qc.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: ama@fabre.act.qc.ca In a similiar vein, can you run windows apps in openstep somehow? Those scoundrels who make firstclass dont have a fc client for unix, and I dont want to have to boot into windows to run it, is their someway to run it on openstep? Maybe a emulator? I heard linux can do this? running on a PC with OPENSTEP, the year 2000 bug-free operating system Email address is invalid, please email macghod@concentric.net
From: leigh@NOSPAMcs.uwa.edu.au (Leigh Smith) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program Date: 24 Apr 1998 09:29:37 GMT Organization: The University of Western Australia Message-ID: <6hpm21$bv3$1@enyo.uwa.edu.au> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <joe.ragosta-0804980858150001@wil77.dol.net> <mazulauf-0804981503390001@sneezy.met.utah.edu> <joe.ragosta-0804981906460001@elk68.dol.net> <01bc4540$d4a15e60$3af0bfa8@davidsul> <6gh8v1$oop$6@ironhorse.plsys.co.uk> <joe.ragosta-0904980642200001@elk33.dol.net> <6gitpj$csn$1@interport.net> <joe.ragosta-0904980134360001@wil68.dol.net> <gEN%.411$ub.297766@news.internetMCI.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: jnelson@jnelson.com In <gEN%.411$ub.297766@news.internetMCI.com> jnelson@jnelson.com wrote: > > Apple Computer *started* as a bunch of hobbiests in a garage. > Ignoring them will be like cutting their own throats. > Agreed. However, Apple was in a nose dive until Jobs+new management could pull it out (and they are still doing that). Clearly the problem wasn't hobbyists (I don't know of another system that has such love as the Mac by its users), but the perspectives of large corporations and institutions making purchasing decisions becoming reluctant buying hardware that could well become orphaned. I just received word that existing associate developers have been upgraded to select for the rest of their renewal period minus support queries, which includes seeding, so I would say that most developers that are already on Rhapsody DR1 seeds will probably continue to receive future seeds. -- Leigh Computer Science, University of Western Australia Smith +61-8-9380-3778 leigh@cs.uwa.edu.au (NeXTMail/MIME) Microsoft - never has so much been made by so few, by pushing so much of so little, on so many with so little resistance.
From: appology for ADULTS ONLY Newsgroups: comp.sys.next.programmer Date: Sun, 19 Apr 1998 14:43:09 PDT Subject: ERROR IN POSTING Organization: Email Platinum v.3.1b Message-ID: <353a6046.0@news.codenet.net> THIS PAST WEEKEND THERE PROBABLY WAS AN ERROR IN POSTING TO YOUR NEWS GROUP. AN ADULTS ONLY MESSAGE FOR http://pics.xxxhosting.net WAS SENT OUT BY ACCIDENT 2x IF YOU HAVE NOT YET READ THIS LETTER WITH THIS URL ABOVE WHEN YOU SEE ADULTS ONLY IN THE FROM OR SUBJECT AREA DELETE IT THANK YOU!
Newsgroups: comp.sys.next.programmer From: valerie@vega (Moukdarath Valerie) Subject: EOF 2.2 with ODBC Message-ID: <ErwqMA.A4C@x-lan.alienor.fr> Sender: news@x-lan.alienor.fr Organization: x&lan Date: Fri, 24 Apr 1998 07:39:45 GMT Hi, I am working with openStep 4.2 on WNT and EOF 2.2. I am connecting to SQLAnywhere with ODBC. I did not succeed in using stored procedure with arguments. It always end with errors saying there are too many arguments. I' ve tried to use a stored procedure without any argument and it is OK. My problem is : how to send a stored procedure with arguments via ODBC. thank for response. valerie.
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody Graphics Questions Date: 24 Apr 1998 12:26:40 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <6hq0e0$ek9$1@pump1.york.ac.uk> References: <zegelin-2404981257030001@dialup35.canb.ispsys.net> <6hp0q0$b59$1@news.digifix.com> On 24 Apr 1998 03:26:56 GMT, Scott Anguish <sanguish@digifix.com> wrote: > You don't need to learn Postscript. As far as it being > advisable... I'd learn it as I need to when I come across options that > aren't covered by classes (most is covered by classes) but you do need to learn some postscript if you're wanting to draw lines (as he was suggesting...) because AFAIK there are no AppKit classes supporting basic drawing. rog.
From: vtnxsovb@jmx.net Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.nsc,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.pen Subject: SHARE THE WEALTH Date: 24 Apr 98 14:25:44 GMT Organization: Money Inc. Message-ID: <3540a0e8.0@www.an.cc.mn.us> ******************************************************* PLEASE READ ON ABOUT HOW TO BE SUCCESSFUL!! ******************************************************* 1. IS THIS REALLY LEGAL? I called a lawyer first. The lawyer was a little skeptical that I would actually make any money but he said it WAS LEGAL if I wanted to try it. I told him it sounded a lot like a chain letter but the details of the system (SEE BELOW) actually made it a legitimate legal business. 2. WOULD THE POST OFFICE BE OKAY WITH THIS? I called them: 1-800-725-2161 and they confirmed THIS IS ABSOLUTELY LEGAL! (See Title 18,h sections1302 NS 1341 of Postal Lottery Laws). This clarifies the program of collecting names and addresses for a mailing list. 3. IS THIS MORAL? Well, everyone who sends me a buck has a good chance of getting A LOT of money ... a much better chance than buying a lottery ticket!!! 4. DOES THIS REALLY WORK? If you are a non believer, pick an address below and mail that person asking how they have done. If it didn’t work there wouldn’t be so many list like this on the Internet. ============ HOW IT WORKS: ============ STEP 1: Get 6 separate pieces of paper and write the following on each piece of paper: PLEASE ADD ME TO YOUR MAILING LIST. $1 US DOLLAR PROCESSING FEE IS ENCLOSED. (THIS IS KEY AS THIS IS WHAT MAKES IT LEGAL SINCE YOU ARE PAYING FOR AND LATER OFFERING A SERVICE). Now get 6 $1.00 bills and place ONE inside EACH of the 6 pieces of paper so the bill will not be seen through the envelope to prevent theft/robbery. Then, place one paper in each of the 6 envelopes and seal them. You should now have 6 sealed envelopes, each with a piece of paper stating the above phrase and a U.S. $1.00 bill. Mail the envelopes to the following addresses: 1) J. Turley P.O. Box 50604 Knoxville, TN 37950-0604 USA 2) K. Agler 109 Fraternity Ct. Chapel Hill, NC 27516 USA 3) Karen Goldstein 2421 Somerset Street Philadelphia, Pa. 19134 USA 4) Heather 1253 Hampshire Drive Whitehall, PA 18052 USA 5) John Pacitti 2600 Mifflin Street Philadelphia, PA 19145 USA 6) B. J. MILLARD 721 LAURA LANE YORK, PA 17402 USA STEP 2: Now take the #1 name off the list that you see above, move the other names up (6 becomes 5, 5 becomes 4, etc...) and add YOUR name as number 6 on the list. (If you want to remain anonymous, put a nickname, but the address MUST be correct. It, of course, MUST contain your country, state/district/area, zip code, etc!!! You wouldn't want your money to fly away, would you?!?!). STEP 3: Now post your amended article to at least 200 newsgroups. Remember, 200 postings is just a guideline. The more you post, the more money you make! NOTE: IN MANY NEWSGROUPS THERE ARE PEOPLE THAT DELETE THIS KIND OF MESSAGES I RECOMEND YOU TO POST 1 OR EVEN 2 TIMES A WEEK TO 300 NEWSGROUPS SO OTHER PEOPLE CAN SEE YOUR MESSAGES.THIS IS A WAY TO INCREASE THE POSSIBILITIES FOR YOU TO GET MORE MONEY!! ------------------------------------------------------------------------ The TOP 200 newsgroups can be found at ---> www.op.net/usenet-stats.html *** BOTS *** Bots are small computer programs on a usenet server. 1) Bots look for certain characters in the "Subject:" field of your newsgroup posting. 2) Bots also look for "multiple postings". 3) If a Bot discovers any of the above, it will delete your posting. 4) Then send you a nasty e-mail. ***OUTSMART THE BOTS*** 1) You will make a lot MORE money if you outsmart the BOTS. 2) Post your message only ONCE. 3) Do NOT use characters such as (! $ % + # & * @ ?) in the "Subject" field. ------------------------------------------------------------------------ > ------------------------------------------------------------------------ > **REMEMBER, THE MORE NEWSGROUPS YOU POST IN, THE MORE MONEY YOU > WILL MAKE!! BUT YOU HAVE TO POST A MINIMUM OF 200** > That's it! You will begin receiving money from around the world within > days! You may eventually want to rent a P.O. Box due to the large > amount of mail you receive. If you wish to stay anonymous, you can > invent a name to use, as long as the postman will deliver it. > **JUST MAKE SURE ALL THE ADDRESSES ARE CORRECT.** >------------------------------------------------------------------------- ================= Now the WHY part: ================= Out of 200 postings, say I receive only 5 replies (a very low example). So then I made $5.00 with my name at #6 on the letter. Now, each of the 5 persons who just sent me $1.00 make the MINIMUM 200 postings, each with my name at #5 and only 5 persons respond to each of the original 5, that is another $25.00 for me, now those 25 each make 200 MINIMUM posts with my name at #4 and only 5 replies each, I will bring in an additional $125.00! Now, those 125 persons turn around and post the MINIMUM 200 with my name at #3 and only receive 5 replies each, I will make an additional $626.00! OK, now here is the fun part, each of those 625 persons post a MINIMUM 200 letters with my name at #2 and they each only receive 5 replies, that just made me $3,125.00!!! Those 3,125 persons will all deliver this message to 200 newsgroups with my name at #1 and if still 5 persons per 200 newsgroups react I will receive $15,625,00! With a original investment of only $6.00! AMAZING! And as I said 5 responses is actually VERY LOW! Average is probable 20 to 30! So lets put those figures at just 15 responses per person. Here is what you will make: at #6 $15.00 at #5 $225.00 at #4 $3,375.00 at #3 $50,625.00 at #2 $759,375.00 at #1 $11,390,625.00 When your name is no longer on the list, you just take the latest posting in the newsgroups, and send out another $6.00 to names on the list, putting your name at number 6 again. And start posting again. The thing to remember is that thousands of people all over the world are joining the Internet and reading these articles everyday, JUST LIKE YOU are now!! So can you afford $6.00 and see if it really works?? I think so... People have said, "what if the plan is played out and no one sends you the money? What are the chances of that happening when there are tons of new honest users and new honest people who are joining the Internet and newsgroups everyday and are willing to give it a try? Estimates are at 20,000 to 50,000 new users, every day, with thousands of those joining the actual Internet. Remember, play FAIRLY and HONESTLY and this will work. You just have to be honest. ** By the way, if you try to deceive people by posting the messages with your name in the list and not sending the money to the rest of the people already on the list, you will NOT get as much. Someone I talked to knew someone who did that and he made practically nothing, and that's after seven or eight weeks! Then he sent the 6 $1.00 bills, people added him to their lists, and in 4-5 weeks he had over $10k. This is the fairest and most honest way I have ever seen to share the wealth of the world without costing anything but your time!!! You also may want to buy mailing and e-mail lists for future dollars. Make sure you print this article out RIGHT NOW! Also, try to keep a list of everyone that sends you money and always keep an eye on the newsgroups to make sure everyone is playing fairly. Remember, HONESTY IS THE BEST POLICY. You don't need to cheat the basic idea to make the money!! GOOD LUCK to all and please play fairly and reap the huge rewards from this, which is tons of extra CASH. Please remember to declare your extra income. Thanks once again... ________________________________________________________________________
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.nsc,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.pen Subject: cmsg cancel <3540a0e8.0@www.an.cc.mn.us> Control: cancel <3540a0e8.0@www.an.cc.mn.us> Date: 24 Apr 1998 14:32:44 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.3540a0e8.0@www.an.cc.mn.us> Sender: vtnxsovb@jmx.net Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: yqrcpsye@jmx.net Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.nsc,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.pen Subject: Work At Home Free Business Date: 24 Apr 98 14:36:38 GMT Organization: Money Inc. Message-ID: <3540a376.0@www.an.cc.mn.us> You have just come across the opportunity you've been waiting for. recently featured in the Wall Street Journal and the FOX News Network,Freedomstarr Communications, Inc. (FCI) proudly offers you the BEST Internet Business Opportunity available. (FREE) FREE WEB PAGE, COMMISSIONS NO COST AT ALL Whether you could use an extra few hundred dollars per month, or up to an additional $5,000 or more monthly, look no further. Freedomstarr is FREE. GO to: http://freedomstarr.com/?pa2018497
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.nsc,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.pen Subject: cmsg cancel <3540a376.0@www.an.cc.mn.us> Control: cancel <3540a376.0@www.an.cc.mn.us> Date: 24 Apr 1998 14:44:25 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.3540a376.0@www.an.cc.mn.us> Sender: yqrcpsye@jmx.net Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Lars Immisch <lars@ibp.de> Newsgroups: comp.sys.next.programmer Subject: Re: How can you store nibs in CVS on NT? Date: Thu, 23 Apr 1998 16:25:17 +0200 Organization: Immisch, Becker & Partner Message-ID: <353F4F4D.15A1@ibp.de> References: <6hhclf$77q@saturn.genoa.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Alex Blakemore wrote: > > Has anyone successfully used CVS to store nibs on NT? > I've ported the wrap, unwrap scripts to NT and they work by themselves, > and I've modified the cvswrappers file to invoke them for nibs. > > But checking binary files into RCS seems to corrupt them. > Its as if the -k 'b' option is totally ignored. I remember problems when the -k b wasn't given when the file was added to the repository, but with my 1.9 cvs setup, binary files are positively working. Are you certain you have a sufficiently recent version of rcs on your server? I am using rcs 5.7. Anything before 5.6 is too old (please check the cvs docs about the exact version required). > It would be helpful to know if anyone has suceeded in using CVS on NT, > if only to know its worthwhile to keep plugging away at this. I cannot say anything about nibs on NT (I would guess you need a cvswrapper in any case to do this), but I am using cvs on NT without problems, so don't stop yet :-) Cheers, Lars -- mailto:immisch@pobox.com http://pobox.com/~immisch Yesterdays yellow yoyo can make you yawn today
Subject: Re: Newby needs help Newsgroups: comp.soft-sys.nextstep,comp.sys.mac.advocacy,comp.sys.next.misc,comp.sys.next.programmer References: <01bc46e5$6b4880a0$24f0bfa8@davidsul> <6god24$lq4$1@news.rt66.com> <01bd6585$58835660$0b0ba8c0@woohoo> <35366F1B.98331849@forsee.tcp.co.uk> In-Reply-To: <35366F1B.98331849@forsee.tcp.co.uk> From: hacker@computation.com Message-ID: <_Z101.966$ub.1227872@news.internetMCI.com> Date: Fri, 24 Apr 1998 15:12:58 GMT NNTP-Posting-Date: Fri, 24 Apr 1998 11:12:58 EDT On 04/16/98, Robert Forsyth wrote: >Yes, but, doesn't Openstep only see the MSDOS 8.3 name. OpenStep 4.2 totally broke the NeXT implementation of DOS file systems. Besides getting confused when you try to rename DOS files under OpenStep, I am unable to copy files to the DOS partitions which have filename lengths greater than 8 characters and just as bad, OpenStep will not copy files into DOS partitions whose filenames contain ANY uppercase characters. This brain dead behavior is inexcusable. NeXT had a decent DOS file system handling capability and OpenStep broke it totally. WHat is NeXT's (now Apple Enterprise Software) problem? Must be too many Friday afternoon beer bashes. - Hacker -- Dark Hacker | e-mail hacker@computation.com Fortress Of Computation | Web http://www.computation.com/pub/hacker/ ____________________________________________________________________ "Life would be so much easier if only we had the source code."
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody Graphics Questions Date: 24 Apr 1998 16:37:50 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <6hqf4u$rlu$1@pump1.york.ac.uk> References: <zegelin-2404981257030001@dialup35.canb.ispsys.net> <6hp0q0$b59$1@news.digifix.com <6hpqta$qlb$2@ns3.vrx.net> On 24 Apr 1998 10:52:26 GMT, Maury Markowitz <maury@remove_this.istar.ca> wrote: > In <6hq0e0$ek9$1@pump1.york.ac.uk> Roger Peppe claimed: > > but you do need to learn some postscript if you're wanting > > to draw lines (as he was suggesting...) because AFAIK there > > are no AppKit classes supporting basic drawing. > > No, but there are basic _functions_ to do this. Lots of them. however, given that the functions directly map to postscript operators, it's a _really_ good idea to know some postscript before using them. (otherwise you have no way of knowing what they actually do, other than guessing, which isn't always the best way to write reliable software!) rog.
From: gmgraves@slip.net (George Graves) Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.programmer.codewarrior Subject: Re: mac out in the cold? Date: Fri, 24 Apr 1998 10:10:24 -0700 Organization: Graves Associates Message-ID: <gmgraves-2404981010240001@sf-usr1-30-158.dialup.slip.net> References: <6hnpj3$spq$4@newsfep3.sprintmail.com> In article <6hnpj3$spq$4@newsfep3.sprintmail.com>, macghod@concentric.net wrote: > OPENSTEP. The best thing since sliced bread. With just some tweaking with > interface builder (Note, I am not a next programmer, this is just what I have > heard) your app will not only run under RHapsody, but also windows 95(!!!), > OPENSTEP, (and maybe winnt? Or just OPENSTEP for winnt?). BUT!, isnt their > a operating system missing? Yes their is, macos! Openstep apps dont run on > the mac! (Good lord, I just know someone will email me and say, duh! havent > you heard of the blue box?) Good lord, you haven't heard about Blue Box! Because Blue Box lets Rhapsody run MAC APPS, it doesn't let the Mac run RHAPSODY APPS. There will be some runtime files available this summer which will allow Allegro (Mac OS 8.2) to run Rhapsody apps, however. > > People often mention, why would windows developers want to continue like they > are, when with openstep, they can put out apps that run on windows, openstep, > and rhapsody! Well, why will mac developers want to program for the mac, > when to program for the mac they have to do a seperate app? They won't. Rhapsody programs will be able to run on Mac OS. Program on > rhapsody only, and you get windows, OPENSTEP, rhapsody. If you want your > software to run on people macs who are using os 8.1 and lower (will 8.2 run > rhapsody apps?) you have to do a whole other application! That's the whole "catch-22" of shifting the Mac from Mac-OS to Rhapsody isn't it? How do you shift that paradigm without losing both developers and users of older OS versions. Blue Box makes Rhapsody backward compatible, but what will make earlier versions of Mac OS forward compatible with Rhapsody? Stay tuned to this station for the next heart rending chapter of our soap opera "Rhapsody in Blue". George Graves
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.next.programmer Subject: Re: Am I doing something wrong?? Date: 24 Apr 1998 17:05:26 GMT Organization: MiscKit Development Message-ID: <6hqgom$i6t$1@news.xmission.com> References: <6hoc5q$86f$1@newsfep2.sprintmail.com> <6hpd0f$h0l$1@newsfep1.sprintmail.com> macghod@concentric.net wrote: > With several programs, I have been unable to run them because I have been > unable to save the prefs! I had the resolution at 640 or 800, and the set > buttons were bellow the bottom of the screen, and their was no scroll bar, > thus no way to save the preferences!! This happened in pppmonitor (I didnt > have a card that supported over 640, and then even in radical news (I set it > to 32 bit color at 800/600) > > Now I have to ask if I am doing something wrong, cause that could be the > case. No, the panels were designed for larger screens are therefore you're SOL with those programs. (Well, not really. See below for possible workarounds.) > It would seem hard to believe that programmers would do this on > purpose? Some apps were written when NEXTSTEP ran on screens that were no smaller than 1120x832 (black hardware). At that time, there was no inkling that NEXTSTEP would ever run on a smaller screen than that. So windows were made that took advantage of the additional screen real estate. If the program wasn't subsequently adjusted when Intel versions of NEXTSTEP were released, then you'll see this problem. Some programmers also probably decided that unless you set you resolution to a certain screen size, they didn't want to bother with you. So there will be some who consciously chose to exclude you, for whatever internal reasons they had. (I'm NOT condoning this.) I've always tried to make sure my programs will work OK, even down to 640 by 480. I even stuck a little resolution ruler on my black monitor to show me how much real estate the various screen sizes would have; that let me design panels that would fit into limited screen layouts. Not everyone cares enough to do that, especially not for freeware. At any rate, if it really bothers you, just fire up InterfaceBuilder and move things around the panel to make it fit on your screen...Or use an app like Virtscreen to give you a larger virtual screen to pan around in (providing access to insanely big windows). -- Later, -Don Yacktman don@misckit.com <a href="http://www.misckit.com/don.html">My home page</a>
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.programmer.codewarrior Subject: Re: mac out in the cold? Date: 24 Apr 1998 10:50:41 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6hpqq1$qlb$1@ns3.vrx.net> References: <6hnpj3$spq$4@newsfep3.sprintmail.com> <6hni4v$453$1@ns3.vrx.net> <6hod6f$86f$3@newsfep2.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: macghod@concentric.net In <6hod6f$86f$3@newsfep2.sprintmail.com> macghod@concentric.net claimed: > You clipped out the part about os 8. So you are saying openstep apps will > run on a g3 running os 8.1? What does 8.1 have to do with anything? What if it's 8.2 that you need? Maury
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody Graphics Questions Date: 24 Apr 1998 10:52:26 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6hpqta$qlb$2@ns3.vrx.net> References: <zegelin-2404981257030001@dialup35.canb.ispsys.net> <6hp0q0$b59$1@news.digifix.com> <6hq0e0$ek9$1@pump1.york.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: rog@ohm.york.ac.uk In <6hq0e0$ek9$1@pump1.york.ac.uk> Roger Peppe claimed: > but you do need to learn some postscript if you're wanting > to draw lines (as he was suggesting...) because AFAIK there > are no AppKit classes supporting basic drawing. No, but there are basic _functions_ to do this. Lots of them. Maury
From: jim@ergotech.com Newsgroups: comp.sys.next.programmer Subject: Re: How can you store nibs in CVS on NT? Date: Fri, 24 Apr 1998 12:11:59 -0600 Organization: Deja News - The Leader in Internet Discussion Message-ID: <6hqh4v$mbm$1@nnrp1.dejanews.com> References: <6hhclf$77q@saturn.genoa.com> <353F4F4D.15A1@ibp.de> > > Alex Blakemore wrote: > > > > Has anyone successfully used CVS to store nibs on NT? If anyone has binaries of RCS and diff that understands binary files on NT I'd like to get them. I searched for some time but all the versions compiled on NT seem to be confused by CR/LF issues, that is, a binary comparison would convert incomming LF to CR/LF combination and the compare would fail on identical files. I spent some time trying to build such a distribution, but other projects intervened and I have never had a chance to try to complete it. If I remember correctly, the source to both RCS and diff recognise the problem and provide a compile time flag to fix it. CVS also claims that it will strip CR/LF on putting text (source) files in the archive, this has not been my experience with 1.9, although it appears, from other peoples experiences that 1.6 correctly handled this problem. Jim -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: Timothy J Luoma <nospam+yes-this-is-a-valid-address@luomat.peak.org> Newsgroups: comp.sys.next.programmer Subject: Re: Am I doing something wrong?? Date: Fri, 24 Apr 1998 13:23:28 -0400 Organization: @Home Network Message-ID: <Pine.NXT.3.96.980424131855.23576A-100000@luomat> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII In-Reply-To: <6hqgom$i6t$1@news.xmission.com> On 24 Apr 1998, Don Yacktman wrote: > Or use an app like Virtscreen to give you a larger virtual screen to pan > around in (providing access to insanely big windows). Minor point: I think Don unintentionally conflated VirtSpace and WideScreen VirtSpace was a nice little program that was bought and turned commercial.... I can never remember the name of the company which bought it. WideScreen is a freeware app which can be found here: ftp://next-ftp.peak.org/pub/next/apps/utils/workspace/WideScreen.0.5.README ftp://next-ftp.peak.org/pub/next/apps/utils/workspace/WideScreen.0.5.NIHS.bs.tar .gz ftp://next-ftp.peak.org/pub/next/apps/utils/workspace/WideScreen.0.5C.NIHS.bs.ta r.gz (the 0.5C version has some color support added, but I believe it is the same in all other respects) TjL
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: duplicating NSInvocations Date: 24 Apr 1998 14:01:27 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6hqk1n$ngh$1@crib.bevc.blacksburg.va.us> I'm trying to copy the arguments of one NSInvocation into another, by iterating through each argument and using the result of -getArgument:atIndex: on the original invocation to set each argument of the new one with -setArgument:atIndex:. However, -getArgument:atIndex: needs to be passed an allocated buffer to fill with the argument, and I don't know how much space to allocate because I don't know the size of the argument. The NSInvocation documentation mentions an -argumentInfoAtIndex: method of NSMethodSignature which returns an NSArgumentInfo structure, but that method doesn't actually exist! So, what do I do? A hack would be just to allocate a "sufficiently big" buffer, which probably isn't too bad since arguments don't tend to be too big (however, structures can be arbitrarily large), but I was hoping for a cleaner solution. (Actually, come to think of it, I could just malloc the frame length and that would be sufficient.) The release notes say that NSArgumentInfo used to be in both NEXTSTEP and OpenStep but was removed from both. (Why?? I need it!) Is there a nice workaround? Can I obtain the argument size somehow from the argument type, using NSMethodSignature's -getArgumentTypeAtIndex:?
From: AUCTION <auction.009@interserv.com> Newsgroups: comp.sys.next.programmer Subject: Adobe Framemaker +SGML (Only $395) Date: Friday, 24 Apr 1998 14:18:20 -0600 Organization: CDE7 Message-ID: <24049814.1820@interserv.com> Adobe Framemaker +SGML (Only $395) =================================== FULL VERSION (5.1 for WIN 95 / NT) http://www.adobe.com/prodindex/framemaker/prodinfosgml.html ONE COPY LEFT ! Brand New....Still in Shrinkwrap ! It Retails for $1000+ Yours for Only $395 !!! =========== WE ACCEPT: =========== 1) VISA 2) MASTERCARD 3) AMERICAN EXPRESS 4) DISCOVER / NOVUS 5) WIRE TRANSFER 6) CASHIER CHECK 7) MONEY ORDER CALL TOLL FREE ================= 1-888-300-5069 *zhETUR
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <24049814.1820@interserv.com> Control: cancel <24049814.1820@interserv.com> Date: 24 Apr 1998 20:04:53 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.24049814.1820@interserv.com> Sender: AUCTION <auction.009@interserv.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: John Jensen <jjens@primenet.com> Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.programmer.codewarrior Subject: Re: mac out in the cold? Date: 24 Apr 1998 13:37:00 -0700 Organization: Primenet (602)416-7000 Message-ID: <6hqt5c$m2h@nntp02.primenet.com> References: <6hnpj3$spq$4@newsfep3.sprintmail.com> <forrestDELETE-THIS!-2304982216240001@term3-19.vta.west.net> Forrest Cameranesi <forrestDELETE-THIS!@west.net> wrote: : Apple announced back in Dec '96 that there will be a Yellow Box for MacOS, : just there already is for Win95/NT (which was much easier, because they : ... Yellow Box for MacOS is very sensible, but my recollection is that the component was strangely missing from early roadmaps ... and that it showed up in mid '97. John
From: jklein@ivy.hampshire.edu (jonathan klein) Newsgroups: comp.sys.next.programmer Subject: Adding window server handler to activate application Date: 24 Apr 98 19:06:31 GMT Message-ID: <3540e2b7.0@192.33.12.30> I'm trying to figure out how to cause a certain keycode to activate a certain application universally, like LaunchBar does with command-space. Both Timothy and Frank (thanks to you both) pointed out that LaunchBar does this by connecting to the PS window server and executing the file CommandSpace.ps in the LaunchBar app wrapper. Looking at this file, it's pretty clear how it works -- the important line being: globaldict /LaunchBarContext get windowPackage1.0 /activateContext get exec My question is: how do I add a named context, such as LaunchBarContext for my application? I assume that I would want to add such a context with a certain callback procedure, and have that callback procedure activate my app... is this correct? -- -jon klein, jklein@ivy.hampshire.edu NeXTmail welcome
From: sadahiro@cs.utexas.edu Newsgroups: comp.sys.next.hardware,comp.sys.next.programmer,comp.sys.next.sysadmin,comp.sys.next.software Subject: seagate barracuda(st15150wc) anyone?T Date: Fri, 24 Apr 1998 20:16:43 -0600 Organization: Deja News - The Leader in Internet Discussion Message-ID: <6hrdhr$rqt$1@nnrp1.dejanews.com> Hi, does anyone use sseagate barracuda(st15150wc)? I knwo it gets hot, but I believe Cubes cooling is sufficient for it. It is wide drive(68pin), but also converted to 50pins. I do not know what it means. Does it mean it exactly behaves as 50pin connector? Or should I still expect not working correctly? Here is spec I got. Part Number ST15150WC Formatted Capacity 4294 MB External Transfer Rate 20 MB/s Sync., 6.9MB/s sustained Fast Wide SCSI-2 Rotational Speed 7200 rpm Average Access 8.0 ms - This is very fast! Just compare to other drives. It doesn't get much better than this. Average Latency 4.17 ms - How fast it seeks on the same cylinder. A direct function of the Rotational Speed of 7,200rpm. Very important number. Buffer 1024 kB - Yes, that's 1 MegaByte! Dimensions 4.0 x 1.63 x 5.97 in (3 1/2 in form factor, half height) SCSI Connector 80-pin SCA 50-pin with adapter, add $24* 68-pin with adapter, add $49* read more about SCA adapters *Includes special savings offer Does this mean that I can use EXACTLY as 50pin scsi-2 HD with 50pin adaptor? If anyone out there uses this drive, please let me know. Thanks a lot! Makoto -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: macghod@concentric.net Newsgroups: comp.sys.next.programmer Subject: newby programmer has simple question Date: 25 Apr 1998 01:43:59 GMT Message-ID: <6hrf4v$3vc$1@newsfep3.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Ok, I have only done a little bit of programming, in c++, and have done c++ on unix for classes. I wanted to compile bytemark. I have user 4.2 and developer installed. It looks like the header files are not being seen? Why does it not find stdio.h and these other header files when they are standard c header files? myhost:5# make nbench0 cc nbench0.c -o nbench0 nbench0.c:32: header file 'stdio.h' not found nbench0.c:33: header file 'stdlib.h' not found nbench0.c:34: header file 'ctype.h' not found nbench0.c:35: header file 'string.h' not found nbench0.c:36: header file 'time.h' not found nbench0.c:37: header file 'math.h' not found nbench0.h:170: undefined type, found `FILE' nbench0.h:237: header file 'windows.h' not found nbench0.h:238: header file 'toolhelp.h' not found nbench0.h:239: undefined type, found `TIMERINFO' nbench0.h:240: undefined type, found `HANDLE' nbench0.h:241: undefined type, found `FARPROC' nbench0.h:249: undefined type, found `FILE' nbench0.c:51: undefined type, found `time_t' nbench0.c:77: illegal statement, missing `;' after `)' nbench0.c:254: undefined type, found `FILE' nbench0.c:280: illegal expression, found `)' nbench0.c:280: illegal expression, found `)' nbench0.c:290: illegal external declaration, found `return' nbench0.c:314: undefined type, found `FILE' nbench0.c:384: illegal expression, found `)' nbench0.c:384: illegal expression, found `)' nbench0.c:579: illegal external declaration, found `return' *** Exit 1 Stop.
From: nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) Newsgroups: comp.sys.next.programmer Subject: Re: newby programmer has simple question Date: 25 Apr 1998 01:51:01 GMT Organization: none Message-ID: <6hrfi5$rli$5@ha2.rdc1.nj.home.com> References: <6hrf4v$3vc$1@newsfep3.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: macghod@concentric.net In <6hrf4v$3vc$1@newsfep3.sprintmail.com> macghod@concentric.net wrote: > Ok, I have only done a little bit of programming, in c++, and have done c++ > on unix for classes. I wanted to compile bytemark. I have user 4.2 and > developer installed. It looks like the header files are not being seen? Why > does it not find stdio.h and these other header files when they are standard > c header files? Do you have the Developer CD? The header files are not included with the regular User CD. TjL -- [Yes that is a valid mail address, until it gets spammed] Unofficial @Home FAQ: http://members.home.com/faqs/ Who decided we _wanted_ a Tarzan for the 90s? Or a new Love Boat, for that matter...
From: macghod@concentric.net Newsgroups: comp.sys.next.programmer Subject: Re: newby programmer has simple question Date: 25 Apr 1998 03:42:50 GMT Message-ID: <6hrm3q$gd5$1@newsfep4.sprintmail.com> References: <6hrf4v$3vc$1@newsfep3.sprintmail.com> <6hrfi5$rli$5@ha2.rdc1.nj.home.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: nospam+yes-this-is-a-valid_address@luomat.peak.org In <6hrfi5$rli$5@ha2.rdc1.nj.home.com> Timothy Luoma wrote: > In <6hrf4v$3vc$1@newsfep3.sprintmail.com> macghod@concentric.net wrote: > > Ok, I have only done a little bit of programming, in c++, and have done c++ > > on unix for classes. I wanted to compile bytemark. I have user 4.2 and > > developer installed. It looks like the header files are not being seen? > Why > > does it not find stdio.h and these other header files when they are > standard > > c header files? > > Do you have the Developer CD? Oh yes, I already installed the cd (OPENSTEP DEVELOPER 4.2) -- running on a PC with OPENSTEP, the year 2000 bug-free operating system Email address is invalid, please email macghod@concentric.net
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <353a6046.0@news.codenet.net> Control: cancel <353a6046.0@news.codenet.net> Date: 24 Apr 1998 10:20:54 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.353a6046.0@news.codenet.net> Sender: appology for ADULTS ONLY Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: "Zico" <ZicoKnows@hotmail.com> Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.programmer.codewarrior Subject: Re: mac out in the cold? Date: Sat, 25 Apr 1998 03:33:49 -0400 Organization: All USENET -- http://www.Supernews.com Message-ID: <6hqf9g$s1a$1@supernews.com> References: <6hnpj3$spq$4@newsfep3.sprintmail.com> <forrestDELETE-THIS!-2304982216240001@term3-19.vta.west.net> x-no-archive: yes Forrest Cameranesi wrote in message ... >[With Rhapsody, y]ou've got legacy and >future apps seamlessly blended, a decent installed base, and full buzzword >compliance, and you could probably sell it for Intel machines (minus blue >box) fairly well too. What more could you ask for? The ability to use both CPUs in my Intel box, like Linux, Windows NT, and BeOS. Z
From: a@b.c Newsgroups: comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.nsc,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.pen Subject: $$$ FREE CASH! $$$ Date: 25 Apr 1998 12:46:22 GMT Organization: FREE $$$ Message-ID: <6hsluu$fgd$552@news.on> $$$ Do YOU need extra CA$H?? $$$ Turn $6 into $60,000 in just 5 weeks!!! This is simple, safe, and it really works! For $6 (U.S), 6 stamps, and about an hour of your time, you could earn a year's salary in a month. Sounds to good to be true, but just imagine. WHAT IF? A little while back, I was browsing these newsgroups, just like you are now, and came across an article similar to this that said you could make thousands of dollars within weeks with only an initial investment of $6.00! So I thought, "Yeah, right, this must be a scam!" but like most of us, I was curious. Like most of us, I kept reading. Anyway, it said that if you send $1.00 to each of the 6 names and addresses stated in the article, you could make thousands in a very short period of time. You then place your own name and address at the bottom of the list at #6, and post the article to at least 200 newsgroups. (There are about 22,000.) or e-mail them to friends, or e-mailing lists... No catch, that was it. Even though the investment was a measly $6, I had three questions that needed to be answered before I could get involved in this sort of thing. 1. IS THIS REALLY LEGAL? I called a lawyer first. The lawyer was a little skeptical that I would actually make any money but he said it WAS LEGAL if I wanted to try it. I told him it sounded a lot like a chain letter but the details of the system (SEE BELOW) actually made it a legitimate legal business. 2. Would the Post Office be ok with this...? I called them: 1-800-725-2161 and they confirmed THIS IS ABSOLUTELY LEGAL! (See Title 18,h sections 1302 NS 1341 of Postal Lottery Laws). This clarifies the program of collecting names and addresses for a mailing list. 3. Is this moral? Well, everyone who sends me a buck has a good chance of getting A LOT of money ... a much better chance than buying a lottery ticket! So, having these questions answered, I invested EXACTLY $7.92 ... six $1.00 bills and six 32 cent postage stamps ... and boy am I glad I did! Within 7 days, I started getting money in the mail! I was shocked! I still figured it would end soon, and didn't give it another thought. But the money just continued coming in. In my first week, I made about $20.00 to $30.00 dollars. By the end of the second week I had a mad total of $1,000.00! In the third week I had over $10,000.00 and it was still growing. This is now my fourth week and I have made a total of just over $42,000.00 and it's still coming in..... It's certainly worth $6.00 and 6 stamps! So now I'm reposting this so I can make even more money! The *ONLY* thing stopping *ANYONE* from enriching their own bank account is pure laziness! It took me all of 5 MINUTES to print this out, follow the directions, and begin posting to newsgroups. It took me a mere 45 minutes to post to over 200 newsgroups. And for this GRAND TOTAL investment of $ 7.92 (US) and under ONE HOUR of my time, I have reaped an incredible amount of money -- like nothing I've ever even heard of anywhere before! 'Nuff said! Let me tell you how this works, and most importantly, why it works. Also, make sure you print a copy of this article now, so you can get the information off of it when you need it. The process is very simple and consists of THREE easy steps. ============ HOW IT WORKS ============ STEP 1: ------ Get 6 separate pieces of paper and write the following on each piece of paper: PLEASE ADD ME TO YOUR MAILING LIST. $1 US DOLLAR PROCESSING FEE IS ENCLOSED. (THIS IS KEY AS THIS IS WHAT MAKES IT LEGAL SINCE YOU ARE PAYING FOR AND LATER OFFERING A SERVICE). Now get 6 $1.00 bills and place ONE inside EACH of the 6 pieces of paper so the bill will not be seen through the envelope to prevent theft/robbery. Then, place one paper in each of the 6 envelopes and seal them. You should now have 6 sealed envelopes, each with a piece of paper stating the above phrase and an U.S. $1.00 bill. Mail the 6 envelopes to the following addresses: #1- R Sauter 3500 Crescent Ct Flower Mound, TX 75028 USA #2- Catherine Watkins P.O. Box 67 Chapel Hill, NC 27514-0067 USA #3- Brett Brandon 807 Brooks Av. Madison, Tn 37115 USA #4- B Jones 0504 E. 600 N. LaPorte, IN 46350 USA #5- tepinga trust 15 Victor Street Avondale, Auckland 1007 New Zealand #6- Mike Ericson P.O. Box 1211 Burnsville, MN 55337 USA STEP 2: Now take the #1 name off the list that you see above, move the other names up (6 becomes 5, 5 becomes 4, etc.) and add YOUR Name as number 6 on the list. (If you want to remain anonymous put a nickname, but the address MUST be correct. It, of course, MUST contain your country, state/district/area, zip code, etc! You wouldn't want your money to fly away, wouldn't you?). STEP 3: Now post your amended article to at least 200 newsgroups. Remember that 200 postings are just a guideline. The more you post, the more money you make! Don't know HOW to post in the news groups? Well do exactly the following: ------------------------------------------------------------------------ HOW TO POST TO NEWSGROUPS FAST WITH YOUR WEB BROWSER: The fastest way to post a newsletter: Highlight and COPY (Ctrl-C) the text of this posted message and PASTE (Ctrl-V) it into a plain text editor (as Wordpad) and save it. After you have made the necessary changes that are stated above, simply COPY (Ctrl-C) and PASTE (Ctrl-V) the text into the message composition window, after selecting a newsgroup, and post it! (Or you can attach the file, without writing anything to the message window.) ------------------------ ------------------------------------------------------------------------ If you have Netscape Navigator 3.0 do the following: 1. Click on any newsgroup like normal, then click on 'TO NEWS'. This will bring up a box to type a message in. 2. Leave the newsgroup box like it is, change the subject box to something flashy, something to catch the eye, as "$$$ NEED CASH $$$? READ HERE! $! $! $" Or "$$$! MAKE FAST CASH, YOU CAN'T LOSE! $$$". Or you can use my subject title. 3. Now click on 'ATTACHMENTS'. Then click on 'ATTACH FILE'. Find your file on your Hard Disk (the one you saved from the text editor). Once you find it, click on it and then click 'OPEN' and 'OK'. You should now see your file name in the attachments box. 4. Now click on 'SEND'/'POST'. You see? Now you just have 199 to go! (Don't worry, it's easy and quick once you get used to it.) NOTE: All the versions of Netscape Navigator's are similar to each other, so you'll have no problem to do this if you don't have Netscape Navigator 3.0. ------------------------------------------------------------------------ ! QUICK TIP! (For Netscape Navigator 3.x and above) You can post this message to many newsgroups at a time, by simply selecting a newsgroup near the top of the screen, hold down the SHIFT, and then select a newsgroup near the bottom of the screen. All of the newsgroups in/between will be selected. After that, you follow/do the basic steps, stated below at this letter, except of step #1. You can go to the page stated below in this letter and click on a newsgroup to open up the newsgroups window. Once you've done this, in the same window go to 'OPTIONS', and then mark 'SHOW ALL NEWSGROUPS' and 'SHOW ALL MESSAGES'. Now you can see all the newsgroups and you can apply easier the above tip. ------------------------------------------------------------------------ If you have MS Internet Explorer do the following: 1. Go to the newsgroups and press 'POST AN ARTICLE'. To the new window type your headline in the subject area and then click in the large window below. There either PASTE your letter (which it's been copied from the text editor), or attach the file which contains it. 2. Then click on 'SEND' or 'OK'. NOTE: All versions of MS Internet Explorer are similar to each other, so you won't have any problem doing this. GENERAL NOTES ON POSTING: A nice page where you'll find all the newsgroups if you want help is http://www.liszt.com/ (When you go to the home page, click on the link 'Newsgroup Directory'). But I don't think you'll have any problem posting because it's very easy once you've found the newsgroups. All these web browsers are similar. It doesn't matter which one you have. (But it makes it very easy if you have Netscape Navigator 3.0 or later. You may download it from the Internet if you don't have it.) You just have to remember the basic steps, stated below. BASIC STEPS FOR POSTING: 1. Find a newsgroup and you click on it. 2. You click on 'POST AN/NEW ARTICLE' or 'TO NEWS' or anything else similar to these. 3. You type your flashy headline in the subject box. 4. Now, either you attach the file containing your amended letter, or you PASTE the letter. (You have to COPY it from the text editor, of course, from before.) 5. Finally, you click on 'SEND' or 'POST' or 'OK', whatever is there. ------------------------------------------------------------------------ **REMEMBER, THE MORE NEWSGROUPS YOU POST IN, THE MORE MONEY YOU WILL MAKE! BUT YOU HAVE TO POST A MINIMUM OF 200** That's it! You will begin receiving money from around the world within days! You may eventually want to rent a P.O.Box due to the large amount of mail you receive. If you wish to stay anonymous, you can invent a name to use, as long as the postman will deliver it. **JUST MAKE SURE ALL THE ADDRESSES ARE CORRECT. ** ------------------------------------------------------------------------- ================= Now the WHY part: ================= Out of 200 postings; say I receive only 5 replies (a very low example). So then I made $5.00 with my name at #6 on the letter. Now, each of the 5 persons who just sent me $1.00 make the MINIMUM 200 postings, each with my name at #5 and only 5 persons respond to each of the original 5, that is another $25.00 for me, now those 25 each make 200 MINIMUM posts with my name at #4 and only 5 replies each, I will bring in an additional $125.00! Now, those 125 persons turn around and post the MINIMUM 200 with my name at #3 and only receive 5 replies each, I will make an additional $626.00! OK, now here is the fun part, each of those 625 persons post a MINIMUM 200 letters with my name at #2 and they each only receive 5 replies, that just made me $3,125.00! Those 3,125 persons will all deliver this message to 200 newsgroups with my name at #1 and if still 5 persons per 200 newsgroups react I will receive $15,625,00! With an original investment of only $6.00! AMAZING! And as I said 5 responses is actually VERY LOW! Average are probable 20 to 30! So lets put those figures at just 15 responses per person. Here is what you will make: at #6 $15.00 at #5 $225.00 at #4 $3,375.00 at #3 $50,625.00 at #2 $759,375.00 at #1 $11,390,625.00 When your name is no longer on the list, you just take the latest posting in the newsgroups, and send out another $6.00 to names on the list, putting your name at number 6 again. And start posting again. The thing to remember is, do you realize that thousands of people all over the world are joining the internet and reading these articles everyday, JUST LIKE YOU are now! So can you afford $6.00 and see if it really works? I think so... People have said, "what if the plan is played out and no one sends you the money? So what! What are the chances of that happening when there are tons of new honest users and new honest people who are joining the internet and newsgroups everyday and are willing to give it a try? Estimates are at 20,000 to 50,000 new users, every day, with thousands of those joining the actual Internet. Remember, play FAIRLY and HONESTLY and this will work. You just have to be honest. By the way, if you try to deceive people by posting the messages with your name in the list and not sending the money to the rest of the people already on the list, you will NOT get as much. Someone I talked to knew someone who did that and he only made about $150.00, and that's after seven or eight weeks! Then he sent the 6 $1.00 bills, people added him to their lists, and in 4-5 weeks he had over $10k. This is the fairest and most honest way I have ever seen to share the wealth of the world without costing anything but our time! You also may want to buy mailing and e-mail lists for future dollars. Make sure you print this article out RIGHT NOW, also. Try to keep a list of everyone that sends you money and always keep an eye on the newsgroups to make sure everyone is playing fairly. Remember that HONESTY IS THE BEST POLICY. You don't need to cheat the basic idea to make the money! GOOD LUCK to all and please play fairly and reap the huge rewards from this, which is tons of extra CASH. Please remember to declare your extra income. Thanks once again... =========================================================== ========== LEGAL? ? ? (Comments from Bob Novak who started this new version.) "People have asked me if this is really legal. Well, it is! You are using the Internet to advertise you business. What is that business? You are assembling a mailing list of people who are interested in home based computer and online business and methods of generating income at home. Remember that people send you a small fee to be added to your mailing list. It is legal. What will you do with your list of thousands of names? Compile all of them into a database and sell them as "Mailing Lists" on the internet in a similar manner, if you wish, and make more money. How do you think you get all the junk mail that you do? Credit card companies, mail order, Utilities, anyone you deal with through the mail can sell your name and address on a mailing list, unless you ask them not to, in addition to there regular business, So, why not do the same with the list you collect. You can find more info about "Mailing Lists" on the internet using any search engine. ." So, build your mailing list, keep good accounts, declare the income and pay your taxes. By doing this you prove your business intentions. Keep an eye on the newsgroups and when the cash has stopped coming (that means your name is no longer on the list), you just take the latest posting at the newsgroups, send another $6.00 to the names stated on the list, make your corrections (put your name at #6) and start posting again. =========================================================== NOTES: *1. In some countries, the export of the country's exchange is illegal. But you can get the license to do this from the post office, explaining the above statements (that you have an online business, etc. You may have to pay an extra tax, but that's OK, the amount of the incoming money is HUGE! And as I said, a few countries have that restriction. *2. You may want to buy mailing and e-mail lists for future dollars. (Or Database or Spreadsheet software.) *3. If you're really not sure or still think this can't be for real, please print a copy of this article and pass it along to someone who really needs the money, and see what happens. *4. You should start getting responses within 1-2 weeks.
From: zegelin@actonline.com.au (Peter) Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody Graphics Questions Date: 25 Apr 1998 11:45:06 GMT Organization: CyberOne Pty Ltd Message-ID: <zegelin-2504982145560001@dialup95.canb.ispsys.net> References: <zegelin-2404981257030001@dialup35.canb.ispsys.net> <6hp0q0$b59$1@news.digifix.com> Cache-Post-Path: genie.cyberone.com.au!unknown@dialup95.canb.ispsys.net In article <6hp0q0$b59$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: > >On 04/23/98, Peter wrote: [snip] > >Regions: > >€ Are regions and their traps (DiffRgn,UnionRgn etc and especially > >BitMapToRgn) fully supported? Is their usage much the same as on > MacOs? A > >pointer to info on regions & Rhaposdy would be great! > > > > There are no regions on Rhapsody.. You have define paths, > which you can then hit-test... > Hmm. No regions as such in Rhapsody, eh. My (unfinished) MacOs App uses regions all over the place. I really will have to Think Different :-) OK, Ive outlined a few places where I find regions useful and I'd be interested to know roughly how I'd handle the same things using Rhapsody. I think I know the answer (always use offscreen drawing?) but I'll ask just in case I've missed something fundamental. 1. Simple graphic object layering. I keep an ordered list of objects and as each object is drawn (near objects first) its region is subtracted from the ClipRgn which means that any object drawn later is automatically clipped by any object drawn earlier. 2. Dragging a selection Create a Bitmap the size of the selection. Draw the selected objects into the Bitmap. Make a region from this using BitMapToRgn() then call DragGreyRgn() to let the system drag my selection around while the mouse is down. 3. Selection Rectangles Each time through the mouse tracking loop: Create 2 regions based on the current selection rectangle. Inset one of the regions by 1 pixel then subtract this smaller region from the the other to create a region consisting of a one pixel thick outline of the current selection rectangle. Do the same for the previous selection rectangle (from the previous iteration of the loop). Get region which includes only those pixels that are NOT in both regions and paint this new region in gray in Xor mode so pixels are flipped. Pretty basic stuff I know but.. Thanks again for any help. Peter
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.nsc,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.pen Subject: cmsg cancel <6hsluu$fgd$552@news.on> Control: cancel <6hsluu$fgd$552@news.on> Date: 25 Apr 1998 13:35:37 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6hsluu$fgd$552@news.on> Sender: a@b.c Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: "Michelle L. Buck" <buck.erik@mcleod.net> Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody Graphics Questions Date: Sat, 25 Apr 1998 09:53:30 -0500 Organization: McleodUSA - http://www.mcleodusa.net Message-ID: <6hsthi$h5g1@odie.mcleod.net> References: <zegelin-2404981257030001@dialup35.canb.ispsys.net> <6hp0q0$b59$1@news.digifix.com> <zegelin-2504982145560001@dialup95.canb.ispsys.net> Peter wrote in message ... >In article <6hp0q0$b59$1@news.digifix.com>, sanguish@digifix.com (Scott >Anguish) wrote: > >> >On 04/23/98, Peter wrote: > >[snip] > > Hmm. No regions as such in Rhapsody, eh. My (unfinished) MacOs App >uses regions all over the place. I really will have to Think Different :-) > > OK, Ive outlined a few places where I find regions useful and I'd be >interested to know roughly how I'd handle the same things using Rhapsody. >I think I know the answer (always use offscreen drawing?) but I'll ask >just in case I've missed something fundamental. > >1. Simple graphic object layering. > I keep an ordered list of objects and as each object is drawn (near >objects first) its region is subtracted from the ClipRgn which means that >any object drawn later is automatically clipped by any object drawn >earlier. > Layering is accomplished by the order of drawing. Using an ordered list of objects and drawing them in order will produce the correct picture. A complex example of this is available in the Draw.app. >2. Dragging a selection > Create a Bitmap the size of the selection. Draw the selected objects >into the Bitmap. Make a region from this using BitMapToRgn() then call >DragGreyRgn() to let the system drag my selection around while the mouse >is down. This is very similar in Rhapsody. Draw the selected objects into an NSImage, return the image from on of the dragging protocol methods, and the system will do the rest for you. > >3. Selection Rectangles > Each time through the mouse tracking loop: > > Create 2 regions based on the current selection rectangle. Inset one of >the regions by 1 pixel then subtract this smaller region from the the other >to create a region consisting of a one pixel thick outline of the current >selection rectangle. > Do the same for the previous selection rectangle (from the previous >iteration of the loop). > Get region which includes only those pixels that are NOT in both regions >and paint this new region in gray in Xor mode so pixels are flipped. > Wow, that sounds like the hard way! In Rhapsody, just use instance drawing (special temporary drawing mode) to draw a rectangle arround the selection. Alternately, composit a transparent rectangle over the selection (Buttons etc do this). Functions for insetting rectangles, rectangle intersections and unions, and clipping are all available in Rhapsody.
From: nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) Newsgroups: comp.sys.next.programmer Subject: Need help: run program post-mount Date: 25 Apr 1998 15:35:55 GMT Organization: none Message-ID: <6hsvsr$dt$1@ha2.rdc1.nj.home.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit The FSSoundPanel lets someone associate a snd file with certain Workspace events (App launch, App quit, Disk mount/umount, and so forth. What I was looking for was a program which would run after a disk mount/umount (or eject, doesn't matter). Does anyone have such a program hidden away somewhere? TjL -- [Yes that is a valid mail address, until it gets spammed] Unofficial @Home FAQ: http://members.home.com/faqs/ Who decided we _wanted_ a Tarzan for the 90s? Or a new Love Boat, for that matter...
Message-ID: <35420B5A.B4B35F53@denalisites.com> From: Brian Wotring <bj@denalisites.com> Organization: Denali Sites MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: NSString Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Sat, 25 Apr 1998 16:05:09 GMT NNTP-Posting-Date: Sat, 25 Apr 1998 12:05:09 EST I'm having problems understanding how to use the [aString characterAtIndex:anIndex] method. I'm trying to take an NSString and put every character into an NSMutableArray of NSStrings. For example, if I had the string: "cat", the result would be an NSMutableArray as follows: 0 -> "c" 1 -> "a" 2 -> "t" Any help on how I might acheive this? -- | Brian Wotring | bj@denalisites.com ( NeXTMail welcome ) | http://tact.dyn.ml.org
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody Graphics Questions Date: 25 Apr 1998 14:04:47 GMT Organization: Idiom Communications Message-ID: <6hsqhv$k1r$5@news.idiom.com> References: <zegelin-2404981257030001@dialup35.canb.ispsys.net> <6hp0q0$b59$1@news.digifix.com> <zegelin-2504982145560001@dialup95.canb.ispsys.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: zegelin@actonline.com.au Peter may or may not have said: [snip] -> Hmm. No regions as such in Rhapsody, eh. My (unfinished) MacOs App -> uses regions all over the place. I really will have to Think Different :-) -> -> OK, Ive outlined a few places where I find regions useful and I'd be -> interested to know roughly how I'd handle the same things using Rhapsody. -> I think I know the answer (always use offscreen drawing?) but I'll ask -> just in case I've missed something fundamental. -> -> 1. Simple graphic object layering. -> I keep an ordered list of objects and as each object is drawn (near -> objects first) its region is subtracted from the ClipRgn which means that -> any object drawn later is automatically clipped by any object drawn -> earlier. Well, postscript does have clipping paths, which need not be contiguous. But, if all you're doing is using regions to keep objects drawn later from overlaying objects drawn earlier, then in postscript all you'd need to do is reverse the order that you're drawing them. -> 2. Dragging a selection -> Create a Bitmap the size of the selection. Draw the selected objects -> into the Bitmap. Make a region from this using BitMapToRgn() then call -> DragGreyRgn() to let the system drag my selection around while the mouse -> is down. In OpenStep, you'd create an NSImage object with a transparent background, draw the selected objects into the NSImage, and then call [NSView dragImage:...] -> 3. Selection Rectangles -> Each time through the mouse tracking loop: -> -> Create 2 regions based on the current selection rectangle. Inset one of -> the regions by 1 pixel then subtract this smaller region from the the other -> to create a region consisting of a one pixel thick outline of the current -> selection rectangle. -> Do the same for the previous selection rectangle (from the previous -> iteration of the loop). -> Get region which includes only those pixels that are NOT in both regions -> and paint this new region in gray in Xor mode so pixels are flipped. Sounds like a lot of work to show a selection rect. In OpenStep, I'd just implement a mouseDown: method in the view that has to show the selection rectangles, and add some code to the drawRect: method that shows the selection. Something like: @implementation myView - (void) mouseDown:(NSEvent *) theEvent { // note where the mouse went down // set a flag to indicate that a selection is happening } - (void) mouseDragged:(NSEvent *) theEvent { //Note where the mouse is now, for the second point defining the selection rect. } - (void) mouseUp:(NSEvent *) theEvent { // reset the flag that says a selection rect is being dragged } - drawRect:(NSRect) rect { // the "normal" drawing code if (dragging) { //draw the selection rect from where the mouse went down, to where the mouse is now. } } -jcr
From: tom@basil.icce.dev.rug.null.nl (Tom Hageman) Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody Graphics Questions Date: 25 Apr 1998 16:43:13 GMT Organization: Warty Wolfs Sender: news@basil.icce.dev.rug.null.nl (NEWS pusher) Message-ID: <Erxo66.528@basil.icce.dev.rug.null.nl> References: <zegelin-2404981257030001@dialup35.canb.ispsys.net> <6hp0q0$b59$1@news.digifix.com> <6hq0e0$ek9$1@pump1.york.ac.uk> rog@ohm.york.ac.uk (Roger Peppe) wrote: > On 24 Apr 1998 03:26:56 GMT, Scott Anguish <sanguish@digifix.com> wrote: > > You don't need to learn Postscript. As far as it being > > advisable... I'd learn it as I need to when I come across options that > > aren't covered by classes (most is covered by classes) > > but you do need to learn some postscript if you're wanting > to draw lines (as he was suggesting...) because AFAIK there > are no AppKit classes supporting basic drawing. Not in OPENSTEP, no. Rhapsody introduces an NSBezierPath class (which can be used to draw lines.) --Tom. <<SPAMBLOCK: remove dev. and null. to reply>>
Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody Graphics Questions From: miba@image.dk (Michael Balle) Date: Sat, 25 Apr 1998 18:56:27 +0200 Message-ID: <1d821q1.14xkw931yvv0i6N@[10.0.0.2]> References: <zegelin-2404981257030001@dialup35.canb.ispsys.net> <6hp0q0$b59$1@news.digifix.com> <zegelin-2504982145560001@dialup95.canb.ispsys.net> <6hsqhv$k1r$5@news.idiom.com> John C. Randolph <jcr.remove@this.phrase.idiom.com> wrote: > Peter may or may not have said: > [snip] > -> Hmm. No regions as such in Rhapsody, eh. My (unfinished) MacOs App > -> uses regions all over the place. I really will have to Think Different :-) > -> > -> OK, Ive outlined a few places where I find regions useful and I'd be > -> interested to know roughly how I'd handle the same things using Rhapsody. > -> I think I know the answer (always use offscreen drawing?) but I'll ask > -> just in case I've missed something fundamental. > -> > -> 1. Simple graphic object layering. > -> I keep an ordered list of objects and as each object is drawn > (near > -> objects first) its region is subtracted from the ClipRgn which means that > -> any object drawn later is automatically clipped by any object drawn > -> earlier. > > Well, postscript does have clipping paths, which need not be contiguous. But, > if all you're doing is using regions to keep objects drawn later from > overlaying objects drawn earlier, then in postscript all you'd need to do is > reverse the order that you're drawing them. > I have been wondering why this region issue hasn't been brought earlier. Is it possible, using the PostScript clipping paths, to do calculations on the paths, like DiffRgn, UnionRgn etc. One place where these operations might come in handy is when you are doing layout of text in a NSTextContainer. If you wan't to make the text flow around a circle, you need a way to calculate the intersection between the proposed text rectangle and the circle. You could probably do it by using some mathematic, but regions would be really nice in this case. Michael
From: agave@lucien.blight.com (Ian Cardenas) Newsgroups: comp.sys.next.programmer Subject: Rhapsody API re: HTML Date: 25 Apr 1998 18:10:22 GMT Organization: CNI nntpcache test server. Message-ID: <893527823.660752@point.cpoint.net> Cache-Post-Path: point.cpoint.net!agave@lucien.blight.com Hi, I've posted this question to Rhapsody-Dev as well but in the interest of reaching the widest audience I thought I'd try here as well. Note: All of the information about the Rhapsody APIs can be found at: http://gemma.apple.com/techpubs/rhapsody/NextLibrary/Documentation/NextDev/HomePages/HomePage.html I would like to know how to display an HTML page in an NSTextView with all the inline images. Once the HTML is in an NSData object you can create an NSTextStorage and call -initWithHTML:. Next, you would get your TextView's LayoutManager and replace the existing TextStorage with the newly created one. This takes care of everything but the images which brings me to my question: How do you get the images to be inlined? It would seem to me you would have to find all the TextAttachments and fetch the image data and then inform the view to update with the new information. I can't determine the process though from looking at the API. Any insight would be appreciated. Thanks, Ian P. Cardenas
From: John Hornkvist <sorry@no.more.spams> Newsgroups: comp.sys.next.programmer Subject: Re: Am I doing something wrong?? Date: Sat, 25 Apr 1998 21:07:28 GMT Organization: Chalmers Tekniska Högskola Sender: john@haddock.cd.chalmers.se (John Hprnkvist) Message-ID: <ErzMoH.KFG@haddock.cd.chalmers.se> References: <Pine.NXT.3.96.980424131855.23576A-100000@luomat> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: nospam+yes-this-is-a-valid-address@luomat.peak.org In <Pine.NXT.3.96.980424131855.23576A-100000@luomat> Timothy J Luoma wrote: > VirtSpace was a nice little program that was bought and turned > commercial.... I can never remember the name of the company which bought > it. WideScreen is a freeware app which can be found here: > > ftp://next-ftp.peak.org/pub/next/apps/utils/workspace/WideScreen.0.5.README > > ftp://next-ftp.peak.org/pub/next/apps/utils/workspace/WideScreen.0.5.NIHS.bs.tar > .gz > > ftp://next-ftp.peak.org/pub/next/apps/utils/workspace/WideScreen.0.5C.NIHS.bs.ta > r.gz > > (the 0.5C version has some color support added, but I believe it is the > same in all other respects) > > TjL Though it's been quite a while since I made the changes, I can confirm that the color version of WideScreen is identical to the monochrome version except for cosmetics. Regards, John Hornkvist Name: nhoj Site: cd.chalmers.se
From: bewebste@mtu.edu (Brian Webster) Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.programmer.codewarrior Subject: Re: mac out in the cold? Date: Sat, 25 Apr 1998 20:03:18 -0400 Organization: MTU Message-ID: <bewebste-2504982003180001@powermac.resnet.mtu.edu> References: <6hnpj3$spq$4@newsfep3.sprintmail.com> <forrestDELETE-THIS!-2304982216240001@term3-19.vta.west.net> <6hqf9g$s1a$1@supernews.com> In article <6hqf9g$s1a$1@supernews.com>, "Zico" <ZicoKnows@hotmail.com> wrote: >x-no-archive: yes > >Forrest Cameranesi wrote in message ... > >>[With Rhapsody, y]ou've got legacy and >>future apps seamlessly blended, a decent installed base, and full buzzword >>compliance, and you could probably sell it for Intel machines (minus blue >>box) fairly well too. What more could you ask for? > > >The ability to use both CPUs in my Intel box, >like Linux, Windows NT, and BeOS. Rhapsody (and NeXTSTEP) has full symmetric multiprocessing support. You can use your multiple CPUs to your heart's content. Anything else? --------------------------------------------------------------------- "Innovation is the ability to integrate a vast array of seemingly unrelated capabilities." - Microsoft, in a quarter-page NY Times ad "And over here, Swiss cheese, spliced with chalk and a beard" - The crazy South Park geneticist Brian Webster bewebste@mtu.edu
From: fractture <fract@sprintmail.com> Newsgroups: comp.sys.next.programmer Subject: gnu libFoundation Date: Sat, 25 Apr 1998 08:29:19 -0700 Message-ID: <3542014F.5BAABB53@sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit libFoundation compiled fine. I have the headers in Foundation wich is in one of my include dirs. libFoundation is in one of my lib dirs.... the objc runtime libs are in the compilers lib and includes in its includes. I can compile objc programs just fine, but not if they use libFoundation. here is the file, compile line, and error messages....could someone let me know how to fix this? #import <Foundation/NSFileHandle.h> #import <Foundation/NSString.h> #import <Foundation/NSData.h> #import <objc/objc.h> /* added these 2 to try getting it to work. */ #import <objc/Object.h> @interface HelloWorld : NSObject /* also tried Object */ { } -hello; @end @implementation HelloWolrd - hello { NSFileHandle *out = [NSFileHandle fileHandleWithStandardOutput]; NSString *theString = @"Hello World!\n"; NSData *finalOut = [NSData dataWithBytes:[theString cString] length:[theString cStringLength] ]; [out writeData: finalOut]; return self; } @end int main(void) { [[HelloWorld alloc] hello]; return 0; } gcc -Wno-import -Wall -g -o hello HelloWorld.m -lobjc -lFoundation -lm /usr/local/lib/libFoundation.a(NSConcreteString.o): In function `_i_NS8BitString__rangeOfCharacterFromSet_options_range_': /home/jik-/libFoundation-0.8.0/libFoundation/Foundation/NSConcreteString.m:81: multiple definition of `__objc_class_name_NXConstantString' /usr/lib/gcc-lib/i586-pc-linux-gnulibc1/pgcc-2.90.27/libobjc.a(NXConstStr.o)(.rodata+0x4): first defined here /usr/local/lib/libFoundation.a(NSObject.o): In function `_i__NSObjectDelayedExecutionHolder__isEqual_': /home/jik-/libFoundation-0.8.0/libFoundation/Foundation/NSObject.m:78: undefined reference to `__objc_class_name_Protocol' /usr/local/lib/libFoundation.a(NSArray.o): In function `_c_NSArray__arrayWithObject_': /home/jik-/libFoundation-0.8.0/libFoundation/Foundation/NSArray.m:70: undefined reference to `__objc_class_name_Protocol' looks to me like the methods and vars in the objc runtime are not being linked or something. Any ideas here?
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody Graphics Questions Date: 26 Apr 1998 03:24:01 GMT Organization: Digital Fix Development Message-ID: <6hu9ch$al0$1@news.digifix.com> References: <zegelin-2404981257030001@dialup35.canb.ispsys.net> <6hp0q0$b59$1@news.digifix.com> <zegelin-2504982145560001@dialup95.canb.ispsys.net> In-Reply-To: <zegelin-2504982145560001@dialup95.canb.ispsys.net> On 04/25/98, Peter wrote: <snip> > > OK, Ive outlined a few places where I find regions useful and I'd be >interested to know roughly how I'd handle the same things using Rhapsody. >I think I know the answer (always use offscreen drawing?) but I'll ask >just in case I've missed something fundamental. > >1. Simple graphic object layering. > I keep an ordered list of objects and as each object is drawn (near >objects first) its region is subtracted from the ClipRgn which means that >any object drawn later is automatically clipped by any object drawn >earlier. This is horribly clumsy. Draw from the back to the front and you get the same result without having to incur all that clipping overhead. > >2. Dragging a selection > Create a Bitmap the size of the selection. Draw the selected objects >into the Bitmap. Make a region from this using BitMapToRgn() then call >DragGreyRgn() to let the system drag my selection around while the mouse >is down. > Draw the selected objects into a bitmap, actually drag the bitmap around.. >3. Selection Rectangles > Each time through the mouse tracking loop: > > Create 2 regions based on the current selection rectangle. Inset one of >the regions by 1 pixel then subtract this smaller region from the the other >to create a region consisting of a one pixel thick outline of the current >selection rectangle. > Do the same for the previous selection rectangle (from the previous >iteration of the loop). > Get region which includes only those pixels that are NOT in both regions >and paint this new region in gray in Xor mode so pixels are flipped. > Instance drawing... -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: fractture <fract@sprintmail.com> Newsgroups: comp.sys.next.programmer Subject: Re: gnu libFoundation Date: Sat, 25 Apr 1998 08:59:40 -0700 Message-ID: <3542086C.E5BC941D@sprintmail.com> References: <3542014F.5BAABB53@sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I found it guys....I was missing -ldl apperently. I messed with the order of the libs a bit, got errors down to undef to 'dlopen' and did a man dlopen.. anyway, thanks ;)
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <23877892958423@digifix.com> Date: 26 Apr 1998 03:50:15 GMT Organization: Digital Fix Development Message-ID: <13743893563222@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1993. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
Message-ID: <3542854F.7341@hotmail.com> From: Clark Cox <coxware@hotmail.com> Organization: Coxware Software MIME-Version: 1.0 Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.oop.powerplant,comp.sys.mac.,programmer.codewarrior,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple developer program - from 12-year-old's perspective References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154><6ggkg1$90k$1@news0-alterdial.uu.net> <SCOTT.98Apr8221520@slave.doubleu.com> <6htoch$a3c$1@nnrp4.snfc21.pbi.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Sun, 26 Apr 1998 00:56:39 GMT NNTP-Posting-Date: Sat, 25 Apr 1998 20:56:39 EST Aaron Lehmann wrote: > > I have a different perspective from which I view Apple's developer program - > from mine, a 12 year old's. > > I am extremely enthusiastic about Rhapsody. It was my hope to join the Apple > developer program and begin working on Rhapsody applications immediately. > Now I cannot afford the cost of the developer program, and there is no way I > can legally get my hands on a copy of Rhapsody before the first public > release. This is effectively forcing me to delay my start. The same goes for > my 14-year old friend. It will be a loss for both Apple and me. I suppose > that a few neat shareware apps wouldn't really matter much to Apple, but > look, a new operating system is comming out and to suceed it needs > applications initially: > > Applications = Developers = Can they get their hands on Rhapsody? > > Now, Apple could have made money off of me if they had kept their prices > down to a resonable level where I could afford it. I think what Apple has to > do is get as many developers as possible so they can have applications ready > when Rhapsody is released. Sending out CD's for low prices would pay off, > because more applications would be available when it ships. Apple should be > actively persuading developers to write for Rhapsody, but how can they if > some developers cannot even afford to because of them? > > I have heard that there is an academic developer program. Is there any way > that a middle school student might be able to join? The price isn't really the limiting factor in this equasion, I'm sorry to say, but you must be 18 to be in any of Apple's developer programs. -- Clark S. Cox, III | ClarkCox3@hotmail.com Coxware Software | Coxware@hotmail.com "The only stupid question is the one not asked" --Unknown e-mail harvester traps: Chairman Reed Hundt: rhundt@fcc.gov Commissioner James Quello: jquello@fcc.gov Commissioner Susan Ness: sness@fcc.gov Commissioner Rachelle Chong: rchong@fcc.gov president@whitehouse.gov mwa234@hotmail.com bestrealtor@marketingmaster.com bstar@sssnet.com info@herbchew.com yyyr4t7@biblioteca.com clickthru@timefreedom.com nantragod@earthlink.net invite@onlinenow.net hdn94-018.hil.compuserve.com zippydj@nevwest.com haniophile@ntr.net Offer@shire.com inetmktg@usa.net empower@empowerlabs.com info@dproducts.be dynamarket@vaprnet.com promotions@the-bookstore.com root@mail.icongrp.com healthy181@aol.com cashrewards@hotmail.com Success@paper.com tei@websecure.net bb77@wyoma.com removeadultinf
From: Pascal Bourguignon <pjb@imaginet.fr> Newsgroups: comp.sys.next.programmer Subject: Re: NSString Date: 26 Apr 1998 05:06:18 GMT Organization: Deficiente Message-ID: <6hufca$ehj$1@news.imaginet.fr> References: <35420B5A.B4B35F53@denalisites.com> Brian Wotring <bj@denalisites.com> wrote: >I'm having problems understanding how to use the [aString >characterAtIndex:anIndex] method. I'm trying to take an NSString and >put every character into an NSMutableArray of NSStrings. For example, >if I had the string: "cat", the result would be an NSMutableArray as >follows: > >0 -> "c" >1 -> "a" >2 -> "t" > >Any help on how I might acheive this? > >-- > > | Brian Wotring | bj@denalisites.com ( NeXTMail welcome ) > | http://tact.dyn.ml.org > > ---------------------------------------------------------------- #import <foundation/NSString.h> #import <foundation/NSUtilities.h> @interface NSString(Split) -(NSMutableArray*)splitCharacters; @end @implementation NSString(Split) -(NSMutableArray*)splitCharacters { unsigned int count=[self length]; NSMutableArray* result=[NSMutableArray arrayWithCapacity:count]; unichar buffer; unsigned int i; for(i=0;i<count;i++){ buffer=[self characterAtIndex:i]; [result addObject: [NSString stringWithCharacters:&buffer length:1]]; } return(result); }//splitCharacters; @end int main(void) { NSLog(@"splited=%@\n",[@"cat" splitCharacters]); return(0); }//main. ---------------------------------------------------------------- cc -o test test.m -lFoundation_s ; ./test Apr 26 07:02:45 [1750] splited=(c, a, t) Of course, this "splited" data structure is quite expansive. If you need to work with characters, you should use C arrays of unichar. At least, optimize the splitCharacters method by using a buffer the same size as the string self. __Pascal Bourguignon__
From: zegelin@actonline.com.au (Peter) Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody Graphics Questions Date: 26 Apr 1998 06:16:56 GMT Organization: CyberOne Pty Ltd Message-ID: <zegelin-2604981617440001@dialup24.canb.ispsys.net> References: <zegelin-2404981257030001@dialup35.canb.ispsys.net> <6hp0q0$b59$1@news.digifix.com> <zegelin-2504982145560001@dialup95.canb.ispsys.net> <6hu9ch$al0$1@news.digifix.com> Cache-Post-Path: genie.cyberone.com.au!unknown@dialup24.canb.ispsys.net In article <6hu9ch$al0$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: > On 04/25/98, Peter wrote: > <snip> > > > > OK, Ive outlined a few places where I find regions useful and > I'd be > >interested to know roughly how I'd handle the same things using > Rhapsody. > >I think I know the answer (always use offscreen drawing?) but I'll > ask > >just in case I've missed something fundamental. > > > >1. Simple graphic object layering. > > I keep an ordered list of objects and as each object is drawn > (near > >objects first) its region is subtracted from the ClipRgn which means > that > >any object drawn later is automatically clipped by any object drawn > >earlier. > > This is horribly clumsy. > > Draw from the back to the front and you get the same result > without having to incur all that clipping overhead. > Yeah, but if I do this (in MacOs) I get flashing unless I use an offscreen buffer then transfer the final result to the screen. Machine speed is probably a factor here. The above method may be "horribly clumsy" but works well for a smallish (< 50) number of items. Anyway, the way you describe is certainly much easier and thanks. Also, please lighten up a bit. Im starting to wish I'd never posted.
From: fractture <fract@sprintmail.com> Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: NSFileHandle and irc Date: Sat, 25 Apr 1998 12:26:45 -0700 Message-ID: <354238F5.B65968E2@sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I am working on learning objective-c. I have decided to start with some socket work, namely irc. So far, I can connect, and send all the info needed to join a channel on a server and say something. But, I am not able to figure out a good loop with tha available methods in NSFileHandle, the class I am using. I have thought to use readInBackgroundAndNotify, but not sure how that would work. this is what I tried, but it dumped. -run { NSData *in; do{ in = [irc availableData]; [self relay:in]; } while(in != NULL); } relay:(NSData *)data is uncomfirmed, but it doesn't use anything unusual, or not used already to get to this point. Should I just forget this and use some other set of classes, or is there a way to do this? Anyone know a good way to go about reading from an irc server?
From: John Hornkvist <sorry@no.more.spams> Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody Graphics Questions Date: Sun, 26 Apr 1998 11:52:02 GMT Organization: Chalmers Tekniska Högskola Sender: john@haddock.cd.chalmers.se (John Hprnkvist) Message-ID: <Es0rMq.385@haddock.cd.chalmers.se> References: <zegelin-2404981257030001@dialup35.canb.ispsys.net> <6hp0q0$b59$1@news.digifix.com> <zegelin-2504982145560001@dialup95.canb.ispsys.net> <6hu9ch$al0$1@news.digifix.com> <zegelin-2604981617440001@dialup24.canb.ispsys.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: zegelin@actonline.com.au In <zegelin-2604981617440001@dialup24.canb.ispsys.net> Peter wrote: >> [Regions vs ordered drawing snipped] > > Yeah, but if I do this (in MacOs) I get flashing unless I use an > offscreen buffer then transfer the final result to the screen. Machine > speed is probably a factor here. The above method may be "horribly clumsy" > but works well for a smallish (< 50) number of items. Since DPS does double buffered drawing by default, you will not have to bother with doing you own buffering, unless you do very complex drawing. And even then it is often better to cache the drawing commands as one or more UserPaths rather than to cache the bitmap. Note that ordered drawing can be a bit of a problem too; using the naive implementation you'll have to draw all objects (or at least check their bounds) every time you redraw a part of the screen. That becomes quite slow when you have a lot of objects to deal with. Draw.app works around this by using an NSImage as a cache. In MagnaCharta I avoid the problem by tiling the view. Which method you should use depends on the size of the view and the number of objects. The approach in Draw is a fair bit simpler, and works well for small views with a reasonable number of objects, so unless performance becomes a problem, that's what I'd recommend using. Regards, John Hornkvist Name: nhoj Address:cd.chalmers.se
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody Graphics Questions Date: 26 Apr 1998 12:49:52 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6hvojg$iil$1@crib.bevc.blacksburg.va.us> References: <zegelin-2404981257030001@dialup35.canb.ispsys.net> <6hu9ch$al0$1@news.digifix.com> <zegelin-2604981617440001@dialup24.canb.ispsys.net> <Es0rMq.385@haddock.cd.chalmers.se> In article <Es0rMq.385@haddock.cd.chalmers.se>, sorry@no.more.spams wrote: > Note that ordered drawing can be a bit of a problem too; using the naive > implementation you'll have to draw all objects (or at least check their > bounds) every time you redraw a part of the screen. That becomes quite slow > when you have a lot of objects to deal with. Draw.app works around this by > using an NSImage as a cache. In MagnaCharta I avoid the problem by tiling the > view. What is "tiling the view"?
From: John Hornkvist <sorry@no.more.spams> Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody Graphics Questions Date: Sun, 26 Apr 1998 17:56:35 GMT Organization: Chalmers Tekniska Högskola Sender: john@haddock.cd.chalmers.se (John Hprnkvist) Message-ID: <Es18IC.40I@haddock.cd.chalmers.se> References: <zegelin-2404981257030001@dialup35.canb.ispsys.net> <6hu9ch$al0$1@news.digifix.com> <zegelin-2604981617440001@dialup24.canb.ispsys.net> <Es0rMq.385@haddock.cd.chalmers.se> <6hvojg$iil$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: nurban@crib.bevc.blacksburg.va.us In <6hvojg$iil$1@crib.bevc.blacksburg.va.us> Nathan Urban wrote: > > What is "tiling the view"? 2tile vt tiled; til·ing (14c) 1: to cover with tiles 2: to install drainage tile in Ð til·er n Splitting the view into several rectangles (tiles). When a graphic is created, moved or resized, you check what tiles it touches. When you need to redraw part of the screen, you check what tiles need to be redrawn, and then do the in order drawing for those tiles. The tiles should be large enough that it is rare that a graphics covers more than one tile. Would there be any interest in an article on the subject of fast drawing for large views with a large number of objects? I have written a draft for one, so Scott willing, I should be able to submit something to StepWise within the next few days. Regards, John Hornkvist Name: nhoj Address: cd.chalmers.se
From: mark@sapphire.oaai.com (Mark Onyschuk) Newsgroups: comp.sys.next.programmer Subject: Re: Apple developer program - from 12-year-old's perspective Date: 26 Apr 1998 15:18:33 GMT Organization: M. Onyschuk and Associates Inc. Message-ID: <slrn6k75m2.5he.mark@sapphire.oaai.com> References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154><6ggkg1$90k$1@news0-alterdial.uu.net> <SCOTT.98Apr8221520@slave.doubleu.com> <6htoch$a3c$1@nnrp4.snfc21.pbi.net> <3542854F.7341@hotmail.com> <slrn6k75gh.5he.mark@sapphire.oaai.com> [mangled an earlier post, here's a repost with the original suggestion] >Aaron Lehmann wrote: >> I have a different perspective from which I view Apple's developer program - >> from mine, a 12 year old's. >> >> I am extremely enthusiastic about Rhapsody. It was my hope to join the Apple >> developer program and begin working on Rhapsody applications immediately. >> Now I cannot afford the cost of the developer program, and there is no way I >> can legally get my hands on a copy of Rhapsody before the first public >> release. This is effectively forcing me to delay my start. The same goes for >> my 14-year old friend. It will be a loss for both Apple and me. I suppose Perhaps you should see if you can locate a local Rhapsody s/w developer; they might be able to lend a hand. I know I'd be mightily impressed to see someone of your age looking into this stuff [It'd remind me of myself *way* back when :-)]. They might see fit to offer you some cycles on a box they own, or otherwise help you out. Regards, Mark --- http://www.oaai.com/ -- M. Onyschuk and Associates Inc. GlyphiX - Graphics "for the rest of us!" OPENSTEP software development
From: mark@sapphire.oaai.com Newsgroups: comp.sys.next.programmer Subject: cancel <slrn6k75gh.5he.mark@sapphire.oaai.com> Control: cancel <slrn6k75gh.5he.mark@sapphire.oaai.com> Date: 26 Apr 1998 15:18:53 GMT Organization: A poorly-installed InterNetNews site Message-ID: <6hvj8t$fiq$1@ns3.vrx.net> ignore Article canceled by slrn 0.9.4.3
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody Graphics Questions Date: 26 Apr 1998 18:05:01 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6i0b2d$ora$1@crib.bevc.blacksburg.va.us> References: <zegelin-2404981257030001@dialup35.canb.ispsys.net> <Es0rMq.385@haddock.cd.chalmers.se> <6hvojg$iil$1@crib.bevc.blacksburg.va.us> <Es18IC.40I@haddock.cd.chalmers.se> In article <Es18IC.40I@haddock.cd.chalmers.se>, sorry@no.more.spams wrote: > In <6hvojg$iil$1@crib.bevc.blacksburg.va.us> Nathan Urban wrote: > > What is "tiling the view"? > Splitting the view into several rectangles (tiles). When a graphic is > created, moved or resized, you check what tiles it touches. When you need to > redraw part of the screen, you check what tiles need to be redrawn, and then > do the in order drawing for those tiles. > The tiles should be large enough that it is rare that a graphics covers more > than one tile. Ah, okay. I've seen this technique applied before (with quadtrees instead of a regular grid), I just wasn't familiar with the term. > Would there be any interest in an article on the subject of fast drawing for > large views with a large number of objects? I have written a draft for one, > so Scott willing, I should be able to submit something to StepWise within the > next few days. Yes, definitely! Since you bring it up, here's something related that I'm trying to do and have been meaning to post about: interactive visualization and manipulation of a large graph structure, sort of like what you can do with Diagram or GlyphiX, except it doesn't have to be as sophisticated. Nodes are represented by shapes (simple polygons or ovals, or images) bounded by rectanges, and edges are splines or Bezier curves (guess the latter would be easier with the new Rhapsody class). Only a portion of the graph will be displayed visible at one time (in the clipped content view of a scrollview), but what is visible could potentially still contain many nodes and edges. I'm trying to figure out how to be able to render it efficiently, allowing for the possibility of either interactively or programmatically moving a node or edge (which would have to force a recalculation of any edge curves attached to a node) and redrawing the affected portions of the view quickly -- as in realtime. If you or anyone else could give me suggestions on how to do this, that would be wonderful. But even if your article isn't directly useful to my problem, I'd still love to see anything on any kind of fast drawing, since I'll be needing to do that later anyway.
Newsgroups: comp.sys.next.programmer Subject: Re: Apple developer program - from 12-year-old's perspective From: miba@image.dk (Michael Balle) Date: Mon, 27 Apr 1998 00:28:59 +0200 Message-ID: <1d84bs8.1dkvqxp1e4rayoN@[10.0.0.2]> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggkg1$90k$1@news0-alterdial.uu.net> <SCOTT.98Apr8221520@slave.doubleu.com> <6htoch$a3c$1@nnrp4.snfc21.pbi.net> <3542854F.7341@hotmail.com> <slrn6k75gh.5he.mark@sapphire.oaai.com> <slrn6k75m2.5he.mark@sapphire.oaai.com> Mark Onyschuk <mark@sapphire.oaai.com> wrote: > >Aaron Lehmann wrote: > I have a different perspective from which I view > >Apple's developer program - > from mine, a 12 year old's. > > I am > >extremely enthusiastic about Rhapsody. It was my hope to join the Apple > > >developer program and begin working on Rhapsody applications immediately. > >> Now I cannot afford the cost of the developer program, and there is no > >way I > can legally get my hands on a copy of Rhapsody before the first > >public > release. This is effectively forcing me to delay my start. The > >same goes for > my 14-year old friend. It will be a loss for both Apple > >and me. I suppose > > Perhaps you should see if you can locate a local Rhapsody s/w developer; > they might be able to lend a hand. > > I know I'd be mightily impressed to see someone of your age looking into > this stuff [It'd remind me of myself *way* back when :-)]. They might see > fit to offer you some cycles on a box they own, or otherwise help you out. > > Regards, Mark I don't even think that Apple will let him attend the developer conference. I remember a very young boy attending the conference in 93 or 94. During one of the session's he even sat next to Sculley. Since that year, I believe there has been a minimum age for attending WWDC. Michael
From: Craig Brozefsky <craig@duomo.onshore.com> Newsgroups: comp.sys.next.programmer Subject: Re: Apple developer program - from 12-year-old's perspective Date: 26 Apr 1998 18:28:31 -0500 Organization: EnterAct L.L.C. Turbo-Elite News Server Message-ID: <87g1j0macg.fsf@duomo.onshore.com> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggkg1$90k$1@news0-alterdial.uu.net> <SCOTT.98Apr8221520@slave.doubleu.com> <6htoch$a3c$1@nnrp4.snfc21.pbi.net> <3542854F.7341@hotmail.com> <slrn6k75gh.5he.mark@sapphire.oaai.com> <slrn6k75m2.5he.mark@sapphire.oaai.com> <1d84bs8.1dkvqxp1e4rayoN@[10.0.0.2]> miba@image.dk (Michael Balle) writes: > I don't even think that Apple will let him attend the developer > conference. I remember a very young boy attending the conference in 93 > or 94. During one of the session's he even sat next to Sculley. Since > that year, I believe there has been a minimum age for attending WWDC. Man, that's what PhotoShop and color laser printers with laminators are for! This type of problem exactly! 8^)
From: zegelin@actonline.com.au (Peter) Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody Graphics Questions Date: 26 Apr 1998 23:44:17 GMT Organization: CyberOne Pty Ltd Message-ID: <zegelin-2704980945100001@dialup46.canb.ispsys.net> References: <zegelin-2404981257030001@dialup35.canb.ispsys.net> <6hu9ch$al0$1@news.digifix.com> <zegelin-2604981617440001@dialup24.canb.ispsys.net> <Es0rMq.385@haddock.cd.chalmers.se> <6hvojg$iil$1@crib.bevc.blacksburg.va.us> <Es18IC.40I@haddock.cd.chalmers.se> Cache-Post-Path: genie.cyberone.com.au!unknown@dialup46.canb.ispsys.net In article <Es18IC.40I@haddock.cd.chalmers.se>, sorry@no.more.spams wrote: [snip] > > Would there be any interest in an article on the subject of fast drawing for > large views with a large number of objects? I have written a draft for one, > so Scott willing, I should be able to submit something to StepWise within the > next few days. > > Regards, > John Hornkvist > Name: nhoj > Address: cd.chalmers.se Thanks heaps for the previous info and yes please! for the article. Regards, Peter
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody Graphics Questions Date: 27 Apr 1998 03:47:59 GMT Organization: Digital Fix Development Message-ID: <6i0v5f$amg$1@news.digifix.com> References: <zegelin-2404981257030001@dialup35.canb.ispsys.net> <6hp0q0$b59$1@news.digifix.com> <zegelin-2504982145560001@dialup95.canb.ispsys.net> <6hu9ch$al0$1@news.digifix.com> <zegelin-2604981617440001@dialup24.canb.ispsys.net> In-Reply-To: <zegelin-2604981617440001@dialup24.canb.ispsys.net> On 04/25/98, Peter wrote: <snip> >> Draw from the back to the front and you get the same result >> without having to incur all that clipping overhead. >> > > > Yeah, but if I do this (in MacOs) I get flashing unless I use an >offscreen buffer then transfer the final result to the screen. Machine >speed is probably a factor here. The above method may be "horribly clumsy" >but works well for a smallish (< 50) number of items. > I used to use off-screen bitmap when I was doing this on MacOS.. There are other issues that come into play using the regions.. for example if you have large pen-widths, you get additional clipping that you don't necessarily want. > Anyway, the way you describe is certainly much easier and thanks. > > Also, please lighten up a bit. Im starting to wish I'd never posted. > I meant that the technique was horribly clumsy. It wasn't intended as a hassle of you personally. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: leigh@NOSPAMcs.uwa.edu.au (Leigh Smith) Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.programmer.codewarrior Subject: Re: mac out in the cold? Date: 27 Apr 1998 04:35:08 GMT Organization: The University of Western Australia Message-ID: <6i11ts$ps5$1@enyo.uwa.edu.au> References: <6hnpj3$spq$4@newsfep3.sprintmail.com> <forrestDELETE-THIS!-2304982216240001@term3-19.vta.west.net> <6hqf9g$s1a$1@supernews.com> <bewebste-2504982003180001@powermac.resnet.mtu.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: bewebste@mtu.edu In <bewebste-2504982003180001@powermac.resnet.mtu.edu> Brian Webster wrote: > In article <6hqf9g$s1a$1@supernews.com>, "Zico" <ZicoKnows@hotmail.com> wrote: > > >x-no-archive: yes > > > >Forrest Cameranesi wrote in message ... > > > >The ability to use both CPUs in my Intel box, > >like Linux, Windows NT, and BeOS. > > Rhapsody (and NeXTSTEP) has full symmetric multiprocessing support. You > can use your multiple CPUs to your heart's content. Anything else? > Actually SMP is yet to appear outside of a NeXT/Apple Lab. To my knowledge it is still in development - (read: currently vapour until further sightings). You can't compare that capability with shipping systems, no matter how good the SMP support is puported to be (and I hope it is!). -- Leigh Computer Science, University of Western Australia Smith +61-8-9380-3778 leigh@cs.uwa.edu.au (NeXTMail/MIME) Microsoft - never has so much been made by so few, by pushing so much of so little, on so many with so little resistance.
From: tom_e@usa.net (Thomas Engelmeier) Newsgroups: comp.sys.next.programmer Subject: Re: Apple developer program - from 12-year-old's perspective Date: Mon, 27 Apr 1998 11:48:44 +0200 Organization: University of Rostock Message-ID: <1d852xi.1vy86c09k5b38N@desktop.tom-e.private> References: <01bc445f$744397e0$1bf0bfa8@davidsul> <ericb-0704981749520001@132.236.171.104> <trumbull-0704981900290001@net44-223.student.yale.edu> <j-jahnke-0704982342480001@192.168.1.3> <352b5da2.0@206.25.228.5> <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154> <6ggkg1$90k$1@news0-alterdial.uu.net> <SCOTT.98Apr8221520@slave.doubleu.com> <6htoch$a3c$1@nnrp4.snfc21.pbi.net> <3542854F.7341@hotmail.com> <slrn6k75gh.5he.mark@sapphire.oaai.com> <slrn6k75m2.5he.mark@sapphire.oaai.com> <1d84bs8.1dkvqxp1e4rayoN@[10.0.0.2]> Michael Balle <miba@image.dk> wrote: > > Regards, Mark > I don't even think that Apple will let him attend the developer > conference. I remember a very young boy attending the conference in 93 > or 94. During one of the session's he even sat next to Sculley. Since > that year, I believe there has been a minimum age for attending WWDC. At least, there was another boy at the 1996 WWDC: He did the presentaion for - how was it called - Colada?... Regards, TomE
From: matthias@:-)factum-gmbh.de (Matthias Schürhoff) Newsgroups: comp.sys.next.programmer Subject: ASCII file from Openstep NT to printer? Date: 27 Apr 1998 12:10:03 GMT Organization: FACTUM Projektentwicklung und Management GmbH Message-ID: <6i1sir$slj@factum.factum-gmbh.de> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Hello, is it possible to send or copy an ASCII file from an Openstep application under Windows NT directly to a printer? Or perhaps it works by using the pasteboard? I know that I can drag the icon from the saved file to a printer on the desktop but it would be easier if it would work directly from the application. With best regards Mattthias =========================================================== Matthias Schürhoff FACTUM Projektentwicklung und Management GmbH, Dortmund E-Mail: matthias@factum-gmbh.de (NeXT & MIME mail ok) WARNING: The return email address field has been altered:-) ----------------------------------------------------------- "A marriage is the most expensive way to find out one's mistakes." - Unknown
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.programmer.codewarrior Subject: Re: mac out in the cold? Date: 27 Apr 1998 14:29:00 GMT Organization: P & L Systems Message-ID: <6i24nc$bll$14@ironhorse.plsys.co.uk> References: <6hnpj3$spq$4@newsfep3.sprintmail.com> <forrestDELETE-THIS!-2304982216240001@term3-19.vta.west.net> <6hqf9g$s1a$1@supernews.com> <bewebste-2504982003180001@powermac.resnet.mtu.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: bewebste@mtu.edu In <bewebste-2504982003180001@powermac.resnet.mtu.edu> Brian Webster wrote: > In article <6hqf9g$s1a$1@supernews.com>, "Zico" <ZicoKnows@hotmail.com> wrote: > > > > What more could you ask for [from Rhapsody]? > > > >The ability to use both CPUs in my Intel box, > >like Linux, Windows NT, and BeOS. > > Rhapsody (and NeXTSTEP) has full symmetric multiprocessing support. You > can use your multiple CPUs to your heart's content. Anything else? > Umm, Rhapsody may well have SMP support, however NEXTSTEP does not. NEXTSTEP is SMP-able: when you boot in verbose mode you get a message saying hat the Mach kernel is configured for one processor, and NeXT had a version of NEXTSTEP running on a dual-PPC machine back in 92/3. One of the main reasons apparently for not supporting SMP on Intel before was the lack of a consistent standard for multi-CPU systems, and lack of resources. Best wishes, mmalc.
From: moellney@simtec.imr.mb.uni-siegen.de (Michael Möllney) Newsgroups: comp.sys.next.programmer Subject: latex2html installing problem Organization: Uni Siegen MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <3544a331.0@rainbow.hrz.uni-siegen.de> Date: 27 Apr 98 15:24:33 GMT Hi! I'm trying to install LaTeX2HTML V98.1 on my Openstep 4.2 System. install-test gives me: ... Checking for availability of DBM or NDBM (Unix DataBase Management)... *** ERROR: AnyDBM_File doesn't define a TIEHASH method at (eval 4) line 1 You will not be able to use latex2html. ... The rest seems to work. Has anybody an idea what goes wrong here ??? Please E-Mail me, thanks, Michael -- Michael Moellney Paul-Bonatz-Straûe 9-11, Raum 426/2 57068 Siegen Tel: +49-271-740-4724 moellney@simtec.imr.mb.uni-siegen.de
From: "George B. Ameer" <george1@ana.porsa.com> Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.programmer.codewarrior Subject: Re: mac out in the cold? Date: Mon, 27 Apr 1998 08:37:03 -0700 Message-ID: <6i28u6$g2a$1@news0-alterdial.uu.net> References: <6hnpj3$spq$4@newsfep3.sprintmail.com> <forrestDELETE-THIS!-2304982216240001@term3-19.vta.west.net> <6hqf9g$s1a$1@supernews.com> <bewebste-2504982003180001@powermac.resnet.mtu.edu> Huh?!? Where have I missed this? If SMP is finally supported I want it NOW! Brian Webster wrote in message ... > >Rhapsody (and NeXTSTEP) has full symmetric multiprocessing support. You >can use your multiple CPUs to your heart's content. Anything else?
Message-ID: <3544A6DC.FDDE570D@rit.edu> Date: Mon, 27 Apr 1998 11:40:12 -0400 From: Jeff Sciortino <jjs2815@rit.edu> Organization: RIT MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Re: Apple developer program - from 12-year-old's perspective References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154><6ggkg1$90k$1@news0-alterdial.uu.net> <SCOTT.98Apr8221520@slave.doubleu.com> <6htoch$a3c$1@nnrp4.snfc21.pbi.net> <3542854F.7341@hotmail.com> <slrn6k75gh.5he.mark@sapphire.oaai.com> <slrn6k75m2.5he.mark@sapphire.oaai.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit If you can get your hands on an old Pentium machine that's compatible, you also might want to post a request on comp.sys.next.marketplace to see if anyone has an old copy of NEXTSTEP academic the would sell cheaply. As we academic types upgrade, we have old copies we're not using lying around of older version. As long as it's version 3.2 or higher, I understand that it's pretty straightforward to port over to rhadsody. -jeff
From: nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.programmer.codewarrior Subject: Re: mac out in the cold? Date: 27 Apr 1998 17:04:00 GMT Organization: none Message-ID: <6i2dq0$6tj$4@ha2.rdc1.nj.home.com> References: <6hnpj3$spq$4@newsfep3.sprintmail.com> <forrestDELETE-THIS!-2304982216240001@term3-19.vta.west.net> <6hqf9g$s1a$1@supernews.com> <bewebste-2504982003180001@powermac.resnet.mtu.edu> <6i28u6$g2a$1@news0-alterdial.uu.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: george1@ana.porsa.com In <6i28u6$g2a$1@news0-alterdial.uu.net> "George B. Ameer" wrote: > Huh?!? > Where have I missed this? If SMP is finally supported I want it NOW! > > Brian Webster wrote in message ... > > > >Rhapsody (and NeXTSTEP) has full symmetric multiprocessing support. You > >can use your multiple CPUs to your heart's content. Anything else? Brian was mistaken about NeXTStep supporting it... I would assume Rhapsody will eventually TjL -- [Yes that is a valid mail address, until it gets spammed] Unofficial @Home FAQ: http://members.home.com/faqs/ NeXTStep/OpenStep/Rhapsody Information & Software: http://www.peak.org/~luomat/
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: pragma for turning off unknown method warnings? Date: 27 Apr 1998 16:47:10 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6i2qse$9h6$1@crib.bevc.blacksburg.va.us> I'm sending a lot of messages to a delegate, which is an 'id'. I get the usual zillion warnings about "cannot find method". For various reasons, I don't want to create a protocol for the delegate, so I can't just tell it that the 'id' conforms to a protocol. I think it's possible to turn off the warnings with a compiler option (but I can't seem to find it in the man page for cc(1)), but I don't want to turn off the warnings _all_ the time, just when compiling this one class. I could futz around with the Makefile so that the warnings are suppressed only for that one file, but I don't really want to. I was wondering if the compiler has any pragmas that I can put in my source file to turn off the warnings?
From: Stanley Newsgroups: comp.sys.next.programmer Date: Mon, 27 Apr 1998 17:25:27 PDT Subject: My First Print Organization: Email Platinum v.3.1b Message-ID: <3544f794.0@news.mountain.net> I would like to invite you to view "HANGIN IT UP", a limited addition print of an original oil painting. You may visit this print at http://www.ovnet.com/~estanley/Working2.html
From: kjNOSPAMray@ix.netcom.com (Ray) Newsgroups: comp.object,comp.sys.oop.macapp3,comp.sys.mac.oop.misc,comp.sys.mac.oop.powerplant,comp.lang.java.programmer,comp.lang.c++,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Sensible Template Classes Date: Mon, 27 Apr 1998 20:30:08 -0700 Organization: Netcom Message-ID: <kjNOSPAMray-2704982030080001@smx-ca10-18.ix.netcom.com> I wish to announce the 0.1 release of the Sensible Template Classes. The Sensible Template Classes (STC) are intended to: (1) provide a set of C++ collection classes that are easier to understand and work with than the "Standard Template Library". (2) provide a set of C++ collection classes that are similar to those of Java JDK, to make porting to and from Java/C++ easier. (3) be portable and freely available. To that end, I am writing this code and making it public domain. This code is intended to be cross-platform. I'm implementing it for MacOS/Metrowerks C++ first, Windows/Visual C++ second, Rhapsody and other Unices third. Your help in porting and bug-fixes is appreciated. Send updates, porting information, and bug-fixes to: C. Keith Ray <mailto:kjray@ix.netcom.com> The latest version will be at <http://pw2.netcom.com/~kjray/> (More precisely <http://pw2.netcom.com/~kjray/stc.sit.bin> -- macbinary stuffit format.) Please NOTE: this 0.1 release is not at all functional. It compiles but it does not link. Almost no functionality has actually been implemented. (Lots of empty function bodies.) No documentation is included, but you can look at the Java collection class documentation on-line in Sub's web site. Your help in completing the coding, as well as porting and bug-fixing will be appreciated. // Keith and Jane ---- // --- ---------- Check out: <http://www.devworld.apple.com/> --------- ---------- <http://www.evangelist.macaddict.com/> ----------- --- --- <http://pw2.netcom.com/~kjray/index.html>
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <35343ad4.0@193.15.242.210> Control: cancel <35343ad4.0@193.15.242.210> Date: 15 Apr 1998 21:39:48 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.35343ad4.0@193.15.242.210> Sender: CatherineW@email.unc.edu Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <3544f794.0@news.mountain.net> From: Stanley Control: cancel <3544f794.0@news.mountain.net> Date: Tue, 28 Apr 1998 06:06:56 GMT Message-ID: <cancel.3544f794.0@news.mountain.net> Sender: Stanley Message <3544f794.0@news.mountain.net> was cancelled by fifi@toby.han.de. Reason: Spam
Newsgroups: comp.sys.next.programmer Subject: Re: Apple developer program - from 12-year-old's perspective References: <6gg8il$6v9$1@ha2.rdc1.sdca.home.com> <B1511189-1AF6A@206.165.43.154><6ggkg1$90k$1@news0-alterdial.uu.net> <SCOTT.98Apr8221520@slave.doubleu.com> <6htoch$a3c$1@nnrp4.snfc21.pbi.net> <3542854F.7341@hotmail.com> <slrn6k75gh.5he.mark@sapphire.oaai.com> <slrn6k75m2.5he.mark@sapphire.oaai.com> <3544A6DC.FDDE570D@rit.edu> From: no_spam_frank@ifi.unibas.ch Message-ID: <354598e6.0@maser.urz.unibas.ch> Date: 28 Apr 98 08:52:54 GMT Jeff Sciortino <jjs2815@rit.edu> wrote: > If you can get your hands on an old Pentium machine that's compatible, > you also might want to post a request on comp.sys.next.marketplace to > see if anyone has an old copy of NEXTSTEP academic the would sell > cheaply. As we academic types upgrade, we have old copies we're not > using lying around of older version. > > As long as it's version 3.2 or higher, I understand that it's pretty > straightforward to port over to rhadsody. Nope, 4.0 or higher -Robert -- Institut fuer Informatik tel +41 (0)61 321 99 67 Universitaet Basel fax. +41 (0)61 321 99 15 Robert Frank Mittlere Strasse 142 rfc822: frank@ifi.unibas.ch (NeXT,MIME mail ok) CH-4056 Basel (remove any no_spam_ from my return address) Switzerland
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: pragma for turning off unknown method warnings? Date: 28 Apr 1998 12:08:07 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <6i4gr7$nc$1@pump1.york.ac.uk> References: <6i2qse$9h6$1@crib.bevc.blacksburg.va.us> On 27 Apr 1998 16:47:10 -0400, Nathan Urban <nurban@crib.bevc.blacksburg.va.us> wrote: > I'm sending a lot of messages to a delegate, which is an 'id'. I get the > usual zillion warnings about "cannot find method". For various reasons, > I don't want to create a protocol for the delegate, so I can't just tell > it that the 'id' conforms to a protocol. I think it's possible to turn > off the warnings with a compiler option (but I can't seem to find it in > the man page for cc(1)), but I don't want to turn off the warnings _all_ > the time, just when compiling this one class. I could futz around with > the Makefile so that the warnings are suppressed only for that one file, > but I don't really want to. I was wondering if the compiler has any > pragmas that I can put in my source file to turn off the warnings? is it really a good idea to turn off these warnings? after all they do represent a genuine potential problem - i.e. the compiler has no way of knowing what types of arguments and return types the method is taking, so on some architectures, it might easily make the wrong decision (e.g. put an argument in the wrong register, or whatever) and you'll get a bug that's very hard to find! what's wrong with just defining a virtual category of NSObject so that the compiler can get the method definitions from *somewhere*, without having to actually use protocols? you don't even have to make it globally visible - a declaration in the same source file is sufficient. you don't actually have to provide an implementation of the category. e.g. @interface NSObject(MyDelegateMethods) - (void)delegateMethod1:(int)x; - delegateMethod2:(id)y; - (struct something)delegateMethod3:(void *)x:(NSRect)r; @end cheers, rog.
From: abuse@mindspring.com Newsgroups: comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software Subject: cmsg cancel <35e0edbf.104873059@news.mindspring.com> Date: 28 Apr 1998 16:08:18 GMT Control: cancel <35e0edbf.104873059@news.mindspring.com> Message-ID: <cancel.35e0edbf.104873059@news.mindspring.com> Sender: CyBorg@cyborg-systems.com (Cy Borg) Cancelled by abuse@mindspring.com
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: pragma for turning off unknown method warnings? Date: 28 Apr 1998 13:57:47 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6i55ar$bp5$1@crib.bevc.blacksburg.va.us> References: <6i2qse$9h6$1@crib.bevc.blacksburg.va.us> <6i4gr7$nc$1@pump1.york.ac.uk> In article <6i4gr7$nc$1@pump1.york.ac.uk>, rog@ohm.york.ac.uk wrote: > On 27 Apr 1998 16:47:10 -0400, Nathan Urban <nurban@crib.bevc.blacksburg.va.us> wrote: > > I'm sending a lot of messages to a delegate, which is an 'id'. I get the > > usual zillion warnings about "cannot find method". > is it really a good idea to turn off these warnings? after all > they do represent a genuine potential problem - i.e. the > compiler has no way of knowing what types of arguments and return > types the method is taking, so on some architectures, it might > easily make the wrong decision (e.g. put an argument in the wrong > register, or whatever) and you'll get a bug that's very hard to find! Well, in traditional delegate style, I'm only sending the message if a runtime check reveals that the delegate responds to it, so it's not a problem. > what's wrong with just defining a virtual category of NSObject > so that the compiler can get the method definitions from *somewhere*, > without having to actually use protocols? you don't even have > to make it globally visible - a declaration in the same source file > is sufficient. Someone else suggested that too. That might be useful, except.. > e.g. > @interface NSObject(MyDelegateMethods) > - (void)delegateMethod1:(int)x; I have some delegate methods which might conflict in name with methods defined in some subclass of NSObject. (i.e., some Foundation or AppKit class might define a method named -delegateMethod1:.. though of course that's not a name I'm really using.) Will putting a category on NSObject screw up those subclasses since they inherit from it? Should I rename my delegate methods?
From: Dan Wellman <wellman@students.uiuc.edu> Newsgroups: comp.lang.objective-c,comp.sys.next.programmer Subject: Autorelease pools in OpenStep Date: 28 Apr 1998 19:59:51 GMT Organization: University of Illinois at Urbana-Champaign Message-ID: <6i5cfn$8s9$1@vixen.cso.uiuc.edu> From Don Yacktman's article about memory management in OpenStep: http://www.stepwise.com/Articles/Technical/HoldMe.html The article says that the default autorelease pool is created at the beginning of and cleared at the end of Openstep's "internal event loop". When exactly does this happen? I've got an object which houses the main execution loop of my program (i.e. creating hundreds of smaller objects that are created and destroyed during this loop, which is repeated several times). I'm running out of memory and I'm wondering if it's because OpenStep is not clearing out the autorelease pools. Any help appreciated! Thanks, dan -- Dan Wellman <> wellman@uiuc.edu <> http://www.cen.uiuc.edu/~wellman/ "A million thoughts in one night can't be wrong" - Cause & Effect
From: don@misckit.com (Don Yacktman) Newsgroups: comp.lang.objective-c,comp.sys.next.programmer Subject: Re: Autorelease pools in OpenStep Date: 28 Apr 1998 23:01:36 GMT Organization: MiscKit Development Message-ID: <6i5n4g$8bb$2@news.xmission.com> References: <6i5cfn$8s9$1@vixen.cso.uiuc.edu> Dan Wellman <wellman@students.uiuc.edu> wrote: > > From Don Yacktman's article about memory management in OpenStep: > > http://www.stepwise.com/Articles/Technical/HoldMe.html > > The article says that the default autorelease pool is created at the > beginning of and cleared at the end of Openstep's "internal event loop". > When exactly does this happen? I believe that the pool is emptied at the end of the event loop, before the next event is dispatched. So, if a mouse click starts some huge operation that greates billions of objects before returning, the release pool on't be emptied during that time. Once the method finally returns (thereby returning control to the event loop), then and only then will the pool be cleared. Is that making sense? In effect, don't expect to see any housecleaning until your action method returns. > I've got an object which houses the main execution loop of my program > (i.e. creating hundreds of smaller objects that are created and > destroyed during this loop, which is repeated several times). I'm > running out of memory and I'm wondering if it's because OpenStep is > not clearing out the autorelease pools. That is a possibility. What you want to do is some of your own release pool management. Say, for example, that you have an expensive operation that is inside a loop. Then, inside the loop, create your own pool before the operation and then release the pool after the operation. while (looping) { NSAutoreleasePool *pool =[[NSAutoreleasePool alloc] init]; // expensive operation here [pool release]; } By doing this, housecleaning will happen (it will be forced) each time you pass through the loop. Learning *where* to put these is the tricky part. You want to do as little of it as possible because it makes your code slightly more complicated and it definitely degrades your performance. Also, it is OK to nest pools, and therefore if an exception breaks you out of the loop, there still won't be any memory leaks caused by pools getting left around. So, read the class docs on NSAutoreleasePool. Lots of good stuff in there you'll need to know...such as how to move an object out of a pool and up a level... -- Later, -Don Yacktman don@misckit.com <a href="http://www.misckit.com/don.html">My home page</a>
Newsgroups: comp.sys.next.programmer From: brianw@sounds.wa.com (Brian Willoughby) Subject: Re: pragma for turning off unknown method warnings? Message-ID: <Es57z2.1sx.0.scream@sounds.wa.com> Organization: Sound Consulting, Bellevue, WA, USA References: <6i2qse$9h6$1@crib.bevc.blacksburg.va.us> <6i4gr7$nc$1@pump1.york.ac.uk> Date: Tue, 28 Apr 1998 21:35:26 GMT On 27 Apr 1998 16:47:10 -0400, Nathan Urban <nurban@crib.bevc.blacksburg.va.us> wrote: > I'm sending a lot of messages to a delegate, which is an 'id'. I get the > usual zillion warnings about "cannot find method". For various reasons, > I don't want to create a protocol for the delegate, so I can't just tell > it that the 'id' conforms to a protocol. I think it's possible to turn > off the warnings with a compiler option (but I can't seem to find it in > the man page for cc(1)), but I don't want to turn off the warnings _all_ > the time, just when compiling this one class. I could futz around with > the Makefile so that the warnings are suppressed only for that one file, > but I don't really want to. I was wondering if the compiler has any > pragmas that I can put in my source file to turn off the warnings? I believe that with objects of type 'id', the compiler is really only concerned with finding a method ANYWHERE with the same name, with no requirement about what class the method belongs to. Are you including the @interface file for your delegate in the class which is calling it? I believe that is all you need to do. The only drawback is that if you include too much, and the compiler finds more than one method with the same name, but differing parameter/return types, then you will get a warning about the conflict. But I've never seen "cannot find method" in this case (rather, the compiler finds too many methods!) You shouldn't need to create a protocol unless you want a clean interface for other programmers to write bundles or otherwise use your delegate, or provide compatible delegate objects. -- Brian Willoughby NEXTSTEP, OpenStep, Rhapsody Software Design Sound Consulting Apple Enterprise Alliance Partner NeXTmail welcome Macintosh Associate Apple is the registered trademark of Apple Computer, Inc. and Apple Records
From: "Frank Mitchell" <frankm@bayarea.net> Newsgroups: comp.lang.objective-c,comp.sys.next.programmer Subject: Re: Autorelease pools in OpenStep Date: 29 Apr 1998 08:02:37 GMT Organization: Bay Area Internet Solutions (408) 260-5000 Message-ID: <01bd7345$584e3520$9842dbcd@frankm> References: <6i5cfn$8s9$1@vixen.cso.uiuc.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Dan Wellman <wellman@students.uiuc.edu> wrote in article <6i5cfn$8s9$1@vixen.cso.uiuc.edu>... > The article says that the default autorelease pool is created at the > beginning of and cleared at the end of Openstep's "internal event loop". > When exactly does this happen? The "internal event loop" resides in the Application object. If you're not using that, you're not autoreleasing. > I've got an object which houses the main execution loop of my program > (i.e. creating hundreds of smaller objects that are created and > destroyed during this loop, which is repeated several times). I'm > running out of memory and I'm wondering if it's because OpenStep is > not clearing out the autorelease pools. Basically, create an NSAutoreleasePool at the top of your loop, and release it at the bottom (or just before you alloc another one). All autoreleased objects go into that pool, and get released when you release it. (It's best to alloc and release it in the same method, for clarity and safety.) I don't have the docs handy, but look at NSAutoreleasePool ... it has two basic methods (the class method that makes a new pool, and the -release method) and a few more for debugging flags. Hope this helps. Frank Mitchell
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: pragma for turning off unknown method warnings? Date: 29 Apr 1998 12:00:33 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <6i74p1$ir4$1@pump1.york.ac.uk> References: <6i2qse$9h6$1@crib.bevc.blacksburg.va.us> <6i4gr7$nc$1@pump1.york.ac.uk> <6i55ar$bp5$1@crib.bevc.blacksburg.va.us> On 28 Apr 1998 13:57:47 -0400, Nathan Urban <nurban@crib.bevc.blacksburg.va.us> wrote: > In article <6i4gr7$nc$1@pump1.york.ac.uk>, rog@ohm.york.ac.uk wrote: > > is it really a good idea to turn off these warnings? after all > > they do represent a genuine potential problem - i.e. the > > compiler has no way of knowing what types of arguments and return > > types the method is taking, so on some architectures, it might > > easily make the wrong decision (e.g. put an argument in the wrong > > register, or whatever) and you'll get a bug that's very hard to find! > > Well, in traditional delegate style, I'm only sending the message if > a runtime check reveals that the delegate responds to it, so it's not > a problem. this isn't sufficient to stop a potential problem, because checking that an object responds to a particular method only checks that it responds to a method of that *name* - but the compiler won't know what types the method expects as its arguments. to do it properly, you either need to provide a prototype to the compiler (by declaring the method somewhere), or check the types by hand, with an NSMethodSignature. if you know the types that the object is going to take, then i'd suggest that the former method is superior (better run-time performance, and typechecking at compile-time) > > what's wrong with just defining a virtual category of NSObject > > so that the compiler can get the method definitions from *somewhere*, > > without having to actually use protocols? you don't even have > > to make it globally visible - a declaration in the same source file > > is sufficient. > > Someone else suggested that too. That might be useful, except.. > > > e.g. > > > @interface NSObject(MyDelegateMethods) > > - (void)delegateMethod1:(int)x; > > I have some delegate methods which might conflict in name with methods > defined in some subclass of NSObject. (i.e., some Foundation or AppKit > class might define a method named -delegateMethod1:.. though of course > that's not a name I'm really using.) Will putting a category on NSObject > screw up those subclasses since they inherit from it? Should I rename > my delegate methods? it shouldn't be a problem. declaring an *interface* for a category of NSObject does not override any methods in NSObject at all. it merely causes the compiler to *think* that any object will implement the method declared, and so won't raise a warning. cheers, rog.
From: x97 <yumi8@hotmail.com> Newsgroups: comp.sys.next.programmer Subject: New to NEXTSTEP, please HELP!!! (easy question) Date: Wed, 29 Apr 1998 09:02:21 -0400 Organization: Ball State University Message-ID: <354724DD.A4C70437@hotmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit hello out there... I am new to NextStep.. a little late maybe... but anyways I got NextStep 3.2 CD and boot diskette for Intel CPU's I finally managed to boot the darn thing.. Now when it reboots for the first time and goes INTO the FIRST configuration menu (where you select devices, monitors, mice, etc etc) ... well the MOUSE does not MOVE!!! How am I supposed to "click" (as the manuals and text files mention so clamly) on the required components??!?! **PLEASE** help... I am desperate... isn't there keyboard control of the mouse cursor at ALL?? please reply asap to: yumi8@hotmail.com THANKS SO MUCH IN ADVANCE!! well since I am here, another question: I use the AHA-154xCF Adaptec SCSI controller, and it seems that the drivers on the 3.2 Boot diskette are not working with it, so I got a Floppy image with the newer driver and that worked (i had to specify that i wanted to load drivers off another disk). My problem is that I have to do that all the time now since the new driver was not installed onto the SCSI hard disk (sd0a) What can I do? (i am still in the configuration of the computer part..I read that I can go and add the driver .. that is IF i ever get to a CLI/shell) Also I got another version of these drivers in ".compressed" format... Well that is not very useful to me since my Nextstep is not running to begin with.. Can I somehow uncompress these files onto a floppy through another OS like: DOS or Linux ? thanks thanks thanks !!!
Newsgroups: comp.sys.next.programmer Subject: Re: New to NEXTSTEP, please HELP!!! (easy question) References: <354724DD.A4C70437@hotmail.com> From: sdroll@NOSPMmathematik.uni-wuerzburg.de (Sven Droll) Message-ID: <35472810.0@uni-wuerzburg.de> Date: 29 Apr 98 13:16:00 GMT ---snip--- >Also I got another version of these drivers in ".compressed" format... >Well that is not very useful to me since my Nextstep is not running to >begin with.. Can I somehow uncompress these files onto a floppy through >another OS like: DOS or Linux ? ---snip--- .compressed should be standard .tar.Z (gzip, gtar on Linux) or .Z -- Sven Droll __ ______________________________________________________/ / ______ __ sdroll@mathematik.uni-wuerzburg.de / /_/ ___/ please remove the NOSPM from my reply-address /_ _/ _/ =====\_/======= LOGOUT FASCISM! ___________________________________________________________________ NeXT-mail or MIME welcome ;-)
From: emarinos@marcon.de (E. Marinos) Newsgroups: comp.sys.next.programmer Subject: Support of Cyrillic Character Sets Date: 29 Apr 1998 03:44:55 GMT Organization: ILK Internet GmbH, Karlsruhe, BRD Message-ID: <6i67nn$11u@fs1.ilk.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi, we have to port our applications, running under OPENSTEP Enterprise 4.2 for Windows NT, so they support cyrillic Character Sets. The complete app should be in cyrillic, from the UI to the database. The underlying Orace DB suports cyrillic. But what's with EOF, the display of data in AppKit, or the data input. Do we need an InputManager. I have currently no idea, but I have to estimate if it's possible or not. Any ideas? Stathis --- MARCON - Evstathios Marinos Consulting Evstathios Marinos | Phone : +49 721 98 22 21 1 Karlstr. 68 | Fax : +49 721 98 22 21 2 76137 Karlsruhe (GERMANY) | E-Mail: emarinos@marcon.de
From: "Jeremy Bettis" <jeremy@hksys.com> Newsgroups: comp.lang.objective-c,comp.sys.next.programmer Subject: Re: Autorelease pools in OpenStep Date: Wed, 29 Apr 1998 09:18:34 -0500 Organization: Internet Nebraska Message-ID: <6i7crs$kdm$1@owl.inetnebr.com> References: <6i5cfn$8s9$1@vixen.cso.uiuc.edu> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit -----BEGIN PGP SIGNED MESSAGE----- > >From Don Yacktman's article about memory management in OpenStep: > >http://www.stepwise.com/Articles/Technical/HoldMe.html > >The article says that the default autorelease pool is created at the >beginning of and cleared at the end of Openstep's "internal event loop". >When exactly does this happen? In an appKit based program, the main loop looks something like this: while (1) { NSAutoreleasePool *pool = [NSAutoreleasePool new]; [[NSRunLoop currentRunLoop] acceptInputForMode:SomeMode beforeDate:somedate]; [pool release]; } Anything you do as a response to an event or a timer will have its autoreleased objects released when the control returns to the main loop. >I've got an object which houses the main execution loop of my program >(i.e. creating hundreds of smaller objects that are created and >destroyed during this loop, which is repeated several times). I'm >running out of memory and I'm wondering if it's because OpenStep is >not clearing out the autorelease pools. If you have a main execution loop, then clearly the process described above does not apply to you. -----BEGIN PGP SIGNATURE----- Version: PGP 5.5.5 iQBVAwUBNUc2uhmtiimFYrNNAQFaNQIAzRJEectRptPoSMe734dMX6rk34QXPaYc QzU8+yvQDPA7r6b4f8R81DSLfbTGwPzyiiTYYUX4ifOP3qBf9AIcfA== =OSRx -----END PGP SIGNATURE-----
From: boehring@biomed.ruhr-uni-bochum.de (Daniel Boehringer) Newsgroups: comp.sys.next.programmer Subject: Re: pragma for turning off unknown method warnings? Date: 29 Apr 1998 16:13:49 GMT Organization: Ruhr-Universitaet Bochum, Rechenzentrum Message-ID: <6i7jjt$4hv$1@sun579.rz.ruhr-uni-bochum.de> References: <6i2qse$9h6$1@crib.bevc.blacksburg.va.us> <6i4gr7$nc$1@pump1.york.ac.uk> <6i55ar$bp5$1@crib.bevc.blacksburg.va.us> > >> @interface NSObject(MyDelegateMethods) >> - (void)delegateMethod1:(int)x; > >I have some delegate methods which might conflict in name with methods >defined in some subclass of NSObject. (i.e., some Foundation or AppKit >class might define a method named -delegateMethod1:.. though of course >that's not a name I'm really using.) Will putting a category on NSObject >screw up those subclasses since they inherit from it? Should I rename >my delegate methods? no, since what you refer to is called "informal protocol" and has effect only at compile time (unless of course you would supply an implementation for a method of a class which has another implementation residing in a category).
Message-ID: <35473A41.4FC3@his.com> Date: Wed, 29 Apr 1998 14:33:39 +0000 From: Royce Priem <priem@his.com> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Password Protected ROM Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Organization: Heller Information Services Hello all - I'm trying to install an IBM 2 Gig drive into my Cube, the drive having already been built and tested on a Turbo slab. After the self-tests it begins the boot process by throwing out the boot command followed by the device and an interesting add-on: sd(0,0,0)- nbu512 At this point it spits out an error like: sc unexpected message:1 and will continue on this course until I tell it to power off. I was told that potentially the problem could be the add-on of nbu-512 to the boot command. I figured I would remove it, but the ROM monitor won't let me in without a password...talk about frustrating!!! I asked someone how to eliminate the password and was told to short the backup battery leads together with the battery still seated. Hasn''t worked after several attempts...any ideas?!?! Thanks in advance... Regards, Royce
From: John Zollinger <john.zollinger@arkona.com> Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer,comp.sys.next.sysadmin Subject: Dark Forest Date: Wed, 29 Apr 1998 16:03:15 -0600 Organization: Arkona, Inc. Message-ID: <3547A3A3.4BA553E@arkona.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Does anyone know if that most excellent utility DarkForest has been ported to OPENSTEP? I miss it greatly now that I am stuck in NT (course I miss many other things as well <sigh>). If it hasn't been ported yet, is the source available somewhere? I wouldn't mind spending some time working on it. Thanks! John Zollinger Arkona, Inc. john.zollinger@arkona.com
From: Student <sxxxxxx@student.macarthur.uws.edu.au> Newsgroups: comp.sys.next.programmer Subject: (no subject) Date: Thu, 30 Apr 1998 12:03:20 +1000 Organization: University of Western Sydney Macarthur Message-ID: <3547DBE8.6A10@student.macarthur.uws.edu.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I'm after information about developing computer programs. No program inparticular but the process, techniques, development and future devlopments thanks.
Newsgroups: comp.sys.next.programmer From: dfevans@bbcr.uwaterloo.ca (David Evans) Subject: Re: Password Protected ROM Message-ID: <893901748.785061@globe.uwaterloo.ca> Sender: news@watserv3.uwaterloo.ca Organization: University of Waterloo References: <35473A41.4FC3@his.com> Cache-Post-Path: globe.uwaterloo.ca!unknown@bcr11.uwaterloo.ca Date: Thu, 30 Apr 1998 02:05:10 GMT In article <35473A41.4FC3@his.com>, Royce Priem <priem@his.com> wrote: > >I asked someone how to eliminate the password and was told to short the >backup battery leads together with the battery still seated. Hasn''t >worked after several attempts...any ideas?!?! > Man, don't do that! Take the battery out and go for lunch or to sleep or something similar. When you come back the password should be gone. -- David Evans (NeXTMail/MIME OK) dfevans@bbcr.uwaterloo.ca Computer/Synth Junkie http://bbcr.uwaterloo.ca/~dfevans/ University of Waterloo "Default is the value selected by the composer Ontario, Canada overridden by your command." - Roland TR-707 Manual
From: DWells <tator3@iname.com> Newsgroups: comp.sys.mac.programmer.codewarrior,comp.sys.mac.programmer.help,comp.sys.mac.system,comp.sys.mac.wanted,comp.sys.mips,comp.sys.msx,comp.sys.newton.marketplace,comp.sys.newton.misc,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.powerpc.advocacy,comp.sys.psion.marketplace,comp.sys.psion.misc,comp.sys.psion.programmer,comp.sys.sgi.admin,comp.sys.sgi.apps,comp.sys.sgi.graphics Subject: EZ EZ MONEY, WHILE YOU WAIT... Date: Wed, 29 Apr 1998 08:35:09 -0400 Organization: Northeast Regional Data Center Message-ID: <35471E7C.1EEF26B5@iname.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit .........For Your Ship To Come In!!!! This is simple, safe, and it really works! For $6 (US), 6 stamps, and about an hour of your time, you could earn a year's salary in a month. Sounds to good to be true, but just imagine. WHAT IF? A little while back, I was browsing these newsgroups, just like you are now, and came across an article similar to this that said you could make thousands of dollars within weeks with only an initial investment of $6.00! So I thought, "Yeah, right, this must be a scam!" but like most of us, I was curious. Like most of us, I kept reading. Anyway, it said that if you send $1.00 to each of the 6 names and addresses stated in the article, you could make thousands in a very short period of time. You then place your own name and address at the bottom of the list at #6, and post the article to at least 200 newsgroups. (There are about 22,000.) or e-mail them to friends, or e-mailing lists... No catch, that was it. Even though the investment was a measly $6, I had three questions that needed to be answered before I could get involved in this sort of thing. 1. IS THIS REALLY LEGAL? I called a lawyer first. The lawyer was a little skeptical that I would actually make any money but he said it WAS LEGAL if I wanted to try it. I told him it sounded a lot like a chain letter but the details of the system (SEE BELOW) actually made it a legitimate legal business. 2. Would the Post Office be OK with this...? I called them: 1-800-725-2161 and they confirmed THIS IS ABSOLUTELY LEGAL! (See Title 18,h sections 1302 NS 1341 of Postal Lottery Laws). This clarifies the program of collecting names and addresses for a mailing list. 3. Is this moral? Well, everyone who sends me a buck has a good chance of getting A LOT of money ... a much better chance than buying a lottery ticket! So, having these questions answered, I invested EXACTLY $7.92 ... six $1.00 bills and six 32 cent postage stamps ... and boy am I glad I did! Within 7 days, I started getting money in the mail! I was shocked! I still figured it would end soon, and didn't give it another thought. But the money just continued coming in. In my first week, I made about $20.00 to $30.00 dollars. By the end of the second week I had a made total of $1,000.00! In the third week I had over $10,000.00 and it was still growing. This is now my fourth week and I have made a total of just over $42,000.00 and it's still coming in..... It's certainly worth $6.00 and 6 stamps! So now I'm reposting this so I can make even more money! The *ONLY* thing stopping *ANYONE* from enriching their own bank account is pure laziness! It took me all of 5 MINUTES to print this out, follow the directions, and begin posting to newsgroups. It took me a mere 45 minutes to post to over 200 newsgroups. And for this GRAND TOTAL investment of $ 7.92 (US) and under ONE HOUR of my time, I have reaped an incredible amount of money -- like nothing I've ever even heard of anywhere before! 'Nuff said! Let me tell you how this works, and most importantly, why it works. Also, make sure you print a copy of this article now, so you can get the information off of it when you need it. The process is very simple and consists of THREE easy steps. ============ HOW IT WORKS ============ STEP 1: ------ Get 6 separate pieces of paper and write the following on each piece of paper: PLEASE ADD ME TO YOUR MAILING LIST. $1 US DOLLAR PROCESSING FEE IS ENCLOSED. (THIS IS KEY AS THIS IS WHAT MAKES IT LEGAL SINCE YOU ARE PAYING FOR AND LATER OFFERING A SERVICE). Now get 6 $1.00 bills and place ONE inside EACH of the 6 pieces of paper so the bill will not be seen through the envelope to prevent theft/robbery. Then, place one paper in each of the 6 envelopes and seal them. You should now have 6 sealed envelopes, each with a piece of paper stating the above phrase and an U.S. $1.00 bill. Mail the 6 envelopes to the following addresses: #1- Coveleski P.O. Box 803 Goshen, NY 10924 USA #2- Dixon G. 17811 S.W. 137th Court Miami, FL 33177 USA #3- Anders Roth Kuuvuorenkatu 1C 59 FIN-20540 TURKU FINLAND #4- Ron Cooper 177 Cochrane Cres. Ft. McMurray, Alta T9K 1H1 Canada #5- Bruce Pelletier P.O. Box 359 Gray, ME 04039 USA #6- D. Wells 18114 SE CR 234 Micanopy, FL. 32667 USA STEP 2: Now take the #1 name off the list that you see above, move the other names up (6 becomes 5, 5 becomes 4, etc.) and add YOUR Name as number 6 on the list. (If you want to remain anonymous put a nickname, but the address MUST be correct. It, of course, MUST contain your country, state/district/area, zip code, etc! You wouldn't want your money to fly away, would you?). STEP 3: Now post your amended article to at least 200 newsgroups. Remember that 200 postings are just a guideline. The more you post, the more money you make! Don't know HOW to post in the news groups? Well do exactly the following: ------------------------------------------------------------------------ HOW TO POST TO NEWSGROUPS FAST WITH YOUR WEB BROWSER: The fastest way to post a newsletter: Highlight and COPY (Ctrl-C) the text of this posted message and PASTE (Ctrl-V) it into a plain text editor (as Wordpad) and save it. After you have made the necessary changes that are stated above, simply COPY (Ctrl-C) and PASTE (Ctrl-V) the text into the message composition window, after selecting a newsgroup, and post it! (Or you can attach the file, without writing anything to the message window.) ------------------------ ------------------------------------------------------------------------ If you have Netscape Navigator 3.0 do the following: 1. Click on any newsgroup like normal, then click on 'TO NEWS'. This will bring up a box to type a message in. 2. Leave the newsgroup box like it is, change the subject box to something flashy, something to catch the eye, as "$$$ NEED CASH $$$? READ HERE! $! $! $" Or "$$$! MAKE FAST CASH, YOU CAN'T LOSE! $$$". Or you can use my subject title. 3. Now click on 'ATTACHMENTS'. Then click on 'ATTACH FILE'. Find your file on your Hard Disk (the one you saved from the text editor). Once you find it, click on it and then click 'OPEN' and 'OK'. You should now see your file name in the attachments box. 4. Now click on 'SEND'/'POST'. You see? Now you just have 199 to go! (Don't worry, it's easy and quick once you get used to it.) NOTE: All the versions of Netscape Navigator's are similar to each other, so you'll have no problem to do this if you don't have Netscape Navigator 3.0. ------------------------------------------------------------------------ ! QUICK TIP! (For Netscape Navigator 3.x and above) You can post this message to many newsgroups at a time, by simply selecting a newsgroup near the top of the screen, hold down the SHIFT, and then select a newsgroup near the bottom of the screen. All of the newsgroups in/between will be selected. After that, you follow the basic steps, stated below in this letter, except step #1. You can go to the page stated below in this letter and click on a newsgroup to open up the newsgroups window. Once you've done this, in the same window go to 'OPTIONS', and then mark 'SHOW ALL NEWSGROUPS' and 'SHOW ALL MESSAGES'. Now you can see all the newsgroups and you can apply easier the above tip. ------------------------------------------------------------------------ If you have MS Internet Explorer do the following: 1. Go to the newsgroups and press 'POST AN ARTICLE'. To the new window type your headline in the subject area and then click in the large window below. There either PASTE your letter (which it's been copied from the text editor), or attach the file which contains it. 2. Then click on 'SEND' or 'OK'. NOTE: All versions of MS Internet Explorer are similar to each other, so you won't have any problem doing this. GENERAL NOTES ON POSTING: A nice page where you'll find all the newsgroups if you want help is http://www.liszt.com/ (When you go to the home page, click on the link 'Newsgroup Directory'). But I don't think you'll have any problem posting because it's very easy once you've found the newsgroups. All these web browsers are similar. It doesn't matter which one you have. (But it makes it very easy if you have Netscape Navigator 3.0 or later. You may download it from the Internet if you don't have it.) You just have to remember the basic steps, stated below. BASIC STEPS FOR POSTING: 1. Find a newsgroup and you click on it. 2. You click on 'POST AN/NEW ARTICLE' or 'TO NEWS' or anything else similar to these. 3. You type your flashy headline in the subject box. 4. Now, either you attach the file containing your amended letter, or you PASTE the letter. (You have to COPY it from the text editor, of course, from before.) 5. Finally, you click on 'SEND' or 'POST' or 'OK', whatever is there. ------------------------------------------------------------------------ **REMEMBER, THE MORE NEWSGROUPS YOU POST IN, THE MORE MONEY YOU WILL MAKE! BUT YOU HAVE TO POST A MINIMUM OF 200** That's it! You will begin receiving money from around the world within days! You may eventually want to rent a P.O.Box due to the large amount of mail you receive. If you wish to stay anonymous, you can invent a name to use, as long as the postman will deliver it. **JUST MAKE SURE ALL THE ADDRESSES ARE CORRECT. ** ------------------------------------------------------------------------- ================= Now the WHY part: ================= Out of 200 postings; say I receive only 5 replies (a very low example). So then I made $5.00 with my name at #6 on the letter. Now, each of the 5 persons who just sent me $1.00 make the MINIMUM 200 postings, each with my name at #5 and only 5 persons respond to each of the original 5, that is another $25.00 for me, now those 25 each make 200 MINIMUM posts with my name at #4 and only 5 replies each, I will bring in an additional $125.00! Now, those 125 persons turn around and post the MINIMUM 200 with my name at #3 and only receive 5 replies each, I will make an additional $626.00! OK, now here is the fun part, each of those 625 persons post a MINIMUM 200 letters with my name at #2 and they each only receive 5 replies, that just made me $3,125.00! Those 3,125 persons will all deliver this message to 200 newsgroups with my name at #1 and if still 5 persons per 200 newsgroups react I will receive $15,625,00! With an original investment of only $6.00! AMAZING! And as I said 5 responses is actually VERY LOW! Average are probable 20 to 30! So lets put those figures at just 15 responses per person. Here is what you will make: at #6 $15.00 at #5 $225.00 at #4 $3,375.00 at #3 $50,625.00 at #2 $759,375.00 at #1 $11,390,625.00 When your name is no longer on the list, you just take the latest posting in the newsgroups, and send out another $6.00 to names on the list, putting your name at number 6 again. And start posting again. The thing to remember is, do you realize that thousands of people all over the world are joining the internet and reading these articles everyday, JUST LIKE YOU are now! So can you afford $6.00 and see if it really works? I think so... People have said, "what if the plan is played out and no one sends you the money? So what! What are the chances of that happening when there are tons of new honest users and new honest people who are joining the internet and newsgroups everyday and are willing to give it a try? Estimates are at 20,000 to 50,000 new users, every day, with thousands of those joining the actual Internet. Remember, play FAIRLY and HONESTLY and this will work. You just have to be honest. By the way, if you try to deceive people by posting the messages with your name in the list and not sending the money to the rest of the people already on the list, you will NOT get as much. Someone I talked to knew someone who did that and he only made about $150.00, and that's after seven or eight weeks! Then he sent the 6 $1.00 bills, people added him to their lists, and in 4-5 weeks he had over $10k. This is the fairest and most honest way I have ever seen to share the wealth of the world without costing anything but our time! You also may want to buy mailing and e-mail lists for future dollars. Make sure you print this article out RIGHT NOW, also. Try to keep a list of everyone that sends you money and always keep an eye on the newsgroups to make sure everyone is playing fairly. Remember that HONESTY IS THE BEST POLICY. You don't need to cheat the basic idea to make the money! GOOD LUCK to all and please play fairly and reap the huge rewards from this, which is tons of extra CASH. Please remember to declare your extra income. Thanks once again... =========================================================== ========== LEGAL? ? ? (Comments from Bob Novak who started this new version.) "People have asked me if this is really legal. Well, it is! You are using the Internet to advertise your business. What is that business? You are assembling a mailing list of people who are interested in home based computer and online business and methods of generating income at home. Remember that people send you a small fee to be added to your mailing list. It is legal. What will you do with your list of thousands of names? Compile all of them into a database and sell them as "Mailing Lists" on the internet in a similar manner, if you wish, and make more money. How do you think you get all the junk mail that you do? Credit card companies, mail order, Utilities, anyone you deal with through the mail can sell your name and address on a mailing list, unless you ask them not to, in addition to there regular business, So, why not do the same with the list you collect. You can find more info about "Mailing Lists" on the internet using any search engine. ." So, build your mailing list, keep good accounts, declare the income and pay your taxes. By doing this you prove your business intentions. Keep an eye on the newsgroups and when the cash has stopped coming (that means your name is no longer on the list), you just take the latest posting at the newsgroups, send another $6.00 to the names stated on the list, make your corrections (put your name at #6) and start posting again. =========================================================== NOTES: *1. In some countries, the export of the country's exchange is illegal. But you can get the license to do this from the post office, explaining the above statements (that you have an online business, etc. You may have to pay an extra tax, but that's OK, the amount of the incoming money is HUGE! And as I said, a few countries have that restriction. *2. You may want to buy mailing and e-mail lists for future dollars. (Or Database or Spreadsheet software.) *3. If you're really not sure or still think this can't be for real, please print a copy of this article and pass it along to someone who really needs the money, and see what happens. *4. You should start getting responses within 1-2 weeks.
Newsgroups: comp.sys.mac.programmer.codewarrior,comp.sys.mac.programmer.help,comp.sys.mac.system,comp.sys.mac.wanted,comp.sys.mips,comp.sys.msx,comp.sys.newton.marketplace,comp.sys.newton.misc,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.powerpc.advocacy,comp.sys.psion.marketplace,comp.sys.psion.misc,comp.sys.psion.programmer,comp.sys.sgi.admin,comp.sys.sgi.apps,comp.sys.sgi.graphics From: news@news.msfc.nasa.gov Message-ID: <cancel.35471E7C.1EEF26B5@iname.com> Control: cancel <35471E7C.1EEF26B5@iname.com> Subject: cmsg cancel <35471E7C.1EEF26B5@iname.com> Organization: http://www.msfc.nasa.gov/ Date: Thu, 30 Apr 1998 07:07:46 GMT Sender: DWells <tator3@iname.com> Make Money Fast post canceled by J. Porter Clark.
From: Norbert Heger <look@signature.please> Newsgroups: comp.sys.next.programmer Subject: Re: duplicating NSInvocations Date: 30 Apr 1998 13:23:06 GMT Organization: Vienna University of Technology, Austria Message-ID: <6i9tvq$7de$1@news.tuwien.ac.at> References: <6hqk1n$ngh$1@crib.bevc.blacksburg.va.us> Originator: root@cray Nathan Urban wrote: > I'm trying to copy the arguments of one NSInvocation into another, > by iterating through each argument and using the result of > -getArgument:atIndex: on the original invocation to set each > argument of the new one with -setArgument:atIndex:. However, > -getArgument:atIndex: needs to be passed an allocated buffer to > fill with the argument, and I don't know how much space to allocate > because I don't know the size of the argument. The NSInvocation > documentation mentions an -argumentInfoAtIndex: method of > NSMethodSignature which returns an NSArgumentInfo structure, but > that method doesn't actually exist! > So, what do I do? A hack would be just to allocate a "sufficiently > big" buffer, which probably isn't too bad since arguments don't > tend to be too big (however, structures can be arbitrarily large), > but I was hoping for a cleaner solution. (Actually, come to think > of it, I could just malloc the frame length and that would be > sufficient.) The release notes say that NSArgumentInfo used to be > in both NEXTSTEP and OpenStep but was removed from both. (Why?? I > need it!) Is there a nice workaround? Can I obtain the argument > size somehow from the argument type, using NSMethodSignature's > -getArgumentTypeAtIndex:? You might try NSGetSizeAndAlignment (defined in NSObjCRuntime.h). Hope this helps. - N.C. _________________________________________________ Norbert C. Heger, Objective Development Austria NeXT OPENSTEP Rhapsody Software Development e-mail: norbert.heger(at)obdev.at web: http://www.obdev.at
From: bmajik@goliath.unl.edu (Matt Evans) Newsgroups: comp.sys.next.programmer Subject: Re: Password Protected ROM Date: 30 Apr 1998 15:57:19 GMT Organization: University of Nebraska-Lincoln Message-ID: <6ia70v$dh7$1@unlnews.unl.edu> References: <35473A41.4FC3@his.com> <893901748.785061@globe.uwaterloo.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In article <893901748.785061@globe.uwaterloo.ca>, dfevans@bbcr.uwaterloo.ca (David Evans) writes: > In article <35473A41.4FC3@his.com>, Royce Priem <priem@his.com> wrote: >> >>I asked someone how to eliminate the password and was told to short the >>backup battery leads together with the battery still seated. Hasn''t >>worked after several attempts...any ideas?!?! >> > > Man, don't do that! Take the battery out and go for lunch or to sleep or > something similar. When you come back the password should be gone. > here's what i did. these are on original 030 cubes btw. i took out the mb. took out the battery, connected a wire across the battery terminals. held it there for a few seconds. put in the battery, powered up, no go, the pw was still there. so then i took it back apart again, and then took the battery out, and then put the wire across the battery terminals, but then tried to power it up again. the machine wont power up with the leads shorted. i tried 3 times, but it didnt' work. however, when i put the battery back in, there was no pw :) machines work fine. im guessing that any attempt ot power the machine up with the battery running as a short very quickly discharges whatever leakages there are in the NVRAM.
From: "michael marlatt" <mmarlatt@tsi.net> Newsgroups: alt.atlanta,alt.jobs.offered,atl.general,atl.jobs,comp.os.os2.marketplace,comp.os.os2.programmer.misc,comp.os.os2.programmer.oop,comp.sys.next.programmer,comp.sys.next.sysadmin,misc.jobs.contract,misc.jobs.offered,prg.jobs,triangle.jobs,us.mi Subject: Systems Programmer/ Coder Date: Thu, 30 Apr 1998 12:56:18 -0400 Message-ID: <6iaaos$6m0$1@news1-alterdial.uu.net> TECHNISOURCE, Inc. Technisource is an innovative and rapidly growing national contract, consulting and staffing organization. Technisource has a long list of satisfied clients and hundreds of high caliber consultants with rewarding careers in the technical industry. Technisource offers the best solution whether our clients' needs are for one engineer, a team or a complete outsourcing unit. Since 1987 Technisource has specialized in the technical consulting industry and is the most qualified resource to turn for your next project. Over 600 opportunities are listed on our website for engineers, programmers, designers, technicians and technical writers. **************************************************************************** *************** For a continuously updated list of opportunities please visit our Web Page at www.tsi.net **************************************************************************** *************** Requirement Number: AT30040 Job Title: Systems Programmer/ Coder Job Location: Atlanta, GA Area TECHNISOURCE has the following opportunity out of its Atlanta office. Skill Set Required: BSCS, JAM/ JPL Programmer, UNIX desired(not required) Tasks and Responsibilities: JAM/JPL Programmer working in code to do screen designs, perform testing and implementation of the JAM/JPL Coded screens. Experienced programmer in JAM/ JPL programming language to do upgrades and maintenance. Please Reply to: Michael Marlatt Technisource, Inc. Dept. - AT30040 / Michael Marlatt 3355 Lenox Road, #810 Atlanta, GA 30326 USA Phone 404-816-9141 / 800-262-9088, Ext. 112 Fax 404-816-8933 Email: mmarlatt@tsi.net WWW: http://www.tsi.net * WHEN FAXING PLEASE USE THE HIGHEST RESOLUTION AVAILABLE (We Use OCR).* **************************************************************************** *********************************
From: nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) Newsgroups: comp.sys.next.programmer Subject: Anyway to programmatically rebuild the Table of Contents file in Mail.app? Date: 1 May 1998 01:33:22 GMT Organization: none Message-ID: <6ib8p2$k6o$8@ha2.rdc1.nj.home.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit One of my Mailboxes keeps getting a corrupted Table of Contents... I can rebuild it by double clicking on the mailbox, but I'd like to find some tricky way of doing it automatically. TjL -- <a href='http://opaldata.com/the_end/index.html'> And you thought the Internet would never end</a> Unix Tip #4872: giving a process a lower priority gets it done faster SEE ALSO: nice(1), renice(8)
From: spammers@ruin.the.internet.channelu.com Newsgroups: comp.sys.next.programmer Subject: Re: Password Protected ROM Date: 1 May 1998 04:23:57 GMT Organization: Michigan State University Message-ID: <6ibiot$6k0$2@msunews.cl.msu.edu> References: <35473A41.4FC3@his.com> <893901748.785061@globe.uwaterloo.ca> <6ia70v$dh7$1@unlnews.unl.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: bmajik@goliath.unl.edu In <6ia70v$dh7$1@unlnews.unl.edu> Matt Evans wrote: > In article <893901748.785061@globe.uwaterloo.ca>, > dfevans@bbcr.uwaterloo.ca (David Evans) writes: > > In article <35473A41.4FC3@his.com>, Royce Priem <priem@his.com> wrote: > >> > >>I asked someone how to eliminate the password and was told to short the > >>backup battery leads together with the battery still seated. Hasn''t > >>worked after several attempts...any ideas?!?! > >> > > > > Man, don't do that! Take the battery out and go for lunch or to sleep or > > something similar. When you come back the password should be gone. > > > > here's what i did. these are on original 030 cubes btw. > > i took out the mb. took out the battery, connected a wire across the battery > terminals. held it there for a few seconds. > put in the battery, powered up, no go, the pw was still there. so then i > took it back apart again, and then took the battery out, and then put the > wire across the battery terminals, but then tried to power it up again. > > the machine wont power up with the leads shorted. i tried 3 times, but it > didnt' work. however, when i put the battery back in, there was no pw :) > > machines work fine. im guessing that any attempt ot power the machine up with > the battery running as a short very quickly discharges whatever leakages > there are in the NVRAM. > I'm pretty sure I've done this a few times. Remove the MB (or just unplug the system). Remove the battery, and short the leads between the +/- terminals where the battery sat. Walk away for 5-10 minutes. Remove short wire, replace battery, reassemble, reboot and wiped NVRAM. Randy rencsok at channelu dot com argus dot cem dot msu dot edu spammers works also :) Randy Rencsok General UNIX, NeXTStep, IRIX Admining, Turbo Software Consulting, Programming, etc.)
From: kc@ignem.omnigroup.com (Ken Case) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer,comp.sys.next.sysadmin Subject: Re: Dark Forest Followup-To: comp.sys.next.advocacy,comp.sys.next.programmer,comp.sys.next.sysadmin Date: 1 May 1998 18:42:54 GMT Organization: Omni Development, Inc. Message-ID: <6id53e$imu$1@gaea.omnigroup.com> References: <3547A3A3.4BA553E@arkona.com> John Zollinger (john.zollinger@arkona.com) wrote: : Does anyone know if that most excellent utility DarkForest has been : ported to OPENSTEP? I miss it greatly now that I am stuck in NT (course : I miss many other things as well <sigh>). It hasn't, but I wrote a replacement for it (OmniDiskUsage) about 9 months ago which I never quite got around to releasing. If you want it, let me know. -- Ken Case kc@omnigroup.com Omni Development, Inc. http://www.omnigroup.com
From: nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) Newsgroups: comp.sys.next.programmer Subject: OpenStep apps don't recognize System Wide CommandKeys? Date: 1 May 1998 20:29:37 GMT Organization: none Message-ID: <6idbbh$81a$1@ha2.rdc1.nj.home.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit In Preferences.app's ``Menu Preferences'' module I have set several "Command-Key Equivalents". NeXTStep apps seem to accept these, but OpenStep apps seem to ignore them. Am I missing something or is there a way to set Global command keys? (ie I like command ~ for Quit rather than command q [too easy to hit by accident] and command-shift-w for Close Window rather than command w) Thanks TjL -- <a href='http://opaldata.com/the_end/index.html'> And you thought the Internet would never end</a> Unix Tip #4872: giving a process a lower priority gets it done faster SEE ALSO: nice(1), renice(8)
Date: Fri, 01 May 1998 12:18:11 -0700 From: michel@nospamplease.matchfonts.com (Michel) Newsgroups: comp.unix.programmer,comp.sys.next.programmer Subject: Font for programmers / Beta test opportunity Message-ID: <michel-0105981218110001@news.jps.net> Organization: Match Software Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-transfer-encoding: 8bit Dear Unix programmers, I have created a font which has encountered a lot of interest in the programming community, among Windows and Macintosh users. This font has been conceived to help programmers get the best listings, and avoid mistakes common with regular fonts. 1. Fixed width font. 2. Exaggerated distinction between similar characters 3. Heavy emphasis on punctuation characters No confusion between similar characters. Emphasized programming symbols make it easy to check for possible errors, and avoid them. I am sorry, but I have no technical knowledge of GUI systems such as X-Window or Next, apart from having worked now and then on a Sun web server through Netscape to access HTML pages. So I cannot venture into trying to supply fonts without the help of actual users. Fontographer can generate fonts for Sun and Next. I was wondering if some of yours, distinguished pundits, would be interested in helping me release Sun and Next versions of the font ? The font is displayed, and described, at http://www.matchfonts.com/pages/program.html I have very basic questions, such as 1. What is the encoding ? ISO Latin 1, or other ? 2. What are the files needed ? 3. What is the most usual archive compacting system (equivalent to ZIP on PC and Stuffit on the Mac), and do versions exist for Mac or PC, so I can prepare them for distribution ? So, if you are interested into trying out this font, while at the same time helping me come up with a satisfactory product for the Unix and Next communities, I will very much appreciate your help. Michel Bujardet --------------------------------------------------------------------- Display fonts, Foreign fonts, Text fonts, Discovery fonts, Custom ... Check http://www.matchfonts.com Online catalog, online shopping with immediate delivery, and free samples for Windows, Macintosh & OS/2.
From: macghod@concentric.net Newsgroups: comp.sys.next.programmer Subject: gcc Date: 2 May 1998 03:32:32 GMT Message-ID: <6ie44g$7ss$3@newsfep3.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I installed the dev packages, and I dont seem to have gcc? I looked on the cd, and gcc isnt in the biniaries, tho it is in the gnu source? I installed the gnu package, and tried to make gcc from the source, but still no gcc (make didnt compile) -- running on a PC with OPENSTEP, the year 2000 bug-free operating system NeXTMail and MIME OK!
From: nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) Newsgroups: comp.sys.next.programmer Subject: Re: gcc Date: 2 May 1998 04:47:18 GMT Organization: none Message-ID: <6ie8gm$2ak$4@ha2.rdc1.nj.home.com> References: <6ie44g$7ss$3@newsfep3.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: macghod@concentric.net In <6ie44g$7ss$3@newsfep3.sprintmail.com> macghod@concentric.net wrote: > I installed the dev packages, and I dont seem to have gcc? I looked on the > cd, and gcc isnt in the biniaries, tho it is in the gnu source? I installed > the gnu package, and tried to make gcc from the source, but still no gcc > (make didnt compile) /bin/cc is based on gnu cc. You can find a compiled gcc at ftp://next-ftp.peak.org/pub/next/apps/devtools/ TjL -- <a href='http://opaldata.com/the_end/index.html'> And you thought the Internet would never end</a> Unix Tip #4872: giving a process a lower priority gets it done faster SEE ALSO: nice(1), renice(8)
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.unix.programmer,comp.sys.next.programmer Subject: Re: Font for programmers / Beta test opportunity Date: 2 May 1998 09:55:46 GMT Organization: Idiom Communications Message-ID: <6ieqj2$eof$3@news.idiom.com> References: <michel-0105981218110001@news.jps.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: michel@nospamplease.matchfonts.com Michel may or may not have said: -> Dear Unix programmers, -> -> I have created a font which has encountered a lot of interest in the -> programming community, among Windows and Macintosh users. Michel, I looked at the URL you gave, and I have to say I don't like the font. IMHO, it seriously reduced the legibility of the perl listing you gave for an example, since the punctuation stands out so much from the rest of the text. The letterforms themselves look fine, but the punctuation is far too heavy, and it actually makes the sample look like it was printed in multiple fonts. If you reduce the line weight on the punctuation characters, then you might be on to something. Also, I tend to go for the smallest font size I can comfortably read, so that I can get 80 rows in a terminal window on my 21" display. I usually use NeXT's "ohlfs" font, which has a rather nice hand-tuned 10-point bitmap, and which prints as courier. I'd like to see how your font looks at *very* small point sizes. -jcr
From: tom@basil.icce.dev.rug.null.nl (Tom Hageman) Newsgroups: comp.sys.next.programmer Subject: Re: Anyway to programmatically rebuild the Table of Contents file in Mail.app? Date: 2 May 1998 16:43:08 GMT Organization: Warty Wolfs Sender: news@basil.icce.dev.rug.null.nl (NEWS pusher) Message-ID: <EsBuLt.ICx@basil.icce.dev.rug.null.nl> References: <6ib8p2$k6o$8@ha2.rdc1.nj.home.com> nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) wrote: > > One of my Mailboxes keeps getting a corrupted Table of Contents... I can > rebuild it by double clicking on the mailbox, but I'd like to find some > tricky way of doing it automatically. Not that I know of. Sounds like a useful addition to the mailapp-utilities though. Any takers? --Tom. (mailapp-utilities' foster parent) <<SPAMBLOCK: remove dev. and null. to reply>>
From: macghod@concentric.net Newsgroups: comp.sys.next.programmer,comp.sys.next.sysadmin Subject: hotline client? Date: 3 May 1998 03:32:46 GMT Message-ID: <6igogu$p6t$1@newsfep1.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I am interested in seeing what this hotline stuff is about, anyone have a hotline client compiled for openstep 4.2 for mach? -- running on a PC with OPENSTEP, the year 2000 bug-free operating system NeXTMail and MIME OK!
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <13743893563222@digifix.com> Date: 3 May 1998 03:50:06 GMT Organization: Digital Fix Development Message-ID: <14190894168021@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1993. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6ih2fr$aob$6724@trader.ipf.de> Control: cancel <6ih2fr$aob$6724@trader.ipf.de> Date: 03 May 1998 06:23:57 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6ih2fr$aob$6724@trader.ipf.de> Sender: catrina cillone<catf@aimnet.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: tom@basil.icce.dev.rug.null.nl (Tom Hageman) Newsgroups: comp.sys.next.programmer Subject: Re: OpenStep apps don't recognize System Wide CommandKeys? Date: 3 May 1998 16:43:16 GMT Organization: Warty Wolfs Sender: news@basil.icce.dev.rug.null.nl (NEWS pusher) Message-ID: <EsDz6r.J8@basil.icce.dev.rug.null.nl> References: <6idbbh$81a$1@ha2.rdc1.nj.home.com> [posted & mailed] nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) wrote: > > In Preferences.app's ``Menu Preferences'' module I have set several > "Command-Key Equivalents". > > NeXTStep apps seem to accept these, but OpenStep apps seem to ignore them. > > Am I missing something or is there a way to set Global command keys? (ie I > like command ~ for Quit rather than command q [too easy to hit by accident] > and command-shift-w for Close Window rather than command w) Apparently Menu Preferences does not update the OPENSTEP user defaults database. To copy your NEXTSTEP command keys to the OPENSTEP defaults database, you can use the following command: defaults write NSGlobalDomain NSCommandKeys \ "`dread System NXCommandKeys | sed 's/^[^ ]* [^ ]* //'`" Hope this helps, Tom. --- __/__/__/__/ Tom Hageman <tom@basil.icce.dev.rug.null.nl> [NeXTmail/Mime OK] __/ __/_/ IC Group <tom@dev.icgroup.null.nl> (work) __/__/__/ <<SPAMBLOCK: remove dev. and null. to reply>> __/ _/_/ Confused? You won't be after the NeXT episode.
From: Guy Moreillon <moreillon@nagra-kudelski.ch> Newsgroups: comp.sys.next.programmer Subject: EOF: controlling when the commit is done. Date: Mon, 04 May 1998 08:43:56 +0200 Organization: NagraVision Message-ID: <354D63AB.4D7BCB8D@nagra-kudelski.ch> Mime-Version: 1.0 Content-Type: multipart/signed; protocol="application/x-pkcs7-signature"; micalg=sha1; boundary="------------msBD822EBD8599793897D45203" To: ask_next@next.com This is a cryptographically signed message in MIME format. --------------msBD822EBD8599793897D45203 Content-Type: multipart/mixed; boundary="------------0B9385501806651B655B6C25" This is a multi-part message in MIME format. --------------0B9385501806651B655B6C25 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello, I have the following problem with EOF: I would like to do a certain number of operations on an editing context, in a precise order, and have these operations carried out in the DB through EOF in the same order, so as to avoid constraints errors in our Oracle database. Moreover, I would like to call stored procedures within the same transaction. Obviously, the editing context does not remember the ordering of the operations that where performed on itself and that is probably why the databaseContext:willOrderAdaptorOperationsFromDatabaseOperations: and databaseContext:willPerformAdaptorOperations:adaptorChannel: delegate methods are declared in the EODatabaseContext class. Also, it seems the only way to call stored procedures within the same transaction as other modifications is to add the calls manually to the list of adaptor operations in one of the two preceding methods. So my question is, is there a way to control when a transaction is commited by a EODatabaseContext? If I could manually keep a transaction open, then I could make a few changes in the editing context, save them, call some stored procedures, make another few changes, save them, and so on until all the operations have been done and saved and only then commit the whole thing. Is there a way to do this ? If not, what's the proper way to handle this situation ? -- Guy Moreillon -- Software Engineer -- NagraVision, Kudelski SA -- Tel: +41 21 732 04 47 --------------0B9385501806651B655B6C25 Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Guy Moreillon Content-Disposition: attachment; filename="vcard.vcf" begin: vcard fn: Guy Moreillon n: Moreillon;Guy org: NagraVision adr: 4 Rue St Denis;;;Echallens;;1040;Switzerland email;internet: moreillon@nagra-kudelski.ch title: Software Engineer tel;work: +41 21 732 04 47 tel;home: +41 21 881 10 27 x-mozilla-cpt: ;0 x-mozilla-html: TRUE version: 2.1 end: vcard --------------0B9385501806651B655B6C25-- --------------msBD822EBD8599793897D45203 Content-Type: application/x-pkcs7-signature; name="smime.p7s" Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename="smime.p7s" Content-Description: S/MIME Cryptographic Signature MIIKbwYJKoZIhvcNAQcCoIIKYDCCClwCAQExCzAJBgUrDgMCGgUAMAsGCSqGSIb3DQEHAaCC CN0wggQnMIIDkKADAgECAhB++XNKXMyGOI7zTlDakBXSMA0GCSqGSIb3DQEBBAUAMGIxETAP BgNVBAcTCEludGVybmV0MRcwFQYDVQQKEw5WZXJpU2lnbiwgSW5jLjE0MDIGA1UECxMrVmVy aVNpZ24gQ2xhc3MgMSBDQSAtIEluZGl2aWR1YWwgU3Vic2NyaWJlcjAeFw05NzExMjEwMDAw MDBaFw05ODExMjEyMzU5NTlaMIIBIzERMA8GA1UEBxMISW50ZXJuZXQxFzAVBgNVBAoTDlZl cmlTaWduLCBJbmMuMTQwMgYDVQQLEytWZXJpU2lnbiBDbGFzcyAxIENBIC0gSW5kaXZpZHVh bCBTdWJzY3JpYmVyMUYwRAYDVQQLEz13d3cudmVyaXNpZ24uY29tL3JlcG9zaXRvcnkvQ1BT IEluY29ycC4gYnkgUmVmLixMSUFCLkxURChjKTk2MTMwMQYDVQQLEypEaWdpdGFsIElEIENs YXNzIDEgLSBOZXRzY2FwZSBGdWxsIFNlcnZpY2UxFjAUBgNVBAMTDUd1eSBNb3JlaWxsb24x KjAoBgkqhkiG9w0BCQEWG21vcmVpbGxvbkBuYWdyYS1rdWRlbHNraS5jaDBcMA0GCSqGSIb3 DQEBAQUAA0sAMEgCQQCwe0v2X73a36yh95CRUjgCgEpRbsKCQONDJfExB8EFrq6b39kgffn1 EBKd74uGfbidUNrU+IJdx1EF529qWRPPAgMBAAGjggFdMIIBWTAJBgNVHRMEAjAAMIGvBgNV HSAEgacwgDCABgtghkgBhvhFAQcBATCAMCgGCCsGAQUFBwIBFhxodHRwczovL3d3dy52ZXJp c2lnbi5jb20vQ1BTMGIGCCsGAQUFBwICMFYwFRYOVmVyaVNpZ24sIEluYy4wAwIBARo9VmVy aVNpZ24ncyBDUFMgaW5jb3JwLiBieSByZWZlcmVuY2UgbGlhYi4gbHRkLiAoYyk5NyBWZXJp U2lnbgAAAAAAADARBglghkgBhvhCAQEEBAMCB4AwgYYGCmCGSAGG+EUBBgMEeBZ2ZDQ2NTJi ZDYzZjIwNDcwMjkyOTg3NjNjOWQyZjI3NTA2OWM3MzU5YmVkMWIwNTlkYTc1YmM0YmM5NzAx NzQ3ZGE1YzVlOTE0MWJlYWRiMmJkMmU4OTIwNmFjNmJmNWQ3MTE0OTk5YTNiZDQzZjRlNTkw NjU0MTANBgkqhkiG9w0BAQQFAAOBgQA7NOi0/6F8Q/Y/pB4MHEmPeEFcYQ74jtq+4q42Y2uX 56PuJcdIvvG0qrZUN/vS2N0Q/FamnBWpx1/mDHEuidHAVG0KjoR9XtxZnmt3Slw8nYBoR8kS W3oSEyAXgP4WE+rCkHuq0ASb+Io2+qJ7vJXpVIv+4I/PDEZhptxZK0TlazCCAnkwggHioAMC AQICEFIfNR3ycH4AK77KWYcE1TkwDQYJKoZIhvcNAQECBQAwXzELMAkGA1UEBhMCVVMxFzAV BgNVBAoTDlZlcmlTaWduLCBJbmMuMTcwNQYDVQQLEy5DbGFzcyAxIFB1YmxpYyBQcmltYXJ5 IENlcnRpZmljYXRpb24gQXV0aG9yaXR5MB4XDTk2MDYyNzAwMDAwMFoXDTk5MDYyNzIzNTk1 OVowYjERMA8GA1UEBxMISW50ZXJuZXQxFzAVBgNVBAoTDlZlcmlTaWduLCBJbmMuMTQwMgYD VQQLEytWZXJpU2lnbiBDbGFzcyAxIENBIC0gSW5kaXZpZHVhbCBTdWJzY3JpYmVyMIGfMA0G CSqGSIb3DQEBAQUAA4GNADCBiQKBgQC2FKbPTdAFDdjKI9BvqrQpkmOOLPhvltcunXZLEbE2 jVfJw/0cxrr+Hgi6M8qV6r7jW80GqLd5HUQq7XPysVKDaBBwZJHXPmv5912dFEObbpdFmIFH 0S3L3bty10w/cariQPJUObwW7s987LrbP2wqsxaxhhKdrpM01bjV0Pc+qQIDAQABozMwMTAP BgNVHRMECDAGAQH/AgEBMAsGA1UdDwQEAwIBBjARBglghkgBhvhCAQEEBAMCAQYwDQYJKoZI hvcNAQECBQADgYEAwfr3AudXyhF1xpwM+it3T4dFFzvj0sHaD1g5jq6VmQOhqKE4/nmakxcL l4Y5x8poNGa7x4hF9sgMBe6+lyXv4NRu5H+ddlzOfboUoq4Ln/tnW0ilZyWvGWSI9nLYKSeq NxJqsSivJ4MYZWyN7UCeTcR4qIbs6SxQv6b5DduwpkowggIxMIIBmgIFAqQAAAEwDQYJKoZI hvcNAQECBQAwXzELMAkGA1UEBhMCVVMxFzAVBgNVBAoTDlZlcmlTaWduLCBJbmMuMTcwNQYD VQQLEy5DbGFzcyAxIFB1YmxpYyBQcmltYXJ5IENlcnRpZmljYXRpb24gQXV0aG9yaXR5MB4X DTk2MDEyOTAwMDAwMFoXDTk5MTIzMTIzNTk1OVowXzELMAkGA1UEBhMCVVMxFzAVBgNVBAoT DlZlcmlTaWduLCBJbmMuMTcwNQYDVQQLEy5DbGFzcyAxIFB1YmxpYyBQcmltYXJ5IENlcnRp ZmljYXRpb24gQXV0aG9yaXR5MIGfMA0GCSqGSIb3DQEBAQUAA4GNADCBiQKBgQDlGb9to1Zh LZlIcfZn3rmN67eehoAKkQ76OCWvRoiC5XOooJskXQ0fzGVuDLDQVoQYh5oGmxChc9+0WDlr bsH2FdWoqD+qEgaNMax/sDTXjzRniAnNFBHiTkVWaR94AoDa3EeRKbs2yWNcxeDXLYd7obcy sHswuiovMaruo2fa2wIDAQABMA0GCSqGSIb3DQEBAgUAA4GBAFJzuppV3Nw/gn2wkJhiKoJM dgBuJT3VwglwVwEMD3cfGKH7HGAOoHU7SSFB/qdcLUxCSdP/KNiM6p3+yQfid4JTI95V885E k/r6TL3KNvNbZrKeyPIMXl7UobQhCTPKO1n8ksI4/K3ZliTgLfqjKfUzaHhOtLyfaTXiqJiU czvEMYIBWjCCAVYCAQEwdjBiMREwDwYDVQQHEwhJbnRlcm5ldDEXMBUGA1UEChMOVmVyaVNp Z24sIEluYy4xNDAyBgNVBAsTK1ZlcmlTaWduIENsYXNzIDEgQ0EgLSBJbmRpdmlkdWFsIFN1 YnNjcmliZXICEH75c0pczIY4jvNOUNqQFdIwCQYFKw4DAhoFAKB9MBgGCSqGSIb3DQEJAzEL BgkqhkiG9w0BBwEwHAYJKoZIhvcNAQkFMQ8XDTk4MDUwNDA2NDM1NlowHgYJKoZIhvcNAQkP MREwDzANBggqhkiG9w0DAgIBKDAjBgkqhkiG9w0BCQQxFgQUyhMKLjmQ/xjTj2QGFz7J8OSk htYwDQYJKoZIhvcNAQEBBQAEQF9WzgdrNEId56ny7wtxr8i+jC59LfnoxIQI1yDvnT+vT7pz kR36SGTrqO/hAmVlmAiM7+873HuNh4PHkmuTiCo= --------------msBD822EBD8599793897D45203--
From: Larry Blische <larry@lkba.com> Newsgroups: comp.sys.next.programmer Subject: Anyone doing OPENSTEP/WEBOBJECTS work in South East Florida? Date: Mon, 04 May 1998 10:26:22 +0000 Organization: LKB Associates, Inc. Message-ID: <354D89BD.8AF52D0F@lkba.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I'm interrested to know if anyone is doing OPENSTEP/WEBOBJECTS related work in south east Florida. I've done contract programming since 1986 and used NEXTSTEP since 1991. I've just relocated to Florida and was wondering if there are any OPENSTEP or Rhapsody projects underway or contemplated which require developers. Thanks for any responses. -- Larry Blische * Consultant/Programmer Desktop Apps : Client/Server : Embedded Systems : Device Drivers : Etc. 6195 Eagles nest Drive * Jupiter, FL 33458 USA * 561.747.7844 mailto:larry@lkba.com * resume at http://www.charm.net/~lkb
From: press@paragon-software.com Newsgroups: comp.object.corba,comp.sys.next.programmer,comp.lang.objective-c,comp.sys.mac.programmer Subject: Paragon Releases OAK CORBA ORB Version 5.2.1 Date: Mon, 04 May 1998 16:56:56 -0600 Organization: Paragon Software, Inc. Message-ID: <6ildj7$utl$1@nnrp1.dejanews.com> Keywords: OAK, CORBA, IIOP, NeXTSTEP, OPENSTEP, Objective-C To: mahesh@paragon-software.com Paragon Releases OAK CORBA ORB Version 5.2.1 Vienna, Virginia, 4 May 1998 - Paragon Software Inc., a leading provider of distributed computing products and services to corporate customers worldwide, is pleased to announce the release of a new version of its flagship OAK CORBA Object Request Broker. OAK is a fast, highly-portable CORBA 2.0 ORB, unique in the world for its availability with Objective-C language bindings. OAK led the way in bringing the benefits of standards-based, cross-platform, cross-language messaging to specialized environments like NeXTSTEP, OPENSTEP, and WebObjects. Now OAK 5.2.1 makes writing distributed systems even easier with CORBA-compliance and usability enhancements like: - Full Interface and Implementation Repositories; - Naming and Events Common Object Services; - Improved developer documentation including detailed walkthroughs on building systems in Objective-C and C++. OAK 5.2.1 is available today. It ships with C++ bindings on all major platforms, including Windows NT, Solaris, HP-UX, and Linux; and with Objective-C bindings for all Apple Enterprise environments including NeXTSTEP, OPENSTEP, WebObjects, PDO, and Rhapsody. OAK is fully interoperable via the Internet Inter-ORB Protocol (IIOP) with ORBs from leading vendors like Borland/Visigenic, Iona, and Sun. For more information on the OAK ORB, or to download an evaluation copy, please visit Paragon's website at: http://www.paragon-software.com/products/oak/ Paragon Software, Inc. 2136 Gallows Road, Suite G Vienna, VA 22027 USA voice 703-876-1700 fax 703-876-1818 info@paragon-software.com http://www.paragon-software.com -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: Gerben_Wierda@RnA.nl Newsgroups: comp.publish.cdrom.hardware,comp.publish.cdrom.software,comp.sys.next.programmer,comp.sys.next.sysadmin Subject: CDR under NEXTSTEP (a unix). Newbie to CDR has several questions, Date: Mon, 4 May 1998 17:42:57 GMT Organization: R&A Sender: news@RnA.NL Message-ID: <EsG17L.HD@RnA.NL> [repost because of server problems] Hello everybody, before I put my questions, let me tell you I did scan the FAQ and several www-pages. I tried out mkisofs, CDDesigner.app (NEXTSTEP app), but some aspects of CD-recording still escape me. Hence my questions. 1. mkisofs does support rockridge extensions and it runs under NEXTSTEP. But NeXT has made extensions to the RR format. As a result, if I try to make an image through mkisofs from - say - /private/etc, I get name conversions on names like rc.standard.org (two dots) or very long names. But NEXTSTEP itself comes with CD's that handle these names perfectly. Question: has someone built the NEXTSTEP format into mkisofs? What *is* the NEXTSTEP extension to RR? (If I know I might add it to mkisofs myself). 2. Can dd be used to create a raw image under unix? Suppose I mount a CD under NEXTSTEP, can I dd from the 'live partition' to create a raw image that can be burned? I think that should work. After all, a raw image is a file ocntaining a raw image of a file system, and dd does just that. Right? If dd can't do it, how can I create a raw image from a disk. And I mean really raw, not an interpretation through mkisofs. 3. What do NEXTSTEP people think of CDDesigner.app? I have been trying the demo of V1.4 but its workings and operation remain a mystery to me. 4. My CD Writer (Yamaha CDR400t, release 1.0m) has been installed. Reading works. I haven't been able to burn anything (even just trial runs) with nscdwrite I get select errors. (Besides, the "Enter Disk" panel is shown half off screen to the right and if I click the bar it disappears (I run a double-headed cube)). If I get these errors, I cannot eject the disk anymore, whatever I try. (I can prevent Workspace manager from mounting it by running the disk command on the drive). I am fresh out of ideas. How do I burn a raw image on a CD with my Yamaha CDR400t? Thanks, -- Gerben_Wierda@RnA.nl (Gerben Wierda) "If you don't know where you're going, any road will take you there" Paraphrased in Alice in Wonderland, originally from the Talmud. Dass man fuer die Philosophie ein Interesse zeigt, bezeugt noch keine Bereitschaft zum Denken -- Martin Heidegger
From: nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) Newsgroups: comp.sys.next.programmer Subject: CommandKeys in OpenStep -- bug or feature? Date: 5 May 1998 05:43:44 GMT Organization: none Message-ID: <6im8ug$l24$1@ha2.rdc1.nj.home.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit In NeXTStep, you could define 'System NXCommandKeys' dwrites which would be defined for all applications. In addition, you could define specific NXCommandKeys for certain apps. In OpenStep, you can define 'NSGlobalDomain NSCommandKeys' which are defined for all OpenStep apps -- HOWEVER if you define specific NSCommandKeys for certain apps, then those apps will IGNORE the keys defined in 'NSGlobalDomain NSCommandKeys' Is there a reason for this? I consider this a bug because I define certain keys I want to be for all apps, such as remapping Quit, Arrange in Front, Close Window, and Preferences... Now if I add NSCommandKeys for one app I have to also include the entries that would be under 'NSGlobalDomain NSCommandKeys'. NeXTStep allowed for Application-specific NXCommandKeys to override the System NXCommandKeys IF there was a conflict but otherwise the Application would use the System NXCommandkey. OpenStep ignores ALL the system-wide NSCommandKeys if ANY NSCommandKeys are defined for this specific Application. It makes it harder to keep consistency (if I change Quit I now have to make sure I change it in all applications where I had to specially define it) and harder to maintain (I have to manually add them). I call it a bug. Anyone else? TjL -- <a href='http://opaldata.com/the_end/index.html'> And you thought the Internet would never end</a> Unix Tip #4872: giving a process a lower priority gets it done faster SEE ALSO: nice(1), renice(8)
From: "Christopher Erker" <C.Erker@worldnet.att.net> Newsgroups: comp.sys.next.programmer Subject: Re: OpenStep on top of Linux? Date: Tue, 5 May 1998 00:38:24 -0400 Organization: AT&T WorldNet Services Message-ID: <6im4u6$que@bgtnsc03.worldnet.att.net> References: <Pine.LNX.3.95.980420091603.21479B-100000@fabre.act.qc.ca> Well if you are just interested in the OS interface, you could look at a product called AfterStep.. It mimics all the GUI features of NeXTSTEP under Linux... I'm an Win NT guy myself so I don't know too much about the particulars of the product, but the screen shots look AMAZING... Sorry, I don't have a URL, but I found it through an altavista search HTH Andre-John Mas wrote in message ... > >If I understand the OpenStep OS, then it is actualy an OS ontop >of an OS, the current bottom layer OS being Mach. Everyone would >love an OS that is controlled by everyone, the only OS that >currentlt fits this description in Linux. So with the GNUstep >project do you think that it would be possible to create GNUstep >to run on top of Linux and still be compatible with OpenStep. > >I wonder if a version of Rhapsody running on top of MkLinux or >any other version of Linux would be a workable technology. > >AJ >
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.unix.programmer,comp.sys.next.programmer Subject: Re: Font for programmers / Beta test opportunity Organization: Is a sign of weakness Message-ID: <SCOTT.98May4230420@slave.doubleu.com> References: <michel-0105981218110001@news.jps.net> <6ieqj2$eof$3@news.idiom.com> In-reply-to: jcr.remove@this.phrase.idiom.com's message of 2 May 1998 09:55:46 GMT Date: 5 May 1998 02:07:39 -0500 In article <6ieqj2$eof$3@news.idiom.com>, jcr.remove@this.phrase.idiom.com (John C. Randolph) writes: Also, I tend to go for the smallest font size I can comfortably read, so that I can get 80 rows in a terminal window on my 21" display. I usually use NeXT's "ohlfs" font, which has a rather nice hand-tuned 10-point bitmap, and which prints as courier. I'd like to see how your font looks at *very* small point sizes. Heck, with 10-point Olhfs, I'm getting 76 lines on my 17" display at 1152x864. I'd expect to get around 90 lines on a 21" display at 1280x1024. If only there were a 1400x1200 display mode, you could get over 100 lines with about the same DPI as my 17" display... Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: dcl@panix.com (David C. Lambert) Newsgroups: comp.sys.next.programmer Subject: Re: Anyway to programmatically rebuild the Table of Contents file in Mail.app? Date: 5 May 1998 10:52:41 -0400 Organization: PANIX Public Access Internet and Unix, NYC Message-ID: <6in93p$m2f@panix.com> References: <6ib8p2$k6o$8@ha2.rdc1.nj.home.com> <EsBuLt.ICx@basil.icce.dev.rug.null.nl> In <EsBuLt.ICx@basil.icce.dev.rug.null.nl> tom@basil.icce.dev.rug.null.nl (Tom Hageman) writes: >nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) wrote: >> >> One of my Mailboxes keeps getting a corrupted Table of Contents... I can >> rebuild it by double clicking on the mailbox, but I'd like to find some >> tricky way of doing it automatically. >Not that I know of. Sounds like a useful addition to the mailapp-utilities >though. >Any takers? I have such a tool written in python; I'll be happy to send it along once I get permission from the person for whom it was written. - David C. Lambert dcl@panix.com
From: embuck@palmer.cca.rockwell.com (Erik M. Buck) Newsgroups: comp.sys.next.programmer Subject: NeXTStep veterans' SIG at WWDC 98 Date: 5 May 1998 17:06:42 GMT Organization: Rockwell International Message-ID: <6ingv2$i2p1@onews.collins.rockwell.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Are there any plans for a NeXTstep veterans' special interest group meeting at WWDC '98? I doubt there are many of us left. I volunteer my hotel room at the Fairmont as a meeting place.
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.publish.cdrom.hardware,comp.publish.cdrom.software,comp.sys.next.programmer,comp.sys.next.sysadmin Subject: Re: CDR under NEXTSTEP (a unix). Newbie to CDR has several questions, Date: 5 May 98 09:28:50 Organization: Is a sign of weakness Message-ID: <SCOTT.98May5092850@slave.doubleu.com> References: <EsG17L.HD@RnA.NL> In-reply-to: Gerben_Wierda@RnA.nl's message of Mon, 4 May 1998 17:42:57 GMT In article <EsG17L.HD@RnA.NL> Gerben_Wierda@RnA.nl writes: But NEXTSTEP itself comes with CD's that handle these names perfectly. Question: has someone built the NEXTSTEP format into mkisofs? What *is* the NEXTSTEP extension to RR? (If I know I might add it to mkisofs myself). Those CD's are formatted with a raw NeXTSTEP filesystem, not with RockRidge. I don't think they work on other systems at all - unless they somehow wrote multiple tables of contents or something. 2. Can dd be used to create a raw image under unix? Suppose I mount a CD under NEXTSTEP, can I dd from the 'live partition' to create a raw image that can be burned? I think that should work. This _should_ work. I would recommend, though, that you dd from an unmounted partition. Doing it from a mounted partition might make the image look like it wasn't unmounted correctly, which could cause problems with the code that mounts the CD. It obviously won't be able to fsck it... Unfortunately, I have no suggestions as the working your CD-R drive. Last time I had problems, it was due to the media itself. It's useful to have a Windows machine around for this type of thing, so you can check if it works with the included software - if not, you have an obvious course to follow (call for tech support!). I've used AerePerennius.app a couple times to write disks formatted with mkisofs (actually the mkhybrid version to make a RockRidge+Joliet disk). I've noticed a number of oddities regarding the "Insert disk in drive 2" and ejecting. I seem to be able to get the system to a point where I can't eject the disk, yet the system seems to think I need to insert another. This with a Sony CDR-924. I suspect that the SCSI implementation (or perhaps the CD-R extensions to SCSI) aren't terribly standard in some cases. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: NeXTStep veterans' SIG at WWDC 98 Date: 5 May 1998 15:44:44 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6inc5c$c0h$2@ns3.vrx.net> References: <6ingv2$i2p1@onews.collins.rockwell.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: embuck@palmer.cca.rockwell.com In <6ingv2$i2p1@onews.collins.rockwell.com> Erik M. Buck claimed: > Are there any plans for a NeXTstep veterans' special interest group meeting > at WWDC '98? I doubt there are many of us left. Far from it, it seems they're all coming back. I'm no veteran, but my boss is. Can I come anyway? Maury
From: embuck@palmer.cca.rockwell.com (Erik M. Buck) Newsgroups: comp.sys.next.programmer Subject: Re: NeXTStep veterans' SIG at WWDC 98 Date: 5 May 1998 21:44:02 GMT Organization: Rockwell International Message-ID: <6io172$i2j1@onews.collins.rockwell.com> References: <6ingv2$i2p1@onews.collins.rockwell.com> <6inc5c$c0h$2@ns3.vrx.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: maury@remove_this.istar.ca In <6inc5c$c0h$2@ns3.vrx.net> Maury Markowitz wrote: > In <6ingv2$i2p1@onews.collins.rockwell.com> Erik M. Buck claimed: > > Are there any plans for a NeXTstep veterans' special interest group meeting > > at WWDC '98? I doubt there are many of us left. > > Far from it, it seems they're all coming back. I'm no veteran, but my boss > is. Can I come anyway? > > Maury > > As far as I am concerned, all are welcome. If we get more than 20 or so people, we will need a bigger room. Even 20 will be very cramped.
Message-ID: <354F6E45.B78399C0@CyclicSYN.Dimensional.COM> From: KLD Van Horn <cyclic@CyclicSYN.Dimensional.COM> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: NX_WINRESIZED Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Tue, 05 May 1998 19:51:31 GMT NNTP-Posting-Date: Tue, 05 May 1998 13:51:31 MDT Organization: Dimensional Communications The NeXTStep documentation "Concepts/Pre3.0_Concepts/05_Events.rtfd describes: Kit-Defined Events For a kit-defined event, data.compound.subtype identifies the kit-defined subevent. The constants corresponding to these subevents are: Constant Subevent Type NX_WINMOVED Window-moved NX_WINRESIZED Window-resized NX_WINEXPOSED Window-exposed NX_APPACT Application-activated NX_APPDEACT Application-deactivate A window-resized subevent's location field contains the window's new location in screen coordinates. data.compound.misc.L[0] contains the window's new width, and data.compound.misc.L[1] contains the window's new height. However, during compilation NX_WINRESIZED is reported as an undefined symbol . All other kit-defined subevents are defined in appkit/Application.h: /* KITDEFINED subtypes */ #define NX_WINEXPOSED 0 #define NX_APPACT 1 #define NX_APPDEACT 2 #define NX_WINMOVED 4 #define NX_SCREENCHANGED 8 /* SYSDEFINED subtypes */ #define NX_POWEROFF 1 /* Additional flags */ etc. How can I intercept and process NX_WINRESIZED subevents without using the AppKit ? Thanks in advance for your efforts ! -Leif
From: mark@sapphire.oaai.com (Mark Onyschuk) Newsgroups: comp.sys.next.programmer Subject: Re: NeXTStep veterans' SIG at WWDC 98 Date: 5 May 1998 22:04:11 GMT Organization: M. Onyschuk and Associates Inc. Message-ID: <slrn6kvkr4.kam.mark@sapphire.oaai.com> References: <6ingv2$i2p1@onews.collins.rockwell.com> <6inc5c$c0h$2@ns3.vrx.net> <6io172$i2j1@onews.collins.rockwell.com> In article <6io172$i2j1@onews.collins.rockwell.com>, Erik M. Buck wrote: >In <6inc5c$c0h$2@ns3.vrx.net> Maury Markowitz wrote: >> In <6ingv2$i2p1@onews.collins.rockwell.com> Erik M. Buck claimed: >> > Are there any plans for a NeXTstep veterans' special interest group >meeting >> > at WWDC '98? I doubt there are many of us left. >> >> Far from it, it seems they're all coming back. I'm no veteran, but my >boss >> is. Can I come anyway? >> >> Maury >> >> >As far as I am concerned, all are welcome. If we get more than 20 or so >people, we will need a bigger room. Even 20 will be very cramped. > [Maury's boss replies] Well, that's one of us anyway. I'm pretty excited, the last time I attended one of these was NeXTworld - and I missed the Stone rave no less :-). If we run out of space, we'll just have to rent out a pub or something. Mark
From: sal@panix3.panix.com (Salvatore Denaro) Newsgroups: comp.sys.next.programmer Subject: Re: OpenStep on top of Linux? Date: 6 May 1998 05:34:24 GMT Organization: PANIX Public Access Internet and Unix, NYC Message-ID: <slrn6kvtj0.3mv.sal@panix3.panix.com> References: <Pine.LNX.3.95.980420091603.21479B-100000@fabre.act.qc.ca> <6im4u6$que@bgtnsc03.worldnet.att.net> On Tue, 5 May 1998 00:38:24 -0400, Christopher Erker <C.Erker@worldnet.att.net> wrote: >Well if you are just interested in the OS interface, you could look at a >product called AfterStep.. >It mimics all the GUI features of NeXTSTEP under Linux... AfterStep *looks* like OpenStep, it doesn't turn X11 into OpenStep. You don't get services, Drag&Drop that _just works_, Interface builder, DPS, Terminal.App, the browser or any of the other things you expect in OpenStep/Mach. >I'm an Win NT guy myself so I don't know too much about the particulars of >the product, but the screen shots look AMAZING... Look != Feel. If I put a cardboard replica of a Porsche 911 over the body of a '88 Ford escort, it wont make the Ford perform any better. -- sal@panix.com Salvatore Denaro "The reality of the software business today is that if you find something that can make you ridiculously rich, then that's something that Microsoft is going to take away from you." -- Max Metral
From: ahoesch@smartsoft.de Newsgroups: comp.sys.next.programmer Subject: SmartBase 1.3 for OPENSTEP Mach Date: 6 May 1998 15:37:35 GMT Organization: Offenes Netz Luebeck e.V. Message-ID: <6iq03v$1d4@merkur.on.on-luebeck.de> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit SUBMISSION SmartBase 1.3 for OPENSTEP ftp.smartsoft.de/SmartBase/SmartBase.1.3.I.b.tar.gz SmartBase has been developed to simplify the development and maintenance of database dependent software systems based on OPENSTEP, Apples new next generation operating system. This goal is achieved by centralizing the business logic for more flexibility, by providing support for remote object instanciation in order to support the design of client/server architectures and by introducing an advanced object paradigm allowing to create data models at a very high abstraction level. Š Centralized business logic Š Remote object support Š Advanced object paradigm SmartBase does not provide data storage mechanisms on its own but makes use of an ordinary relational database management system for this purpose. You can think of SmartBase as an objective extension for your database. Because SmartBase is based on Apples Enterprise Objects Framework it principly can be used with any RDBMS, that's supported by EOF or comes with an EOF-Adapter. SmartBase is built for Intel. You need to have OPENSTEP Mach and EOF installed to use this product. Moreover you need a RDBMS. The current release only supports OpenBase (Version 5.1.8 or higher). See http://www.openbase.com to get a demo version of OpenBase for Mach. RELEASE NOTES Changes for version 1.3 Release ---------------------------------------------- - OBModeler has been redesigned to support cut & paste, modification of classes and attributes after their creation and comments for classes and attributes. Note that the obmodel file format has been changed. You can read old models with the new OBModeler but not vice versa. - Many minor bugs have been fixed. Changes for version 1.2 Release ---------------------------------------------- - Access Privileges can now correctly be set also for complex attributes (including attributes of type DATA). The responsible bug has been fixed. - Some more minor bugs have been fixed. Changes for version 1.1 Release ---------------------------------------------- - Only classbundles that have been changed since the last time the application has run are downloaded from the server again. This reduces network traffic and allows SmartBase to be used with very slow connections too. - Objects cannot only be instanciated on client computers but also on the server now - Some minor bugs have been fixed - Changes have been made to the framework libraries in order to realize the remote object mechanism. - The SmartBase demos have been updated to reflect the framework changes. INSTALLATION Simply untar SmartBase.1.3.I.tar.gz and install all five packages on your workstation. Read the documentation in /LocalLibrary/Documentation/SmartBase for further details. If you encounter any problems or have suggestions regarding the further development of SmartBase, do not hesitate to contact us. We look forward to hear from you. For further information, contact: Smartsoft Development Email: info@smartsoft.de Special thanks to Scott Keith from OpenBase Inc. for his kind support!!!
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: NeXTStep veterans' SIG at WWDC 98 Date: 6 May 1998 09:05:22 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6ip94i$oum$13@ns3.vrx.net> References: <6ingv2$i2p1@onews.collins.rockwell.com> <6inc5c$c0h$2@ns3.vrx.net> <6io172$i2j1@onews.collins.rockwell.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: embuck@palmer.cca.rockwell.com In <6io172$i2j1@onews.collins.rockwell.com> Erik M. Buck claimed: > As far as I am concerned, all are welcome. If we get more than 20 or so > people, we will need a bigger room. Even 20 will be very cramped. Great! Well put me down for a five-ten minute demo of CyberDog, and if possible another for our app? Maury
From: Mike Talvensaari <mtalvens@adobe.com> Newsgroups: comp.sys.next.programmer Subject: Link errors in Project Builder for WebObjects Framework Date: Wed, 06 May 1998 14:32:17 -0700 Organization: Adobe Systems, Inc. Message-ID: <3550D6E1.55CD@adobe.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I'm having a problem with completing compiles in Project Builder. When I try to build my WebObjectsFramework project I get the following error relating to the .def file, and the build succeeds, but the framework.dll won't load into the application when I run it. Building... Generating SAFE.def....d:/NeXT/NextDeveloper/Executables/Utilities/egrep: permission denied done Then, if I rebuild, I get this permission error and the build fails. Installing product... d:/NeXT/NextDeveloper/Executables/Utilities/chmod -R +w d:\LocalDeveloper/Frameworks/SAFE.framework d:/NeXT/NextDeveloper/Executables/Utilities/rm -rf d:\LocalDeveloper/Frameworks/SAFE.framework (cd C:/Netscape/Server/docs/WebObjects/SAFE && d:/NeXT/NextDeveloper/Executables/Utilities/tar cf - SAFE.framework) | (cd d:\LocalDeveloper/Frameworks && d:/NeXT/NextDeveloper/Executables/Utilities/tar xf -) d:/NeXT/NextDeveloper/Executables/Utilities/tar: permission denied : Invalid argument d:/NeXT/NextDeveloper/Executables/Utilities/tar: can't write to -: make: *** [install-products] Error 2 Currently, I'm working around this by running a script that deletes my .def and changes all my permissions (chmod -R 775) on my project and LocalDeveloper directories. Once I run the script the build succeeds, most of the time. We have one other person on our site that is experiencing the same problem, but other developers here are having no problems. We have not been able to trace the problem to some sort of setup issue. Any advice or help would be appreciated. Mike ---- Mike Talvensaari Adobe Systems, Inc.
From: finton@cs.wisc.edu Newsgroups: comp.sys.next.programmer Subject: Trying to teach the old SavePanel new tricks Date: 6 May 1998 22:27:18 GMT Organization: University of Wisconsin, Madison Message-ID: <6iqo46$mu8$2@news.doit.wisc.edu> I'm trying to get a save panel to allow the user to select directories only. (This is for an application where the user has multiple files to save to a directory). I can use delegate methods to cause a SavePanel to only display and accept directories, but the panel won't let me hit "OK" until I type a filename, which is wrong for this application. Am I missing something? My other alternative appears to be to use an OpenPanel and reset the title and prompt, and tell it to allow choosing directories but not files, and disallow multiple selections. But this solution doesn't allow the user to specify a new directory to create; it only allows selection of existing directories. And on NT / Yellowbox, the panel's button says "Open," which is confusing, to say the least. So neither the stock OpenPanel nor the SavePanel are exactly what I want. Before I think about writing my own panel with NSBrowser, does anyone know of existing code which would do what I want to do? Thanks, David Finton
From: macghod@concentric.net Newsgroups: comp.sys.next.programmer,comp.sys.next.sysadmin,comp.sys.next.marketplace Subject: WTB appkit Date: 7 May 1998 21:06:18 GMT Message-ID: <6it7oa$ceg$2@newsfep4.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit On Apples web site their is a pdf of the appkit. This is 1600 pages, and since I prefer hard copy, it is too big to print out. ANyone have this for sale? I also posted this to the non marketplace groups, because I would also like to know if I can buy this from apple or anywhere else? -- Running on Openstep 4.2, the best os in the world, and about to get better! Mac/openstep qa work desired, email me for resume NeXTMail and MIME OK!
From: kiss@slip.NOSPAM.net (Richard Kiss) Newsgroups: comp.sys.next.programmer Subject: Questions: NSNotificationCenter, NSBrowser Date: Fri, 08 May 1998 00:35:39 -0800 Organization: Slip.Net (http://www.slip.net) Message-ID: <kiss-0805980035400001@ip112.santa-clara6.ca.pub-ip.psi.net> I'm working on an GUI app on OS4.2 and have some questions that are hopefully quite easy to seasoned vetrans. I apologize if I'm a bit confused in some of these questions -- it's been a couple months since I touched PB, in part because of these dead ends. Feel free to respond to just a subset of these questions. Please post. NSNotificationCenter (1) Judging from the headers, NSNotificationCenter appears to be thread safe. When a NSNotification is sent to the sender, does it block until it is broadcast to all the listeners, or just the listeners in this thread, or is it non-blocking? How does it get the NSNotification to objects in other threads? Through an NSConnection? (2) How is the performance of the NSNotificationCenter? Is it feasable to broadcast lots of messages, sometimes up to several a second? (3) I'd like to make the app support a plug-in architecture. Instead of publishing a large protocol of methods that the plug-in objects can implement, I'm considering just allowing the plug-ins to register for the messages their objects are interested in. I'm currently broadcasting the messages myself, but it seems that NSNotificationCenter essentially does just what I need. Is there any serious disadvantage to this? Is it efficient, considering the overhead of packaging up the NSNotification versus the optimization that NSNotificationCenter has presumably undergone? Obviously, these questions are all related. NSBrowser (4) I have an NSBrowser view (all leaf nodes, so it's essentially a list) that contains objects, described in text. This is easy. How do I add mixed text/graphics? Some of the objects represent files, so I'd to display icons. (5) What about drag & drop of the icons? This is a vague question, so perhaps just a pointer to the correct object and interesting methods in the Application Framework docs. (6) The objects in the browser are not necessarily homogenous. The browser is embedded in a panel with verb-labelled buttons. Not all verbs are applicable to all objects. I want it so clicking the button sends the selected object(s) a message. I also want it so that only applicable verb buttons are clickable. Help! (7) I don't really understand the "multiple selection/dotted-outline selection" bit of NSBrowsers. How do I get the selected set, versus the "dotted-outline" selection. I want to turn off the dotted-outline selection, since it's visually confusing, and only retain the highlight multiple selection. Can I do this? How? (8) I found tables very simple, but browsers are, by comparison, a bit confusing. Since it's a list, perhaps I should be using a single-column table instead. Comments? Could I get the icons here? Misc (9) Can anyone point me to some sample "Inspector panel" code? Thanks very much for your help! -- Richard
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: Network objects fail with only light usage. Date: Fri, 08 May 1998 10:44:11 +0200 Organization: Square B.V. Message-ID: <3552C5DB.587642A0@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I'm currently investigating if it would make sense for us to enhance our applications with distributed objects. The documentation makes it seem real easy but in reality network objects are complex. I have written a very simple daemon which accepts connections from clients. Those clients then are queried for a dictionary which is put in an array. The client then asks for all the dictionaries the server has. If the client exits (breaks the connection) the server deletes the dictionary. The server and clients don't give themselves directly over the network but NSProtocolCheckers as a 'firewall'. To test it I start four clients which connect and then quit, disconnecting. The clients are in a batch file that runs around. For about five minutes everything is fine, but then they can't get a connection to the proxy anymore. At first this is only a few times, but as we progress it gets more difficult for a client to get the proxy. The connection never fails. I think this is not even a heavy stress test, although in real life there won't be as many connects/disconnects. We expect only five to ten each day, but I don't know what happens if the server has been up for a month! Does anybody know why the proxy becomes unreachable after so many connects? Maybe it matters, I'm developing under Windows NT. Regards, Maurice le Rutte. ------------------------------------------------------------------- part of function that registers the server: NSConnection *connection; NSProtocolChecker *protocolChecker; /* * Put an extra barrier between us and the mean outside world. * This will force them to behave if they call us. */ protocolChecker = [ [NSProtocolChecker alloc] initWithTarget: self protocol: @protocol(SQCentralServerProtocol)]; [self registerNotifications]; connection = [NSConnection defaultConnection]; [connection registerName: [self registeredName]]; [connection setRootObject: protocolChecker]; The method -serverProxy tries to connect to the proxy if the instance variable serverProxy is nil. Method of client that gets the proxy: - (id <SQCentralServerProtocol>)serverProxy { if(serverProxy == nil) { /* * Get the connection to the central server. */ connection = [ [NSConnection connectionWithRegisteredName: @"Square Central Server" host: nil] retain]; if(connection == nil) fprintf(stderr,"No Connection!\n"); /* * Get a proxy of the server. */ serverProxy = [ [connection rootProxy] retain]; if(serverProxy == nil) fprintf(stderr,"No Proxy\n"); /* * Set the protocol for the proxy to reduce method call overhead. */ [(NSDistantObject *)serverProxy setProtocolForProxy: @protocol(SQCentralServerProtocol)]; } return serverProxy; } -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto:mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+ Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" and "to trick into believing or accepting as genuine something false and often preposterous"
From: sal@panix3.panix.com (Salvatore Denaro) Newsgroups: comp.sys.next.programmer,comp.sys.next.sysadmin,comp.sys.next.marketplace Subject: Re: WTB appkit Date: 8 May 1998 20:31:37 GMT Organization: PANIX Public Access Internet and Unix, NYC Message-ID: <slrn6l6qt9.alg.sal@panix3.panix.com> References: <6it7oa$ceg$2@newsfep4.sprintmail.com> In article <6it7oa$ceg$2@newsfep4.sprintmail.com>, you wrote: >On Apples web site their is a pdf of the appkit. This is 1600 pages, and >since I prefer hard copy, it is too big to print out. ANyone have this for >sale? I also posted this to the non marketplace groups, because I would also >like to know if I can buy this from apple or anywhere else? Go to Amazon.com and look up "Apple Rhapsody" There is a book by Jesse Feiler that will pop up. Also on that page (IIRC) are links to some OpenStep books you might want to look up. The Feiler book is almost a re-print of the Apple dev notes. It isn't a full 1600+ page ref guide, but it is (IMHO) good enough for use as a quick ref guide. If you *MUST* have a printout, go down to Kinko's or some other print shop. many/most will bulk brint PDFs for you. My local printshop will even bind it for me for a fee. If you do this, do yourself a favor and print it in 400 page chunks with a full TOC in each chunk. Also, bring *two* Zip disks. One in Mac and one in PC format. Hope this helps. -- sal@panix.com Salvatore Denaro "The reality of the software business today is that if you find something that can make you ridiculously rich, then that's something that Microsoft is going to take away from you." -- Max Metral
From: rafkah@aol.com (Rafkah) Newsgroups: comp.sys.next.programmer Subject: what exactly is Smalltalk ? Message-ID: <1998050822010800.SAA10901@ladder01.news.aol.com> Date: 08 May 1998 22:01:08 GMT Organization: AOL, http://www.aol.fr I've been talking about Smalltalk a lot. what is it exactly ? what are his advantages against oc, c++ or java ? - Raphie rafkah@aol.com "Enjoy the life when Bill Gates is not President Of The World yet"
From: mpaque@wco.com (Mike Paquette) Newsgroups: comp.sys.next.programmer Subject: Re: Questions: NSNotificationCenter, NSBrowser Date: Fri, 8 May 1998 16:18:52 -0700 Organization: Electronics Service Unit No. 16 Message-ID: <1d8plt6.11oa4aslb163gN@carina55.wco.com> References: <kiss-0805980035400001@ip112.santa-clara6.ca.pub-ip.psi.net> Richard Kiss <kiss@slip.NOSPAM.net> wrote: > (1) Judging from the headers, NSNotificationCenter appears to be thread > safe. When a NSNotification is sent to the sender, does it block until it > is broadcast to all the listeners, or just the listeners in this thread, > or is it non-blocking? How does it get the NSNotification to objects in > other threads? Through an NSConnection? All of the above! There's a default NSNotificationCenter per app, and a default NSNotificationQueue per thread. The queue is what pushes notifications our for a thread. When your app posts a notification to a queue, you can choose synchronous delivery with the 'NSPostNow' postingStyle. Asynchronous delivery can be done using 'NSPostASAP', which posts as soon as control gets back to the run loop, or with 'NSPostWhenIdle', which posts when the run loop is idle. In all cases coalescing of identical notifications is done before dispatch, unless a coalesce mask is used in the call to NSNotificationQueue. > (2) How is the performance of the NSNotificationCenter? Is it feasable to > broadcast lots of messages, sometimes up to several a second? This shouldn't be a problem. The NEXTIME framework used NSNotification heavily to keep it's UI elements synched up. The NSMovieController (the thing with the play/pause button, position slider, and stepping buttons) both posted and kept it's state up to date using notifications. Slider position was updated at up to 4 notifications/second, with notification overhead at a small fraction of a percent of total time. The NEXTIME player app used notifications to drive it's inspector, including notification of changed key window (movie to be inspected), movie play/pause state, and current position (by time and frame). > (3) I'd like to make the app support a plug-in architecture. Instead of > publishing a large protocol of methods that the plug-in objects can > implement, I'm considering just allowing the plug-ins to register for the > messages their objects are interested in. I'm currently broadcasting the > messages myself, but it seems that NSNotificationCenter essentially does > just what I need. Is there any serious disadvantage to this? Is it > efficient, considering the overhead of packaging up the NSNotification > versus the optimization that NSNotificationCenter has presumably > undergone? Obviously, these questions are all related. Notifications work well for lazy binding such as you describe. I wouldn't use it for thousands of messages/sec, but for the applications I've described (up to maybe 10 messages/sec) it works well with minimal overhead. -- Mike Paquette mpaque@wco.com
From: macghod@concentric.net Newsgroups: comp.sys.next.programmer,comp.sys.next.advocacy Subject: whats up with zoneraiders?!?/! Date: 8 May 1998 23:28:19 GMT Message-ID: <6j04ej$nvc$1@newsfep4.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Zoneraiders is out, and it is for rhapsody for ppc ONLY. I thought something written for rhapsody would work on BOTH rhapsodies? And openstep as well? Why in gods name would zoneraiders be ppc only? Also, I am starting to learn programming in openstep. So I make a simple app, lets say currencyconverter. What would I do to get this to work on rhapsody for ppc and intel? A while ago on a thread on the developer program, I said I thought programmers would need rhapsody to develop for rhapsody, and the openstep programmers wondered why everyone would think apps have to be written differently to work on rhapsody. Do they or dont they? -- Running on Openstep 4.2, the best os in the world, and about to get better! Mac/openstep qa work desired, email me for resume NeXTMail and MIME OK!
From: Arnaud <debayeux@easynet.fr> Newsgroups: comp.sys.next.programmer Subject: How to get multiple views with tabs in OS 4.2 ? Date: Sat, 09 May 1998 02:29:25 +0200 Organization: [posted via] Easynet France Message-ID: <3553A365.D013CEF7@easynet.fr> Mime-Version: 1.0 Content-Type: multipart/alternative; boundary="------------83E086C8E0C05D28C44660F1" --------------83E086C8E0C05D28C44660F1 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello, In Interface Builder, and in other programs, I have seen the following interface : [Image] Each tab (Instances, Classes, Sounds, Images) selects a specific view. The tiff files for the graphics effectively exists in the Interface Builder resources. How do you program this ? Why isn't it in the palette window of Interface Builder ? I have seen it in Nextstep 3.3 screenshots. Does it exist in 3.3 and is it hidden in 4.2 ? Is there something like a MiscKit of palettes available for 4.2 ? Thanks for your help From a novice Openstep developer... --------------83E086C8E0C05D28C44660F1 Content-Type: multipart/related; boundary="------------05C4ECC0D9E268701256003F" --------------05C4ECC0D9E268701256003F Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit <HTML> Hello, <P>In Interface Builder, and in other programs, I have seen the following interface : <P><IMG SRC="cid:part1.3553A365.A16C0495@easynet.fr" HEIGHT=240 WIDTH=356> <P>Each tab (Instances, Classes, Sounds, Images) selects a specific view. <BR>The tiff files for the graphics effectively exists in the Interface Builder resources. <BR>How do you program this ? <BR>Why isn't it in the palette window of Interface Builder ? <BR>I have seen it in Nextstep 3.3 screenshots. Does it exist in 3.3 and is it hidden in 4.2 ? <BR>Is there something like a MiscKit of palettes available for 4.2 ? <P>Thanks for your help <P>From a novice Openstep developer...</HTML> --------------05C4ECC0D9E268701256003F Content-Type: image/gif Content-ID: <part1.3553A365.A16C0495@easynet.fr> Content-Transfer-Encoding: base64 Content-Disposition: inline; filename="C:\WINDOWS\TEMP\nsmailF9.gif" R0lGODlhZAHwAPcAAAAAAAAQAAAQEBAAABAAEBAQEBAQIRAhIRgYGCEQECEQISEQMSEYISEh ACEhECEhISEhMSExMSkpKSkpMTEQEDEhADEhEDEhMTEhQjEpKTExMTExQjFCUjFCYzkYCDkY GDkpCEIhEEcxBUIxGDkhIUohIUoxIUI5KVIxEFJCEGMxEGNCEHMxEHNCEHNCIYRCEEI5MUpC MVI5MVJCMTk5OUIxQkJCQkIxUkJCUlIxUlJCUkIxY0JCY1IxY0Ixc1Ixc0JKMUJSUkpKSlpS OVpSQmNKQmNSQmNaSnNSEHNSIX9WFIxSGIRSIZRSIWtSQmtaQlJSUlpSUlJSWlpSWkJSY1JC Y1JSY2NSY0JCc0JSc1JCc1JSc2NSc3NSc1JChGNShF5aTnNjQo5jG3toT4RjUoRzUntaWoRz WmNjY2NrY2tja2Njc3Nrc2NjhHNjhHNjlG1zbXOEc3NziIR7gznBUnuEe4SJiYSUlIxzWpSE WpSEY5CIhIyMjJSIkJSUnJSllKVSEJxaKaVjEKVjIbVjIaVjMcZjMaVjQqVzEKVzIbVzIcZz IaVzMbVzMcZzMbVzQrWEMcaEMZx/Y8ZzQrp4V8aEQsaUQsaUUtZzQtaEQtaEUtaUQtaUUtaU Y+eEUueUUuelUtalY+eUY+elY/+UY/+lY6WEa6WMa62Ua62cc+eUc+elc+e1c/+Uc/+lc/+1 c62ce7WUe7Wce72le72lhMachMalhMathOelhP+lhP+1hL2ljM6tjNa9jMbGlNa9lN61lN69 lP+llM6tnNa9nNbGnJSlpaWUpaWlpaWtpaW1pa2tpbWlpc7Gpda1paWtra2lra2trb21rdbG re/GraWltaW1taXGtbWltbW1tbXGtd69tefWtf/Wtf/ntbW1vdbGve/evbW1xrXGxrXWxsa1 xsbGxsbWxs7OxufGxufWxv/Gxv/WxtbWzsbG1sbW1sbn1tbG1tbW1tbn1t7n1s7e3ufn3tbW 59bn59b/597n5+fW5+fn5/f39+fn/////ywAAAAAZAHwAAAI/gABCBxIsKDBgwgTKlzIsKHD hxAjSpxIsaLFixgzajSIrKPHjyBDihxJsqTJkyhTqlzJsqXLlzBjypxJs6ZHgcg26tzJs6fP n0CDCh0KBefQo0iTKl3KtKnOogByAvhHtapVq1Cdat3KtavXnlfDVoUKVeq/k1m/ql3Ltu3X syazlhV4ViGytG7z6t3LN2NdKCGL3h04d6pUhIP7KjZIdWDjqf8cRxY7uXJYyAUvQ666uDPB s0UBewydGEDhuglLe1b8GPNmupElf44Ne3Zm2rJdr2bdkfRg0XKNGo4qMire3XpbP+asu3lr 57ib156OfC/c375Ln8450nj1vspj/jeubDs3denPqaf/zvZ6YtG9CQuv2101+7bha5M3D728 ZKvqRXefWu6R9VFwUdHFXXH2DUggbcuJ99yEuFF4G2MQCuhgV9eFlh2CZi0YkncbtpfhdMyd ZyF/0um2XolbweXbjPIlOFx9x8HI1Xgnqljhj/6dh6KGOjoFmmkHIgmignblWKSRAAYY5Iqw AajZVU++lRN8o5Fo2nxoZSnmmAO6N9KSU0Gh5ppsrknmm3B29k+bdKpZo1ROxqnnnnzleVBh fAYqKIyADmroobuVVeeijDbq6KOQRirppJRWaumlmGaq6aacdqopoqCGmqinpJZq6qmopqrq qqy2Spar/rDGKuustNZqK6YC3arrrrz26uuvjuYK7LDEFmvssZcKi+yyzDbrrK7KPivttNRW i6tpkSIjD2Xcduvtt+BeJQ8ydTQKRx82pYsSH2xAQQMNi56r7rwksesuvHXKS+++INn7bqTR NkqVPt88w+/Bz/CBxgR1duSHGtbaoAED+LLpMMTVSkwxnRdbC4XGFTca8KJUZQPNwSh3ZIcQ 2K7ZURpTePzxAwCE/HLMHttAs83IwCzzzDU/OjLH/2QTTcpI2wGABGvC0bMUP38MwANNPx21 DVNXnQbUV2cdbMuMyoPOyUgjTfOafewRRdRqFtBy2muz7TbaarPdNtiLDm3x/j/flF02y2qy kQzGbO8MheCER2044narabjIeNO57TInEYeS5X73uzQUfNxlKUFu8oo1A5x7/jlhnkZO6eil 41os65A7+o8+l6uEeeYfnY2MH5+zia3qtM69e7K+p76p8LwTX+zcsTOKTNG1c4fTTcQN1JH1 10v1EfbXT9+9SoAjkzjAebsZ7flgo24n8I7uLL7ydKrf8u+vru9h/ZO6P/6k9NNvfvpf+p/v 2Aepx+WNgNsyWOW+t6DqZU9E1XMgBB/YQAqahA8AmIDp+Fe++/nvg3YKnYcIuKjRbbBSAZtf CFcorA+CcHUAYMAJKfW7LwXQgyz03whXaCnYHVBg/vZIiQQruL0RUXCIN0miiFRCs+TRsIPF ux8PeahDS7nNidcS4Q5ZKEIVSrFSV8zUCyOnwzJOEVPM++GinpcNIR6Reko8zBCR2D3ueQ8l gOtdneYnPwD+T1l6e5QBP0XFQuLwkPjLX9Bct0U+trCLoLNfpgbZpkC24x+Ui57lbifBTb5x iRYEye1KgsEM6DF+htQiF6E4KR+iMIpejGUXVfnFSLnylY00pCxZWalbDpBR/+CH7T7ZSWIq 0ZMTJGJKznbKXK5yjLOsZaTSiMtVNtKMI3whGEkYrGfqUpK75KajqFlJ4LFxmMnUHgOtl6A4 cu8w2xulSfJ4OkAKsIa//hSg+TBFSf6pTy77xCZZ+MjPRepRm/KxnyPzOUmDlnNR20oG7nCH QaY1bk1Yo9pF1ZTRWolTUx3dI/D+EcSJ4o6ZG4UCORu3UlV5cVUtTaSantc3k2aOnhvtZ+Ec yqp/qkqnQ5ucTTNX0ZRKjXQp9aXdlDq02Q31pB+lVkyjNlWqqq5KVTnnU/0GuHB59atgDatY x0rWspr1rGgt6x23yta2uvWtcI3rWuNK17ra9a54Xclc88rXvvr1rybdK2AHS9jCGpYlgj2s YhfLWMAmtrGQjaxkt/rYyVr2spidV2Uzy9nOevZy8pyhUUdL2tLSyUwMDIyoVstaoUABtepU /m1LABCMXgCjF7W9bW5xa1ve6ra3uw0ucIf72+L69rjCNW5ykUtc5iq3ucuNLnSn+9zqOve6 0rVudrFLXe5qt7vbDS94x3vb5xUntILxaoJ+wd5gsPcX7m3ve+ML3/naV774rW9+6cvf++r3 v/3dr38DDOABG1jACC5wggnM4AMr+MENXrCDIwzhCVtYwhiucHvNOyL0RiVcggEAMH5x2xGX mMS2RbGJU3ziFrP4xSuOsYpn7GIZ15jGMMaxjXN84x7z+Mc7DrKOh+xjIReZyEBGspGTfOQm M/nJK+awKD38PHCFWMNYprCWL5xlLm85w1/uMpi9TOYxm1nMaA6z/prLLOV4nunDVk6QeL9L Z+/aec53Jm+e64znPuvZz3z+s6ADTeg9G7rNcBRJeqP0nzmt98wEHmqZ0zzpNe/XGcWgNKQr zelNe1rToA4GolMLkkWnCKshXrKqf4GMadAjrV6thzSQoeRaO3nVtx5yM9hxCUqgAsq2Bnau hy3sYuPa2MGOMmwFa+r0YGW9lgYwMujBj2rjo9r8uLa1sa3tbHMb2/PYxzzuIQ95vEMe54jH OdxhDnOMIx7GMEO0591pDQ+jG+zIRCYKYYJP09vf9f53qAPO3lHHttRwPvVmHJ0TQDP3efnI tj8ijo+JS5ziFq84ximOj47D4x7wMPc7/t7hDnKU3N1vAMKgDe1wQIPDG6LghCY2UYgKmKLl K8d5oXXO8pyv3ODMTrizx/LogbM6mN1O+re3zfRsi/se8yg3utXtjnO02xzaKIccIADwrg88 HOwQhSg+QXZNMIICYSC41wW+ZliLpShrny/QqQziBB1byM/T+MX3rve+++PaHe/4uEM+cpKv exzuFoc23PCAZDue2HcnsTq4gQtOiJ0Uo4j5ISpggVs8HtmQ//zdiZQRuIfex3N/c5VJT5Ur s/153A4KMsqhDXHIYQFsVzuaiUGNUDiCEZQYBSlKQYpPeOLsFSBD7uPO/AeTHiNFWb5+U69o oXcrxDyvLcQl/g6AHXDA++Dngfh3IP7yj58H4P+++tPv/X/U/hqM9/nO5Y/cbYghCQ4oQBKE 8QpXZP4Tj9AAFlAB9Jd9BZhzz3cRRXGAxkV9slV3OaF2sKdtAIADO3ADO2CBPNADG8gDWsAD P9CBHDiCFmiBGGiBHSEAO1A02ZANa6AAzSd9DeYLJFAACWCD65ALrtAKpSAKjUABFUABMah7 MmhfCWgRcGd00rZsdBdnORF5N5Z3Ftd939cB5IcFHogFWqCFWuiBPICFPlB+5Jd+KcgB/3AN 4mANbnAAoteGp8djTlAABlAA66ALOtgKovAInFcBbgh6fgiFQHaEFWF6f4hjDohw/s8DMBH4 esFkbQBwguT3gVowiZS4hVqwEBe4AwKwiSuYDdaADS9YhEOIYLyAAAmQgztIfI8AhBQgikro ivQliBQRfURYcEyoehDRcAe4fRVHhZGohVmgBVngBVvgBcOYBQQxB3EwB0ahfmX4D9aQDcrQ BgfAgD03f8q1CzBQh66wg6MQgANojQaIjSwnixOxgOS4XIeYJKv3LUX3bxNYbRXIgZdoERm4 AynYicoAijD4irV4ZumQC6XAg6SwikE4iv44hOYoEUkog+vYJQYnWyLWhyMmhRHXfV+IBTYC AMvYkcxoFCNihRywiQJghspQDdPIhm8IiCxJZNzojZSw/ocUOZMrKXpzkioAQJMv9pDxEZGI mJOMGHvoB4K5KEqZmI//gA0dEYr/mJBdloMDKXwG2YpNCYu6JxSiyJODwWEbyT3Yt4v/gHEA IH5YgIzhMkpj6X3PeJLG8AXVmI7iKH/cOAqt8I0CSIBwmZfXuJfB4HZiAQBxWV63WH0+yY45 2XXxyA/z6IEL1y1oeY8kuYIe8YIIaZX5BZWkIHZTWZlVSXCwZROH6ZCD+YBFVBA9OZE1aZGK +YuNyS1oOZTPaAzVYAxcUAA62ZI1CQx1WJfFF5MDeJvAmZuBeDA5KZzKdl6q95k/+Yqw54hD +QOtSRmPiYFIiQzGcAxWoACW/umUDZaDolB8eciK2zmek6acNEFb/6iV6VWaBHGaBsiL/oCR WBidfzki6DeSm2iGxoAMf1CbgcmXcqkLmDd24IiXAKqX43hn5jkTgImgvaWeH4YSV4aYSCeP OMCB0HmWI3KPSHkMf4CdBsCZIvoLOQgKxecJP3iQ3NmZcbegMhGaRAihXClH3pNqM6maGCl+ 9HkZIamWyFCS/2AMftCftmmcwSls68AKYhdzevibRvqkhVhkLhoTxRmlOzmaiDilrhdwzZlt j5iBPbCjWHIQmRiZ/3AMfuAHVkAA5MmZOfgJY4ei4smiK+p1UwoT6Ml8MuoeXUkcXwmX8OmL PCCm/gByfjywfoj6D3/gB3ewBQXwnwk6Z0kac5/ACQUKqZiKoELxn3sqoe/YaYlZgZlIqJzR gfcIiZD4D2mqprZZp65KYOvgCpzgCbOKfEL4qnRKb355GTHYqWHyhDf6D3rnixxAqo0xhlYo kuxnhnfAqFsgAEeKm6C3DqugCTK3CZvnpFYqrdyabHf6ElXarYKJnIR5k0KziFzaiPwwD1+K gcYaGSWYiRg4r5moqnfQB2vaprhKrZswc2Y3pxhGC/q6nd/qEnmakDKai7QFlhxHrBCxrBDL rH5gB1IgAJl6oHtGrf5qCYVwl8Q1C7YwC7UAsiJLsiMbsg6asgpKnJyK/qXsiFgSGEzhxq4m eKo2W7M4m4k5m6p9YAd7sKYjup3roAr6pm+2Kl+0QAu2kLSwIAuwAAunIAunMLVOm7QDW4sF O1u96rIQeRdeCxhg+7Vei5qFmHcddw/EGrFqi6gR+w92YAd8UAUBEK1022PrEAqaUAmZUAk1 95sjWwuwEAupYAqnYAp5IAl6gAdlgAdnIAmFCwsjK67b6q3EeaQyGraYK7aiAZTp+nTy0K7y Grr0OrqiW7r/0LN2EARsmqusuw6ioLeVgAmFIJ5MO7V6cAaMOwa6ewS8OwZnoAdRm7QCy7rN l7WIlZVc25Oau7xhu7CA+g+Dh7ZrO71sW71u/msHdSC3Fxup1HW3e1sJfHuXsyC4p5AHeKC7 T3AEQ2AEQ1AEQ+AETzAGZSAJqXCy2wupxqtXLUuuspW5/uu1EUih8hB147KrV2EHczAHODAA QUunrhsJmBAJkzAIrEgLTXu7u0sEM7DBG2wERHAEY+C4sCCwtNDAcScUW8u/iMi8LEy2Lfk8 73AP50ZyJDcO7oB4tVd70fiJJ3mSyDCb1jmkjDrEVFAFVIC92FsF4Vq3k2u3oTAJEhwJfVsB tCC4pnAGu2sEHHwCG1wER4AHeRALkDsLklvGSmbAAGK5ybuV//u/6Opvz2NuU7du7TYOtFcO LZgN2MDDzAASHoqm/qt6r3ZQBXRAB8qgjAo8AFc7hOvQCVHsCGJAuxZ8vkdgBFocA1wcAx+M B5LwtEtrwk6Zv7aDvCr8sm28vM4LoDBsblV3cjiseGnoidbQw9ZZy4u6qM3arG97B8pwDddQ B8rIAwFwvyqbXOzge5DgCJDQsQMosqlwCro7BFvcxU8AxrKQCiiLsdzLZ6IsRPvbYap3ypkb wEZnXnLgBud8zm7gBmvAzuzcBWvABfEcz1bABfV8z1tgBVWgz1BQBTogBThQBQGNAzmAAzAK ytLXDZ2gCI6wCJBcwVCLxZU8A5i8wR88BnrwtIu8ot18OSkMzoTJwv/rwm2IDHLgFExs/sY9 xmuO0Agt3QR76ASmUAZjoMEzcAJcPANejAenEAu0UAtNHNSjV7lQerkiLbZvPHDEoGDPswY4 oANPrQFpeAx3MAdrYAU4UAMaAAEPoAAPgAEaUAMbUANngKsIbV/doAkMrQiGEMlBWAFYTASX nNObfAojvNGu2NGVQ8ogLVtHnbmpvM3mlQZrQNhpYAUcsAZ1IAd1cNU4sAEbcAALEAEbAAWF zQZrAAF4QMzaPGi81giKANowHQIjYAENIANHcNM4rdPV3NMgy9nbjFx6bRIN2tnj2teIKM5I fZh1GsfL+LbAjMBocAU4cANjfdw3AAVWENxvuwdrcAFljdcs/soOaa0IhGDdYhACIoACIqAC K/ABIEDXX2zXw3vWClm53Cmjfy22JG2cVXbLQnydi7qf9F3L9A3Ef4CSyJDZm63SQp1sx8xr hJAIA94E250C3u0CnNfF4+3TQO3fEK5kKFzUaywYuh22Sb187eh2zx3dxGvW7DAKjtANjjAI hIAIhCAIKIAC3t0CK0ABHjADdX3XHy7d5i1m6n3hYxvbhUmliwvbmcoOq5AJ3dAEiTAIiDAI R74CLY4EKlABJVAE8ZsHpoDNZGzbWB7bWk5/Rq3j5Cya+wIARnAGp2DWII5v3QAIgjAIgpDi bN4EK8ACK/ACLz4CF60HelC15W3m/jbO55WW4+sNGO1tpcjABT4RAUPA01Aa4TfGDmIgBk0g BoMg6YMw6ZHeBErg4ihQATIwA+37xWEMuf+96Cld6qhX4cYR6NfzqqdwBrjr6q/u6owr67A+ 67F+67Ze5jfOduygBEuwBIGwBJcO6YJw6UzABEFo0V/cybBgCzX+7LvOZqjeG6oe2EBezNd+ jSR7l6Y9gN3OedwugOwLwo4bC6+N7eie5dmujtOu45sL7X4e7WAWX8FA7wdW7/Yu7/G+72oG 6Do+6Iw+6gIf8ARv6gNv8NLa5eud4fDe5w7f8BCv73w9Zaqn6oK+5emO8eqe8eve8RxPjgqv 218e8fxO/vIPL/EmH6PTbvHA6mR0cltQQGJQEGRqwmRscvDGSSe2NfNJBgUv1iYI/ws+X/A1 JmIwZvQ2BvBIr2JLT+ouFvJHzfAIBgUNFvMMFvMQhvXuZfUpn2BYn2Zaz15U/15fP55hD8oC gV9pf2ZrH19rH+/+vt7Wbl1Df1xDT/U7rya6xfPBUPPBxfe95fM+X1uD3wuCb/h4D/N+b/iM X8yAv/cwz/h9P/hrslt13/d7ryaWz/czX/OZn/hUv/jAUPOF//HIJRDHJRC6NRC5tbCAqfq4 hfq7BfvBwPq2RVuy3wuAWVu7/3Ptruojv2BnD19YX/zxNfbtVfz2xfXJL/TH/v/8akL8Yk/2 Yl/2tcj89WX10d/30+/88xX22t/8Qi/2Wz/24S/9Qq/83q/5EJ+TvO3+8BWahzn/8+9e8M9e 8F/77UX/+I//bg8QwX4NFEhw4K+CCA8GQ/YP2UOIyAAAiBgRCkVkUDJu1NiRo0Rgv4D1Cjmy JEmRvaCsXJkSikgoIaGgnHnypcmUwGKiHHkz5q+aO2+qhFnyZ82cSXEu5dn0pE6WO3X2dJkU ac+oValOzSlVqNafXJFedaq0LNOnZp8CAMZWJFu2TQG8DQkApd22eenGnTuyr1u9eH8JRqs2 ZcOKDycmfnjxo0fIGyUqNFiZMpSEBTEDRRj1JWWg/p8Tfr4s8GWwz6k1m86K2nLmhbFfy+Zs WfXBlp0rbwaqOapv3MF7D9x8mrNr3aBhz2aunHbCuYODRY8+eCIAgXOnE6Q+EHt369ezK9Qe fbvz5gURJ17M2PFj+B6xk+wVjL79kfhV3qcf1P5M/QIkCbP7dtLPv5gwE6q/AQUEkL/8IKxP QgHzM7DAniAk8MEMLexwwggBNHDBDgnU6b8JOQRxxQpZpPDF/Pjqxa7p/KIPr/lonLE+GWmU McK28MtxQh2BdNHI9SpqLzHHInNSsuqWU060hVIjbrUqiZONt9qSQ+5LK7UcLr3nyjTIuDO9 DJPLL48TbU3WTFNTzNqK/mPTTPTylBI27QYjDyHzwMPOu/EK9fO886qjLrtByXwtyYiWrOi9 J58E7KymhrKpKJhWQmuls6LKVCuitlIJVJkwTaswVgf81FSoaAqrVFcH9NQlr6Dq6lauem0p qK1mXVVVw9CKKy+Sjr0uL7cEw9EkZaGdqCTAqp2LMGKZghQiSS2iqFL4JtsTz3Ed1ZNcdM0t 91x102U3T9LWldfdedu199168aV3oG0Vw4hJjMCNbD4jWzQYxiMTPrhghBdWuGGIGZb44YlR qzhiijN2eGOMOb74Y40l7Feifyf9Nr5wo9z33nxbXllfll92Oc2ZYba5Zpxl1lndkbuFiFKU /gceduhiszWa6FaRPrpopZtm+umkoV46aqqnttrpqrG2uueSvQ065Zt3Djtmssc2O+ey0T5b bLXbTpdrxjIK+Gv5PA4ZZLvz7nhvvPm++2+9+xYccL8DJxxkuN07WWAo2XY87cfXhnxyySt3 m/LLb04c4Mbo3uhSrUOXWvSsRze9dNSvPl311Eln/XXXld7cZM89EjfyzHPHfXfMebfcd917 /2V2r2v/3HDkC1d+8OSZX/5w6JuP/nnpDSf+58U9n4zsf7r3/nvwwxd/fPLL/+d34dFXP/j1 gQf++sbmZlwj0KU2/3788y8/9tX5b71/AP5PgLALIAG15RD2dA17/sbTyO3apT8IRjB/7KOg +9JXwQtaED3wk1vn5icRvUlQhCMUn/NMOL0TVo96K0whC1HYCw42yXjbgxkJbUjC9mVQhznk IQZ7uJAYys9z9UvaDY0oQgMm0X9KLOASndhEKFYtiAxUjM2OeEX97dCHW9QgF7W4silSkWAa y991zHhGNKZRjWfM3wtV2EI4utGFb5RjHKf4wQaqjExljFsf/Ri3i0ywi4P8YiF/aEh5hRGP IBkdH9f4SEiaMZD4Y+IAn2jJKFZSk5dM4h1n2CdzOTKSo1TjJO+HSEIeUpVeXKVsFGm8MW5M lKSk5XVMab465pKOu5xjL+PISyQhUEkK/owfFWk4r1nWkpa3LB8rnZnKZ6Lygq/8IBFVlUxl jpKZ5NtkJjn5TW+GE5PjNJonP+jAd2Ezm5Dc5vik2cp3RhOe6KPmEPOmznWusZ3i06Uv+/lL fwLzn9Azp/b0eK4yekahC2VoQ/cZPnlGFJoTjafM6sk4aw6tjPlUZhvBSc6PdhOk4hRpJ4UZ KWJ2cJHoXFcZsTg+ieSvojOdp0RpWq6Lfi2WDXPpS8MXU0oKVKgBJSpAjTpQhxWUccdEV099 +j2g3s+mNaUoVW9KrpyCK6NqcepTuxdVXIZUrCQd60jNWlLZnZRbKZXhOUGZTvwBwKtQlSv+ rnpXq+bVpllF/tlOK9ZVr4K1fEgtKmGPOtTDFlapX2PqHuM6V+8JlnxTpWxVKxtPvloqdIB9 qmTHh1ayhvasZQXtaA+jVn8BMntLfWtLHwtZh9T1lHq1bG3xulfUkky1HrRnxzjrU8/yE7GG Je5wjVtYuy0WXI1F6GshG1yI2pa2t5WuMzMLn62i5bcvhS74Svtd0oZXtOCVmnJRxlI8bfeK /eju96j73unGN3jXHdg9netV9soWl8dNbH+Li1wAD868T2LuuNR7xPwKEr6XlW91cUffj2TX KQc2YoIpKV7TZpi8Gsbw1nLrs2KutLXpve9TLTxbBqfYwQveGYTr5tsS+/TE+w3w/n/9y18b 53jAYHtgjF8642auuMEsJrLmPsxWIWJ0sz7GIpC52WEojzfKHJbyaRkDYpV+0opMXm97vVdk FYNZyDg98m7FaN/76dfEXu5ejnH85hrD+ag7jkyBy0ThGzp5fGEecp/53DIXf27JaYZtbD1a ZSonesOL7jCdP4Le5uD5hmw+n5/HfGkxOyrQ9EOz+VL6R1ArqY1ydnOcTX1jfzradgc1cFw5 WksFWzrTsqY1Q8qsON4quZGFfiqjEe1rRU9Zk6puXL54/dRZ/xnTtebXrTl35r0dG4uopnap q03q4RI7jyuT9hWZnexl/3nTjERdt2/4a2GnG93rJqe2/qs4M3OP8NvzVjaRx+3XkMVbf9c+ tbX9je3nudvO4CY4vcN9cILcW93BZjewHb7wh5tF4CMueL0NXnFM39tweFkRxwWkI5Dz6GP4 Bji/Tf7vftNx4jKbCGhajp5GxTxfrEb4xW0ubmfTToyumxZPWq6XwNxF6Mi61qWmZS2+pKQv M2o4xJ3edKgf8MpINibF2fUd7kjnUFsvj3cOQp1FHarliLpWQWRec4unPeM5L94iSY4wu4Qc R0KiO1xqJPId2ShIybJPyO8+95QH/uQlR/nEVs6y7nTdT1r3+rW4DqisL37xZ1c8zTF++Zuz TeGr+wvRlT4YuXweWn4RfbXq/jJ0uDDr6QxnfcRbr7WVz+zsjMc6ZWY/ebMXauyBijyi0P57 zEdT437jePHvExf9GD9Ges/7jOyerJ73nUiFpz7hrd/vw7fP6moPPvAzn57Nn3Xpq3d9+cl/ /kzGfnfb9z73v/9+Mk/dzG6//uAFX/371z+X2Xd//9v//wULP/OLugF8PQJEvwFSv+5bQPgD wBwaPvuLQPyTQP2rwIBjuwUSMQf0PwbkwJoSQAQsQBEMQRL8NQVswA7cQBUUHgicQBe0QArM Pxk0PAwMMS1LQRxEQR0EtBrsoJ07QCA0QCEcwSA0KfnDNQ30QCVcwR20oBaEwRecwSiMQSqE oR5s/iuDYsIlzMEttBwQLMISHMIwJMLWO8EuPEMtTEOyeUIphEI3bEM4PDn+U8Mm5EI63JMv FEMw3EM97EMybAozvENBtEMcZMMqfMNDjMMp1LErXC3Gsrw6jEQ05MI8/MMxvEQ+tMTVCURC lMRBREFDXEREFEVFHEXracSq+8RJ9MROtLUjfLZQ4xZM9MNZ1ERaJCmucShd3EVe7EVf/EVg DEZhHEZiLEZjPEZkTEZlJMaJ0DdnfEZojEZpnMZuw7JYZJLcukZt3EZu7EZv/EZwDEdxHEdy LMdYhIJs9EZr7EZ0NEd3fEd4jEd5nEd6rEdsNMd15MZ2tEd+7Ed//EeATwzIUNtHcszHbSRI gUxIhVxIhmxIUENIcTRIbYRIh6xIi7xIjHRHigRHibzGjcxIkAxJkRzJxkjHbnw1lExJlVxJ lmxJl3xJmIxJmWTJgAAAOw== --------------05C4ECC0D9E268701256003F-- --------------83E086C8E0C05D28C44660F1--
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: what exactly is Smalltalk ? Date: 9 May 1998 00:43:12 GMT Organization: Idiom Communications Message-ID: <6j08r0$9q6$1@news.idiom.com> References: <1998050822010800.SAA10901@ladder01.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: rafkah@aol.com Rafkah may or may not have said: -> I've been talking about Smalltalk a lot. what is it exactly ? what are his -> advantages against oc, c++ or java ? -> -> - Raphie Check out the FAQ in comp.lang.smalltalk. -jcr
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer,comp.sys.next.advocacy Subject: Re: whats up with zoneraiders?!?/! Date: 8 May 1998 21:15:50 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6j0ao6$k5s$1@crib.bevc.blacksburg.va.us> References: <6j04ej$nvc$1@newsfep4.sprintmail.com> In article <6j04ej$nvc$1@newsfep4.sprintmail.com>, macghod@concentric.net wrote: > Zoneraiders is out, and it is for rhapsody for ppc ONLY. I thought > something written for rhapsody would work on BOTH rhapsodies? And openstep > as well? Why in gods name would zoneraiders be ppc only? Maybe they didn't have an Intel box to test it on so didn't want to risk releasing it even if it should theoretically work, or maybe they did something really low-level (there are bound to be _some_ hardware dependencies, though I can't think of what they might have done!). > Also, I am starting to learn programming in openstep. So I make a simple > app, lets say currencyconverter. What would I do to get this to work on > rhapsody for ppc and intel? Check the appropriate architectures in the Build Options panel. Build. > A while ago on a thread on the developer program, I said I thought > programmers would need rhapsody to develop for rhapsody, and the openstep > programmers wondered why everyone would think apps have to be written > differently to work on rhapsody. Do they or dont they? OpenStep apps can be moved forward to Rhapsody pretty much without trouble. Every new release NeXT (now Apple) tweaks the libraries so sometimes minor things break between releases, and that will inevitably include the transition from OPENSTEP 4.2 to Rhapsody. But like I said, it's pretty minor stuff, and a lot of the major OpenStep developers didn't have to change any of their code at all, even for large apps. (You'll have to rework the .nib files to have Apple menu structures and such, though. Might need to fine-tune the placement of UI elements in IB too, since some of them changed a bit.)
From: izumi@pinoko.berkeley.edu (Izumi Ohzawa) Newsgroups: comp.sys.next.programmer Subject: NEXTSTEP_App.nib and WINDOWS_App.nib Date: 9 May 1998 03:38:44 GMT Organization: University of California, Berkeley Distribution: world Message-ID: <6j0j44$9hb$1@agate.berkeley.edu> If I get source code with only NEXTSTEP_App.nib, how do I create WINDOWS_App.nib, and vice versa? Are these supposed to be created by hand individually, or is there some way to covert one to the other? -- Izumi Ohzawa <izumi@pinoko.berkeley.edu>
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer,comp.sys.next.advocacy Subject: Re: whats up with zoneraiders?!?/! Date: 8 May 1998 21:17:27 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6j0ar7$k6q$1@crib.bevc.blacksburg.va.us> References: <6j04ej$nvc$1@newsfep4.sprintmail.com> <6j0ao6$k5s$1@crib.bevc.blacksburg.va.us> In article <6j0ao6$k5s$1@crib.bevc.blacksburg.va.us>, nurban@vt.edu wrote: > In article <6j04ej$nvc$1@newsfep4.sprintmail.com>, macghod@concentric.net wrote: > > Zoneraiders is out, and it is for rhapsody for ppc ONLY. I thought > > something written for rhapsody would work on BOTH rhapsodies? And openstep > > as well? Why in gods name would zoneraiders be ppc only? > Maybe they didn't have an Intel box to test it on so didn't want to > risk releasing it even if it should theoretically work, I guess another possibility would be they just didn't install the binaries necessary for building other architectures, conceivably to save disk space. But I'm just guessing..
From: nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) Newsgroups: comp.sys.next.programmer,comp.sys.next.advocacy Subject: Re: whats up with zoneraiders?!?/! Date: 9 May 1998 01:31:54 GMT Organization: none Message-ID: <6j0bma$l3h$3@ha2.rdc1.nj.home.com> References: <6j04ej$nvc$1@newsfep4.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: macghod@concentric.net In <6j04ej$nvc$1@newsfep4.sprintmail.com> macghod@concentric.net wrote: > Zoneraiders is out, and it is for rhapsody for ppc ONLY. I thought > something written for rhapsody would work on BOTH rhapsodies? Compiling for PPC and Pentium should be as easy as selecting a checkbox before compiling. > And openstep as well? An OpenStep application can be ported to Rhapsody relatively easily from what I understand, but Rhapsody != OpenStep and OpenStep != Rhapsody > Why in gods name would zoneraiders be ppc only? Perhaps they dislike supporting monopolies and therefore want to give people apps which only run on PPC. My guess is the only people who know the real answer are the people who wrote it. TjL -- Nothing worth saying can be said in less than four lines.
From: "Michelle L. Buck" <buck.erik@mcleod.net> Newsgroups: comp.sys.next.programmer Subject: Re: NEXTSTEP_App.nib and WINDOWS_App.nib Date: Sat, 9 May 1998 00:33:08 -0500 Organization: McleodUSA - http://www.mcleodusa.net Message-ID: <6j0pul$hl21@odie.mcleod.net> References: <6j0j44$9hb$1@agate.berkeley.edu> Izumi Ohzawa wrote in message <6j0j44$9hb$1@agate.berkeley.edu>... > >If I get source code with only NEXTSTEP_App.nib, >how do I create WINDOWS_App.nib, and vice versa? > >Are these supposed to be created by hand individually, >or is there some way to covert one to the other? > >-- >Izumi Ohzawa <izumi@pinoko.berkeley.edu> Just copy NEXTSTEP_App.nib to WINDOWS_App.nib and you will have a working windows App. To be completely compliant, you should rearrange your menus and a few other details, but it is not necessary.
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer,comp.sys.next.advocacy Subject: Re: whats up with zoneraiders?!?/! Date: 9 May 1998 04:14:54 GMT Organization: Digital Fix Development Message-ID: <6j0l7u$6vb$1@news.digifix.com> References: <6j04ej$nvc$1@newsfep4.sprintmail.com> <6j0ao6$k5s$1@crib.bevc.blacksburg.va.us> In-Reply-To: <6j0ao6$k5s$1@crib.bevc.blacksburg.va.us> On 05/08/98, Nathan Urban wrote: >In article <6j04ej$nvc$1@newsfep4.sprintmail.com>, macghod@concentric.net wrote: > <snip> >> Also, I am starting to learn programming in openstep. So I make a simple >> app, lets say currencyconverter. What would I do to get this to work on >> rhapsody for ppc and intel? > >Check the appropriate architectures in the Build Options panel. Build. No, he's only got OpenStep, not Rhapsody. You'll need to have a Rhapsody installation to compile a Rhapsody binary.. You need move the source tree over, and recompile. You'll probably also want to tune the UI in IB > >> A while ago on a thread on the developer program, I said I thought >> programmers would need rhapsody to develop for rhapsody, and the openstep >> programmers wondered why everyone would think apps have to be written >> differently to work on rhapsody. Do they or dont they? > >OpenStep apps can be moved forward to Rhapsody pretty much without >trouble. Every new release NeXT (now Apple) tweaks the libraries >so sometimes minor things break between releases, and that will >inevitably include the transition from OPENSTEP 4.2 to Rhapsody. >But like I said, it's pretty minor stuff, and a lot of the major >OpenStep developers didn't have to change any of their code at >all, even for large apps. (You'll have to rework the .nib files >to have Apple menu structures and such, though. Might need to >fine-tune the placement of UI elements in IB too, since some of >them changed a bit.) > If you are careful when you write the software, then you usually can just move stuff over. Yes, there have been some small changes in Rhapsody.. I expect more small ones until we hit CR1. However we're talking small changes... -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: John Hornkvist <sorry@no.more.spams> Newsgroups: comp.sys.next.programmer Subject: Re: How to get multiple views with tabs in OS 4.2 ? Date: Sat, 9 May 1998 12:39:09 GMT Organization: Chalmers Tekniska Högskola Sender: john@haddock.cd.chalmers.se (John Hprnkvist) Message-ID: <EsowH9.45n@haddock.cd.chalmers.se> References: <3553A365.D013CEF7@easynet.fr> <EsovB6.41y@haddock.cd.chalmers.se> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: sorry@no.more.spams In <EsovB6.41y@haddock.cd.chalmers.se> John Hornkvist wrote: > In <3553A365.D013CEF7@easynet.fr> Arnaud wrote: > >I have seen it in Nextstep 3.3 screenshots. Does it exist in 3.3 and is it > hidden in 4.2 ? > >Is there something like a MiscKit of palettes available for 4.2 ? > > > IIRC there is a solution to your problem in the MiscKit, and I think Rhapsody > will have a class for it. That should be the MiscKit _for OPENSTEP_, to avoid confusion... :)
From: mark@sapphire.oaai.com (Mark Onyschuk) Newsgroups: comp.sys.next.programmer Subject: Re: Questions: NSNotificationCenter, NSBrowser Date: 9 May 1998 11:32:44 GMT Organization: M. Onyschuk and Associates Inc. Message-ID: <slrn6l91bb.rra.mark@sapphire.oaai.com> References: <kiss-0805980035400001@ip112.santa-clara6.ca.pub-ip.psi.net> <1d8plt6.11oa4aslb163gN@carina55.wco.com> In article <1d8plt6.11oa4aslb163gN@carina55.wco.com>, Mike Paquette wrote: >Richard Kiss <kiss@slip.NOSPAM.net> wrote: >> (1) Judging from the headers, NSNotificationCenter appears to be thread >> safe. When a NSNotification is sent to the sender, does it block until it >> is broadcast to all the listeners, or just the listeners in this thread, >> or is it non-blocking? How does it get the NSNotification to objects in >> other threads? Through an NSConnection? > >All of the above! [snip] >I wouldn't use it for thousands of messages/sec, but for the applications >I've described (up to maybe 10 messages/sec) it works well with minimal >overhead. Great post - I didn't realize that queues were vended on a per-thread basis. A couple of observations: To allow notifications to be posted across threads (ie. thread A posts so that an object receives the message executed by thread B) we use the multi-threded messaging scheme covered in the on-line documentation for NSConnection. We're preparing a bit of an article and download on our site http://www.oaai.com/ which wraps that NSConnection stuff in a nice easy-to-use category of NSThread. We've also found that NSNotificationCenter really degrades in performance with many object registered to receive many notifications - looks like an N^2 degredation. As a result, we'd constructed a simplified version of the notification center for GlyphiX. I think that the EOF folks also did this to speed notifications in their (I'm sure) notification intensive framework. Mark
From: John Hornkvist <sorry@no.more.spams> Newsgroups: comp.sys.next.programmer Subject: Re: How to get multiple views with tabs in OS 4.2 ? Date: Sat, 9 May 1998 12:13:54 GMT Organization: Chalmers Tekniska Högskola Sender: john@haddock.cd.chalmers.se (John Hprnkvist) Message-ID: <EsovB6.41y@haddock.cd.chalmers.se> References: <3553A365.D013CEF7@easynet.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: debayeux@easynet.fr In <3553A365.D013CEF7@easynet.fr> Arnaud wrote: > >Hello, > >In Interface Builder, and in other programs, I have seen the following interface : > >Each tab (Instances, Classes, Sounds, Images) selects a specific view. >The tiff files for the graphics effectively exists in the Interface Builder resources. >How do you program this ? >Why isn't it in the palette window of Interface Builder ? >I have seen it in Nextstep 3.3 screenshots. Does it exist in 3.3 and is it hidden in 4.2 ? >Is there something like a MiscKit of palettes available for 4.2 ? > >Thanks for your help > >From a novice Openstep developer... NOTE: Please don't post images to usenet. IIRC there is a solution to your problem in the MiscKit, and I think Rhapsody will have a class for it. If you need to do it on your own, one solution is to use a view that takes a contentview, like a scrollview. Use a controller that keeps your subviews in a dictionary, indexed by title. When a tab is clicked, the controller can read its title, select the proper view from the dictionary and set the contentview for the scrollview accordingly. Regards, John Hornkvist Name: nhoj Address: cd.chalmers.se
From: macghod@concentric.net Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Yellowbox programming, and ragosta Date: 10 May 1998 00:00:43 GMT Message-ID: <6j2qnb$r2p$2@newsfep1.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit In the past, I am sure I have seen joe ragosta say that openstep programming to multiple platforms (those supported by Openstep that is) is as simple as clicking a box and recompiling. I would like Joe to walk me through this process to show his knowledge of this. To everyone else (besides Joe), I did the simple first app from the openstep docs, currencyconverter. Followed everything it said to do, compiled it, no problems. I then went to the build options and clicked all the boxes for other platforms. GOt a whole buncha errors :( Then only selected one box, but not the platform I am on, whole buncha errors :( -- Running on Openstep 4.2, the best os in the world, and about to get better! Mac/openstep qa work desired, email me for resume NeXTMail and MIME OK!
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Yellowbox programming, and ragosta Date: 10 May 1998 00:16:04 GMT Organization: MiscKit Development Message-ID: <6j2rk4$skr$2@news.xmission.com> References: <6j2qnb$r2p$2@newsfep1.sprintmail.com> macghod@concentric.net wrote: > To everyone else (besides Joe), I did the simple first app from > the openstep docs, currencyconverter. Followed everything it said > to do, compiled it, no problems. I then went to the build options > and clicked all the boxes for other platforms. GOt a whole buncha > errors :( Then only selected one box, but not the platform I am > on, whole buncha errors :( Well, we already told you the likely cause of your problems. Now that you've given more specifics, I can positively say that one of the previous suggestions--your dev. tools are not properly installed--is in fact the case. This means that you installed the developer tools "thin". If you want to cross compile, you must install them "fat". This is because in order to link a multiple architecture binary, you must have multi-architecture libraries and frameworks on your system. (This should be obvious.) Installer.app on OPENSTEP makes this easy to do, and the installation instructions explain it. But you have to check the boxes to install for all the architectures when you do the install. As I reacall, you're using OPENSTEP 4.2, so you just need to RTFM; it should be easy to reinstall it the way you need it. On RDR1, it isn't so easy because for some dumb reason Apple set it to automatically install the tools thin, so you have to break out the command line and re-install them FAT by hand. Figures that they'd go and BREAK things that worked just fine before. :-P (Thank goodness it is only the install that is broken.) This has been documented many times (publically) on rhapsody-dev@omnigroup.com, so I'm not breaking any NDAs by saying this. Again, RTFM. Once you get things installed right, it should be no tougher than clicking the check box to get things to cross compile. -- Later, -Don Yacktman don@misckit.com <a href="http://www.misckit.com/don.html">My home page</a>
From: ethorne@chicagonet.net (Edwin E. Thorne) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Yellowbox programming, and ragosta Date: Fri, 8 May 1998 19:22:33 -0600 Organization: ISPNews http://ispnews.com Message-ID: <1d8q5gj.17qnilxg9rrdzN@pppsl552.chicagonet.net> References: <6j2qnb$r2p$2@newsfep1.sprintmail.com> <macghod@concentric.net> wrote: > In the past, I am sure I have seen joe ragosta say that openstep programming > to multiple platforms (those supported by Openstep that is) is as simple as > clicking a box and recompiling. > > I would like Joe to walk me through this process to show his knowledge of > this. Nah, don't ask Joe, I can tell you what to do. 1. Raise your PC above your head with both hands. 2. Throw it down against a cement floor as hard as you can. 3. Sweep up the pieces. 4. Obtain another PC and resume with step # 1. 5 Repeat steps 1 through 4 until Openstep works as desired, or until all available PCs are exhausted, whichever comes first. > To everyone else (besides Joe), I did the simple first app from the openstep > docs, currencyconverter. Followed everything it said to do, compiled it, no > problems. I then went to the build options and clicked all the boxes for > other platforms. GOt a whole buncha errors :( Then only selected one box, > but not the platform I am on, whole buncha errors :( > > > -- > Running on Openstep 4.2, the best os in the world, and about to get better! > Mac/openstep qa work desired, email me for resume > NeXTMail and MIME OK! Edwin ------------------------------------------------ Created on an Apple Macintosh Performa 6400/180
From: macghod@concentric.net Newsgroups: comp.sys.next.programmer Subject: Re: Yellowbox programming, and ragosta Date: 10 May 1998 00:51:39 GMT Message-ID: <6j2tmr$2nt$2@newsfep1.sprintmail.com> References: <6j2qnb$r2p$2@newsfep1.sprintmail.com> <6j2rk4$skr$2@news.xmission.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: don@misckit.com In <6j2rk4$skr$2@news.xmission.com> Don Yacktman wrote: > macghod@concentric.net wrote: > Well, we already told you the likely cause of your problems. Now that you've > given more specifics, I can positively say that one of the previous > suggestions--your dev. tools are not properly installed--is in fact the case. I am sorry, but I am a bit confused? This is the first post in this vein by me, are you maybe thinking of something else? I did start a thread about casidy and green, but that wasnt about ME. > This means that you installed the developer tools "thin". If you want to > cross compile, you must install them "fat". This is because in order to link > a multiple architecture binary, you must have multi-architecture libraries > and frameworks on your system. (This should be obvious.) Installer.app on > OPENSTEP makes this easy to do, and the installation instructions explain it. > But you have to check the boxes to install for all the architectures when you > do the install. As I reacall, you're using OPENSTEP 4.2, so you just need to > RTFM; it should be easy to reinstall it the way you need it. I did read the manual. And I must say it was not very helpful. The manual I have is"installing and configuring OPENSTEP" Basically the instructions boild down to "install all the packages in /NextCD/Packages" And I did do this -- Running on Openstep 4.2, the best os in the world, and about to get better! Mac/openstep qa work desired, email me for resume NeXTMail and MIME OK!
From: jurgen@oic.de (Juergen Moellenhoff) Newsgroups: comp.sys.next.programmer Subject: Re: Yellowbox programming, and ragosta Date: 10 May 1998 02:45:04 GMT Organization: OIC, Bochum, Germany Message-ID: <6j34bg$jfi$1@nexus.oic.de> References: <6j2qnb$r2p$2@newsfep1.sprintmail.com> <6j2rk4$skr$2@news.xmission.com> <6j2tmr$2nt$2@newsfep1.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit In <6j2tmr$2nt$2@newsfep1.sprintmail.com> macghod@concentric.net wrote: > > In <6j2rk4$skr$2@news.xmission.com> Don Yacktman wrote: But you > > have to check the boxes to install for all the architectures when > > you do the install. As I reacall, you're using OPENSTEP 4.2, so > > you just need to RTFM; it should be easy to reinstall it the way > > you need it. > > > I did read the manual. And I must say it was not very helpful. The > manual I have is"installing and configuring OPENSTEP" > > Basically the instructions boild down to "install all the packages in > NextCD/Packages" And I did do this Hmmm... From my "CD-Manual" (OPENSTEP 4.2/Mach): ---------------------------------------------- Installing the libraries 1.. 2.. 3. Select the additional architectures you wish to develop for. Important: You should select every architecture that might run the applications you create. 4... 5... ---------------------------------------------- Bye, Juergen
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <14190894168021@digifix.com> Date: 10 May 1998 03:49:44 GMT Organization: Digital Fix Development Message-ID: <1562894772821@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1993. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: Henry McGilton <henry@trilithon.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Yellowbox programming, and ragosta Date: Sat, 09 May 1998 22:48:16 -0700 Organization: Trilithon Research and Trading Message-ID: <35553FA0.AB984F57@trilithon.com> References: <6j2qnb$r2p$2@newsfep1.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit macghod@concentric.net wrote: * To everyone else (besides Joe), I did the simple first * app from the openstep docs, currencyconverter. Followed * everything it said to do, compiled it, no problems. I * then went to the build options and clicked all the boxes * for other platforms. GOt a whole buncha errors :( Then * only selected one box, but not the platform I am on, * whole buncha errors :( Of course, we all notice that you carefully avoided detailing precisely what those errors happened to be. As an analogy, I today I installed an entire new system with Windows NT, and when I tried to connect to the ISP, I GoT a WhOlE BuNcHa ErRoRs! Based on your inability to describe the problems you claim to be having (over and above the intrinsic problems), I can conjecture that you didn't install all of the software neccessary to build for multiple platforms. If my conjecture turns out to be the case, would you kindly let the rest of the anxiously awaiting long-term NextStep/OpenStep community know so we can exhale and get back to business (since 1993) as usual? ........ Henry ============================================================= Henry McGilton | Trilithon Software, and, Boulevardier, Java Composer | Pacific Research and Trading -----------------------------+------------------------------- mailto:henry@trilithon.com | http://www.trilithon.com =============================================================
From: mmalcolm crawford <m.crawford@shef.ac.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Yellowbox programming, and ragosta Date: 10 May 1998 12:57:31 GMT Organization: University of Sheffield, UK Message-ID: <6j487r$ibo$1@bignews.shef.ac.uk> References: <6j2qnb$r2p$2@newsfep1.sprintmail.com> In-Reply-To: <6j2qnb$r2p$2@newsfep1.sprintmail.com> On 05/10/98, macghod@concentric.net wrote: > In the past, I am sure I have seen joe ragosta say that openstep > programming to multiple platforms (those supported by Openstep that > is) is as simple as clicking a box and recompiling. >[...] > To everyone else (besides Joe), I did the simple first app from > the openstep docs, currencyconverter. Followed everything it said > to do, compiled it, no problems. I then went to the build options > and clicked all the boxes for other platforms. GOt a whole buncha > errors :( Then only selected one box, but not the platform I am > on, whole buncha errors :( > Oh heavens, I can just see this is going to join various other unpleasant Urban Myths about OPENSTEP (like DPS is slow)... I am quite sure that you have some sort of installation problem. Did you install all the relevant binaries? You should only be clicking boxes for binaries you have installed -- if you're trying to compile for Sparc, for example, on Prelude to Rhapsody, and don't have the Sparc binaries installed, you'll have problems. Similarly, if you're running RDR, as far as I could tell from demos of Project Builder, it still has buttons for compiling on NeXT and Sparc, and yet as far as I'm aware versions of RDR were never produced for those platforms. So again, if you tried compiling for either of those platforms, you'll be sunk. Cross-compilation has been working fine on NEXTSTEP and OPENSTEP since NEXTSTEP 3.1, for much more complicated applications than the one you're describing. Unless you have gone out of your way to break things, I can see no reason why it should suddenly fail for you. I trust when you have discovered your error you will post a suitable retraction so that whenever anyone cites your original post in the future, we can direct them to the followup... mmalc. --
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Yellowbox programming, and ragosta Date: 10 May 1998 10:51:22 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6j4eta$rbo$1@crib.bevc.blacksburg.va.us> References: <6j2qnb$r2p$2@newsfep1.sprintmail.com> In article <6j2qnb$r2p$2@newsfep1.sprintmail.com>, macghod@concentric.net wrote: > In the past, I am sure I have seen joe ragosta say that openstep programming > to multiple platforms (those supported by Openstep that is) is as simple as > clicking a box and recompiling. > To everyone else (besides Joe), I did the simple first app from the openstep > docs, currencyconverter. [...] whole buncha errors :( Others have pointed out your problem. What I don't understand is, why are you singling out Joe?? If you've been paying attention at all, you'd have noticed that the rest of us _also_ say that it's as simple as clicking a box and recompiling, as long as you stay within the Yellow Box. (At least for different architectures all running OS/Mach, and not Mach->Windows or or something.. that can have added complications.)
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: Re: Yellowbox programming, and ragosta Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer References: <6j2qnb$r2p$2@newsfep1.sprintmail.com> <6j487r$ibo$1@bignews.shef.ac.uk> Message-ID: <3555ce8a.0@news.depaul.edu> Date: 10 May 98 15:58:02 GMT In comp.sys.mac.advocacy mmalcolm crawford <m.crawford@shef.ac.uk> wrote: > On 05/10/98, macghod@concentric.net wrote: <Macghod's public flailing, questioning something that many people other than Joe Ragosta have been doing for years, snipped> <Mmalc's wisdom also snipped, only with more respect> >I trust when you have discovered your error you will post a suitable >retraction so that whenever anyone cites your original post in the future, we >can direct them to the followup... And, Macghod, make sure you post back to EACH AND EVERY ONE of the newsgroups you compulsively posted this to Do you have 'Usenet Turet's Syndrome' or something, that makes you spastically cross-post to the entire known universe? -- "... and subpoenas for all." - Ken Starr
From: macghod@concentric.net Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Yellowbox programming, and ragosta Date: 10 May 1998 17:49:27 GMT Message-ID: <6j4pb7$jkq$1@newsfep3.sprintmail.com> References: <6j2qnb$r2p$2@newsfep1.sprintmail.com> <35553FA0.AB984F57@trilithon.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: henry@trilithon.com In <35553FA0.AB984F57@trilithon.com> Henry McGilton wrote: > macghod@concentric.net wrote: > Based on your inability to describe the problems you claim > to be having (over and above the intrinsic problems), I can > conjecture that you didn't install all of the software > neccessary to build for multiple platforms. Yes, you are right. My irritation with Joe lead me to be out of line, and rude, and I apologize for it. I will try to change my ways.
From: macghod@concentric.net Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Yellowbox programming, and ragosta Date: 10 May 1998 17:54:13 GMT Message-ID: <6j4pk5$jkq$2@newsfep3.sprintmail.com> References: <6j2qnb$r2p$2@newsfep1.sprintmail.com> <6j487r$ibo$1@bignews.shef.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: m.crawford@shef.ac.uk In <6j487r$ibo$1@bignews.shef.ac.uk> mmalcolm crawford wrote: > Oh heavens, I can just see this is going to join various other unpleasant > Urban Myths about OPENSTEP (like DPS is slow)... Yes, and I apologize. I acted out of line. I will change the way I act here. > I am quite sure that you have some sort of installation problem. Did you Yes, you are right, as usual. The default install only installs the libraries "thin" not fat. I screwed up, and being more irritated than usual at joe, I posted that out of line post. > I trust when you have discovered your error you will post a suitable > retraction so that whenever anyone cites your original post in the future, we > can direct them to the followup... yes, the original post was a great embarrasment, and %100 incorrect.
From: <paw@mcmail.com> Newsgroups: comp.sys.next.programmer Subject: CHEQUES CASHED Message-ID: <35562e3d.0@news1.mcmail.com> Date: 10 May 1998 23:46:21 GMT CHEQUES CASHED: Rates from 3.25% including "Not Neg" & "Account Payee Only". If you require a cheque cashing and do not wish to use a bank account, please call us today. Totally discreet,efficient & confidential service. Tel 07970 622645 or E-mail paw@mcmail.com for info. (UK ONLY SORRY).
From: spamcan@iswest.com Organization: Internet Specialties West, Inc. Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <35562e3d.0@news1.mcmail.com> Date: 10 May 1998 22:51:01 GMT Control: cancel <35562e3d.0@news1.mcmail.com> Message-ID: <cancel.35562e3d.0@news1.mcmail.com> Sender: <paw@mcmail.com> Spam cancelled. Autocancel spam type: spam
From: "Michael J. Peck" <mjpeck@nstar.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Yellowbox programming, and ragosta Date: Sun, 10 May 1998 23:39:31 +0000 Organization: ISPNews http://ispnews.com Message-ID: <35563AB3.B8A9AAC@nstar.net> References: <6j2qnb$r2p$2@newsfep1.sprintmail.com> <6j4eta$rbo$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Nathan Urban wrote: > Others have pointed out your problem. What I don't understand is, > why are you singling out Joe?? If you've been paying attention at all, > you'd have noticed that the rest of us _also_ say that it's as simple as > clicking a box and recompiling, as long as you stay within the Yellow Box. I think the point was that while "the rest of [you]" know what you're talking about, and can answer the question... ...draw your own conclusions. It's a matter of "legitimacy of claims". MJP
From: "Craig Halley" <craig@mpv.com> Newsgroups: comp.sys.next.programmer Subject: Re: How to get multiple views with tabs in OS 4.2 ? Date: Sun, 10 May 1998 22:18:06 -0500 Organization: Southwestern Bell Internet Services, Richardson, TX Message-ID: <6j5qn3$c2u$1@nnrp1> References: <3553A365.D013CEF7@easynet.fr> The NSTabView is definitely part of the Rhapsody Developer Release, so you'll have access to it under all environments where that applies (Yellow Box for Windows, Rhapsody, etc.). Someone else mentioned that it's in the MiscKit, but I can't confirm that. If you want to try it on your own, it's not too difficult... with the exception of the tabs themselves. I'd recommend doing a simplified version that doesn't use overlapping tabs, which can get a little tricky. Use a matrix of buttons at the top of an NSView, and set its contentView based on which button is pressed. You can change the background color of the selected button, instead of depressing it, to make it behave more like a tabbed display. There's lots of examples of SwitchView's out there (the idea behind a tab view), and just about every NEXTSTEP developer I know has done one just for kicks, so I'd start with that if you're going to roll your own. Check the FTP sites for code to get some hints from. Craig Halley MPV, Inc.
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: NSConnection-getRootProxy only 256 calls possible? Date: Mon, 11 May 1998 11:10:22 +0200 Organization: Square B.V. Message-ID: <3556C07E.296300A0@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit [In an earlier post I asked why distant objects fail with only a light usage] I have found out when the programs start to recieve nil as an answer. After only a mere 256 calls the program seems to block and will not respond anymore. Does anybody know how I can prevent this from happening. I can't sell my customer that he should quit the server after 256 client connections!! Maurice le Rutte. -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto:mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+ Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" and "to trick into believing or accepting as genuine something false and often preposterous"
From: maul@anke.imsd.uni-mainz.DE (Mathias Maul) Newsgroups: comp.sys.next.programmer Subject: class "Zombie"? Date: 11 May 1998 12:22:43 GMT Organization: Johannes Gutenberg-Universitaet Mainz, Germany Message-ID: <6j6qij$b38$1@bambi.zdv.Uni-Mainz.DE> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi there, has anyone ever encountered this error message when trying to run an executable from gdb? ---> objc: class `Zombie' not linked into application An uncaught exception was raised Typed streams library error: class error for 'Zombie': class not loaded Program exited with code 0377. <--- Compiling and linking (NeXTStep 3.3/i386) went without errors or warnings; I also double-checked my .nib files for instances of objects that might have been removed in the project. Needless to mention, there is (and has been) no class called "Zombie" in my project. Any hints will be greatly appreciated ¼ foo!, Matt.
From: macghod@concentric.net Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: future of mac programming Date: 11 May 1998 20:18:17 GMT Message-ID: <6j7me9$so3$1@newsfep4.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I had been teaching myself mac programming, up till a month ago, when I said to myself "with rhapsody the mac api's are going to be gone, yellowbox is Apple's future" and stopped with the mac api, and started learning openstep programming. With todays "carbon" anouncement I am VERY confused ?? -- Running on Openstep 4.2, the best os in the world, and about to get better! Mac/openstep qa work desired, email me for resume NeXTMail and MIME OK!
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 11 May 1998 16:39:26 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> References: <6j7me9$so3$1@newsfep4.sprintmail.com> In article <6j7me9$so3$1@newsfep4.sprintmail.com>, macghod@concentric.net wrote: > I had been teaching myself mac programming, up till a month ago, when I said > to myself "with rhapsody the mac api's are going to be gone, yellowbox is > Apple's future" and stopped with the mac api, and started learning openstep > programming. With todays "carbon" anouncement I am VERY confused ?? I think we need to wait a bit longer and learn more about Apple's plans. It looks like MacOS and Rhapsody will be unified into one OS with improved MacOS-based "Carbon" APIs.. but it doesn't sound like Yellow Box is being dropped. If you have existing MacOS apps, then it's to your advantange to do a more minimal Carbon port to get them to run "native" on MacOS X; if you want to do a Yellow port you kind of might as well do a full rewrite since otherwise there's not much point because a simple port wouldn't really take advantage of the Yellow APIs. But if you're writing a _new_ app, then it's definitely in your interests to do it in Yellow, since the APIs will still be a lot better than Carbon. Or so I surmise, so far.
From: M Rassbach <mark@milestonerdl.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 11 May 1998 15:20:57 -0500 Organization: Inc.Net http://www.inc.net Message-ID: <35575DA8.B7A1148F@milestonerdl.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Nathan Urban wrote: > In article <6j7me9$so3$1@newsfep4.sprintmail.com>, macghod@concentric.net wrote: > > With todays "carbon" anouncement I am VERY confused ?? Join the club. > I think we need to wait a bit longer and learn more about Apple's plans. A day? A week? A month? or 3 years? > but it doesn't sound like Yellow > Box is being dropped. But there is no conformation of the status of YellowBox staying either.
From: macghod@concentric.net Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: rhapsody for intel and "carbon" Date: 11 May 1998 22:00:27 GMT Message-ID: <6j7sdr$4gn$1@newsfep4.sprintmail.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: nurban@crib.bevc.blacksburg.va.us In <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> Nathan Urban wrote: > In article <6j7me9$so3$1@newsfep4.sprintmail.com>, macghod@concentric.net wrote: > > > I had been teaching myself mac programming, up till a month ago, when I said > > to myself "with rhapsody the mac api's are going to be gone, yellowbox is > > Apple's future" and stopped with the mac api, and started learning openstep > > programming. With todays "carbon" anouncement I am VERY confused ?? Another thing, will carbon based apps run on rhapsody for intel? -- Running on Openstep 4.2, the best os in the world, and about to get better! Mac/openstep qa work desired, email me for resume NeXTMail and MIME OK!
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 11 May 1998 18:17:43 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6j7te7$vah$1@crib.bevc.blacksburg.va.us> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> In article <35575DA8.B7A1148F@milestonerdl.com>, mark@milestonerdl.com wrote: > Nathan Urban wrote: > > I think we need to wait a bit longer and learn more about Apple's plans. > A day? A week? A month? or 3 years? How about a day, when they have the talk on Rhapsody and Yellow Box? > > but it doesn't sound like Yellow Box is being dropped. > But there is no conformation of the status of YellowBox staying either. Scott Anguish says otherwise.. http://www.stepwise.com/SpecialCoverage/WWDC98/index.html though I don't know where he got his information from.
From: Arnaud <debayeux@easynet.fr> Newsgroups: comp.sys.next.programmer,comp.sys.next.hardware Subject: 3D acceleration under OS 4.2 ? Date: Tue, 12 May 1998 00:11:24 +0200 Organization: [posted via] Easynet France Message-ID: <3557778C.D3D40D37@easynet.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello, anyone using a 3D hardware under Openstep 4.2 Mach/Intel ? Even if you developed a custom driver, I'm interested ! Thanks for your help.
From: rriley@yahoo.com (Rex Riley) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Yellowbox programming, and ragosta MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit References: <6j2qnb$r2p$2@newsfep1.sprintmail.com> <6j487r$ibo$1@bignews.shef.ac.uk> <3555ce8a.0@news.depaul.edu> Message-ID: <xGL51.1323$sy4.2938729@news.rdc1.sdca.home.com> Date: Mon, 11 May 1998 23:17:49 GMT NNTP-Posting-Date: Mon, 11 May 1998 16:17:49 PDT Organization: @Home Network In <3555ce8a.0@news.depaul.edu> Jonathan W Hendry wrote: > In comp.sys.mac.advocacy mmalcolm crawford <m.crawford@shef.ac.uk> wrote: > > On 05/10/98, macghod@concentric.net wrote: > > <Macghod's public flailing, questioning something that many people > other than Joe Ragosta have been doing for years, snipped> > > <Mmalc's wisdom also snipped, only with more respect> > > >I trust when you have discovered your error you will post a suitable > >retraction so that whenever anyone cites your original post in the future, we > >can direct them to the followup... > > And, Macghod, make sure you post back to EACH AND EVERY ONE of the > newsgroups you compulsively posted this to > > Do you have 'Usenet Turet's Syndrome' or something, that makes > you spastically cross-post to the entire known universe? > > Macghod, having carped long and hard on seemingly insignificant problems, is Steve Job's posterchild for why you don't release DR software into the cheap seats. Maybe someday he'll keep his posts where they belong, instead of Chicken Little posting all over the place... -r
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer,comp.sys.next.advocacy Subject: Re: whats up with zoneraiders?!?/! Organization: Is a sign of weakness Message-ID: <SCOTT.98May9211551@slave.doubleu.com> References: <6j04ej$nvc$1@newsfep4.sprintmail.com> <6j0ao6$k5s$1@crib.bevc.blacksburg.va.us> <6j0ar7$k6q$1@crib.bevc.blacksburg.va.us> In-reply-to: nurban@crib.bevc.blacksburg.va.us's message of 8 May 1998 21:17:27 -0400 Date: 11 May 1998 17:50:49 -0500 In article <6j0ar7$k6q$1@crib.bevc.blacksburg.va.us>, nurban@crib.bevc.blacksburg.va.us (Nathan Urban) writes: In article <6j0ao6$k5s$1@crib.bevc.blacksburg.va.us>, nurban@vt.edu wrote: > In article <6j04ej$nvc$1@newsfep4.sprintmail.com>, > macghod@concentric.net wrote: > > Zoneraiders is out, and it is for rhapsody for ppc ONLY. I > > thought something written for rhapsody would work on BOTH > > rhapsodies? And openstep as well? Why in gods name would > > zoneraiders be ppc only? > > Maybe they didn't have an Intel box to test it on so didn't want > to risk releasing it even if it should theoretically work, I guess another possibility would be they just didn't install the binaries necessary for building other architectures, conceivably to save disk space. But I'm just guessing.. Or perhaps the version of Rhapsody they have never gave them the option of installing the support for cross-compiling? -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar Subject: cmsg cancel <6j82dn$9ad$112@news.on> Control: cancel <6j82dn$9ad$112@news.on> Date: 11 May 1998 23:43:14 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6j82dn$9ad$112@news.on> Sender: FAST CASH @fgdfg.asdf Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: don.brown@cesoft.com (Donald Brown) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 11 May 1998 19:03:08 -0500 Organization: Prairie Group Message-ID: <MPG.fc14db42395aa139896a2@news.supernews.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> In article <6j7me9$so3$1@newsfep4.sprintmail.com>, macghod@concentric.net says... > I had been teaching myself mac programming, up till a month ago, when I said > to myself "with rhapsody the mac api's are going to be gone, yellowbox is > Apple's future" and stopped with the mac api, and started learning openstep > programming. With todays "carbon" anouncement I am VERY confused ?? Here is the path they have described today: You learn and program with the Mac APIs. There is a small percentage of those that are going away and being replaced, but you will mostly be working with the APIs we've been using for decades. Go ahead and learn the Mac API. For now, avoid INITs and cdevs and other such cute stuff. Until we know more, many of them will break in the OSX future. Write apps. In your apps, avoid strange tricks between applications. Do >NOT< send an Apple Event from one app to another, pass the address of a buffer, and have the second app write into the buffer in the first app's heap. Avoid the system heap. If you're not already working with Rhapsody/OpenStep/Yellow Box/etc., hold off until we hear more. The word I thought I heard this morning is that those are not the future, at least for client software. Of course, last year's WWDC, the three big things were CHRP, Newton/eMate, and Rhapsody. The first two are dead and the third is AWOL. So, you need to be flexible. Donald
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 11 May 1998 20:16:57 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6j84dp$vln$1@crib.bevc.blacksburg.va.us> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> In article <MPG.fc14db42395aa139896a2@news.supernews.com>, don.brown@cesoft.com (Donald Brown) wrote: > Here is the path they have described today: > You learn and program with the Mac APIs. There is a small percentage of > those that are going away and being replaced, but you will mostly be > working with the APIs we've been using for decades. Go ahead and learn > the Mac API. Better advice: if you're writing a new app, forget the Mac API. OpenStep is infinitely better. Might as well develop good habits from the beginning, and not waste your time. > If you're not already working with Rhapsody/OpenStep/Yellow Box/etc., > hold off until we hear more. The word I thought I heard this morning is > that those are not the future, at least for client software. I didn't hear anything of the sort. Want to provide a quote? I listened to the whole thing. You have to realize that the vast majority of the people at the keynote are _existing Mac developers_, which means that they already _have_ MacOS apps. For them, the thing they care about the most is Carbon, so they don't have to rewrite their apps. For someone doing _new_ development, there's little reason to learn Carbon. Check out how many of the talks they're giving are on Rhapsody-related things, and tell me that Yellow Box is going away. > Of course, last year's WWDC, the three big things were CHRP, > Newton/eMate, and Rhapsody. The first two are dead and the third is > AWOL. AWOL?? RDR1 already in the hands of developers, RDR2 available now, RCR1 available in 3rd quarter.
From: stevenj@alum.mit.edu (Steven G. Johnson) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Mon, 11 May 1998 20:39:00 -0400 Organization: Massachusetts Institute of Technology Message-ID: <stevenj-ya02408000R1105982039000001@news.mit.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <6j7sdr$4gn$1@newsfep4.sprintmail.com> <6j7tij$vb7$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit nurban@vt.edu wrote: >macghod@concentric.net wrote: >> Another thing, will carbon based apps run on rhapsody for intel? > >Good question. My guess would be "no", but there might be a MacOS X for >Intel down the road.. My guess would be "yes." Think of MacOS X as Rhapsody with Latitude++ (Carbon). Both Rhapsody and Latitude *already* support multiple processors natively. That is too much of an advantage for Apple to throw it away for no reason. Of course, we don't know yet if Carbon has anything to do with Latitude--anyone from Metrowerks care to comment? =) Cordially, Steven G. JOhnson
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: 11 May 1998 18:20:03 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6j7tij$vb7$1@crib.bevc.blacksburg.va.us> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <6j7sdr$4gn$1@newsfep4.sprintmail.com> In article <6j7sdr$4gn$1@newsfep4.sprintmail.com>, macghod@concentric.net wrote: > Another thing, will carbon based apps run on rhapsody for intel? Good question. My guess would be "no", but there might be a MacOS X for Intel down the road.. that could likely require an emulation solution though, depending on how portable the Carbon stuff is after they stripped out those 2000 calls.
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 11 May 1998 15:32:01 -0700 Organization: Primenet Services for the Internet Message-ID: <B17CCA7D-359B7@206.165.43.153> References: <6j7me9$so3$1@newsfep4.sprintmail.com> nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit It's an interesting issue, this fragmentation of the MacOS market into ever smaller pieces. Here is what I see: For MacOS developers, there are now at least 5 markets: 68K/PPC System 7.1 & later. 68040/PPC MacOS 8.0 PPC MacOS 8.1 PPC MacOS 8.2 and better (Carbon-based) Yellow Box (Rhapsody, 8.x + libraries (?), WIntel 95/98/NT and MacOS X). Each developer is going to have to look at their installed base, their potential revenue from upgrades and new users, etc., and decide which market to target. In my opinion, for all but a tiny fraction of Apple's developers, this is the most schizophrenic scheme possible. Everyone but the upper-crust, high-end developers who target those who purchase the latest and greatest of everything are going to be scratching their heads on this one. Nothing new for Jobs. ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: stevenj@alum.mit.edu (Steven G. Johnson) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 11 May 1998 21:06:14 -0400 Organization: Massachusetts Institute of Technology Message-ID: <stevenj-ya02408000R1105982106140001@news.mit.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit don.brown@cesoft.com (Donald Brown) wrote: >If you're not already working with Rhapsody/OpenStep/Yellow Box/etc., >hold off until we hear more. The word I thought I heard this morning is >that those are not the future, at least for client software. For goodness sakes, don't make this kind of decision until a couple of days have passed, and we hear from more WWDC sessions. Actually, all of the available evidence at this point indicates that Yellow is alive and well--for example, just look at all of the WWDC sessions on Yellow Box developement. It sounds like MacOS X *is* Rhapsody, with the full Yellow APIs and development environment, but in addition supporting a subset of the MacOS Toolbox API (Carbon) for legacy apps. As S. J. said, Rhapsody was a great start, it just didn't go far enough in supporting people's current investment in the MacOS [these are very close to his exact words]. Sigh...I wish Jobs had been clearer on this point; it would have prevented 24 hours of hysteria. (Which is, I think, about as long as these worries will last.) Cordially, Steven G. Johnson
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 12 May 1998 01:17:29 GMT Organization: MiscKit Development Message-ID: <6j87v9$sam$1@news.xmission.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> nurban@crib.bevc.blacksburg.va.us (Nathan Urban) wrote: > In article <35575DA8.B7A1148F@milestonerdl.com>, > mark@milestonerdl.com wrote: > > But there is no conformation of the status of YellowBox staying either. > > Scott Anguish says otherwise.. > > http://www.stepwise.com/SpecialCoverage/WWDC98/index.html > > though I don't know where he got his information from. I do. Let's just say someone at Apple in a position to know specifically *asked* Scott to post that comment. That's all I can say. -- Later, -Don Yacktman don@misckit.com <a href="http://www.misckit.com/don.html">My home page</a>
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Yellowbox programming, and ragosta Date: 12 May 1998 01:17:52 GMT Organization: P & L Systems Message-ID: <6j8800$1cc$6@ironhorse.plsys.co.uk> References: <6j2qnb$r2p$2@newsfep1.sprintmail.com> <6j487r$ibo$1@bignews.shef.ac.uk> <6j4pk5$jkq$2@newsfep3.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: macghod@concentric.net In <6j4pk5$jkq$2@newsfep3.sprintmail.com> macghod@concentric.net wrote: > In <6j487r$ibo$1@bignews.shef.ac.uk> mmalcolm crawford wrote: > > Yes, and I apologize. > On behalf of all of us, I'm sure, thank you for this. > > I am quite sure that you have some sort of installation problem. Did you > > Yes, you are right, as usual. > Well, I'm afraid I'll have to correct you on that too :-) I do sometimes get things badly wrong, not least very recently in my expectations of how things would be announced at WWDC. I hope I'll always have the good grace to admit my errors in future too, though. Best wishes, mmalc.
From: don.brown@cesoft.com (Donald Brown) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 11 May 1998 20:17:04 -0500 Organization: Prairie Group Message-ID: <MPG.fc15f0828d669629896aa@news.supernews.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <6j84dp$vln$1@crib.bevc.blacksburg.va.us> In article <6j84dp$vln$1@crib.bevc.blacksburg.va.us>, nurban@crib.bevc.blacksburg.va.us says... > In article <MPG.fc14db42395aa139896a2@news.supernews.com>, don.brown@cesoft.com (Donald Brown) wrote: > > > Here is the path they have described today: > > > You learn and program with the Mac APIs. There is a small percentage of > > those that are going away and being replaced, but you will mostly be > > working with the APIs we've been using for decades. Go ahead and learn > > the Mac API. > > Better advice: if you're writing a new app, forget the Mac API. > OpenStep is infinitely better. Might as well develop good habits from > the beginning, and not waste your time. Depends on what you're doing, and who your target market is. The word from this morning is, there is a future in programming for the Mac APIs. There also is a present. Even when Rhapsody ships, I have strong doubts how widely it will be accepted. Not certainties, doubts. > > If you're not already working with Rhapsody/OpenStep/Yellow Box/etc., > > hold off until we hear more. The word I thought I heard this morning is > > that those are not the future, at least for client software. > > I didn't hear anything of the sort. Want to provide a quote? I listened > to the whole thing. Nope, no exact quote. To rephrase myself, I don't think it will be a focus of Apple in the near future. > You have to realize that the vast majority of the people at the keynote > are _existing Mac developers_, which means that they already _have_ > MacOS apps. For them, the thing they care about the most is Carbon, > so they don't have to rewrite their apps. For someone doing _new_ > development, there's little reason to learn Carbon. Check out how many > of the talks they're giving are on Rhapsody-related things, and tell me > that Yellow Box is going away. The reason to learn Carbon is that, with few changes, that code will also work with OS 8 or OS 7.x. > > Of course, last year's WWDC, the three big things were CHRP, > > Newton/eMate, and Rhapsody. The first two are dead and the third is > > AWOL. > > AWOL?? RDR1 already in the hands of developers, RDR2 available now, > RCR1 available in 3rd quarter. In the list of the "three things most important to Apple", Rhapsody didn't place. It fell below Java! I call that AWOL. May show up later in the WWDC, but AWOL during the keynotes. Donald
From: schuerig@acm.org (Michael Schuerig) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Tue, 12 May 1998 03:23:44 +0200 Organization: Completely Disorganized Message-ID: <1d8waq9.1t0soxr6xzkjtN@rhrz-isdn3-p44.rhrz.uni-bonn.de> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> Mail-Copies-To: never Donald Brown <don.brown@cesoft.com> wrote: > In your apps, avoid strange tricks between applications. Do >NOT< send > an Apple Event from one app to another, pass the address of a buffer, and > have the second app write into the buffer in the first app's heap. Avoid > the system heap. From <http://developer.apple.com/macosx/pdf/CarbonPaper.pdf> page 46: <EXCERPT> Carbon will fully support the Apple Event Manager. Standard Apple events previously defined by Apple--for example, the required suite of Apple events discussed in Inside Macintosh: Interapplication Communication--will be supported in Mac OS X and play the same roles as they do in Mac OS 8. See also "Open Scripting Architecture" (page 66) for information on scripting support. </EXCERPT> And page 66: <EXCERPT> Carbon will fully support the Open Scripting Architecture (OSA), including AppleScript-specific interfaces. See "Apple Event Manager" (page 46) for related information. </EXCERPT> So if there's any safe means of IAC, the it's AppleEvents. Michael -- Michael Schuerig mailto:schuerig@acm.org http://www.uni-bonn.de/~uzs90z/
From: macghod@concentric.net Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 12 May 1998 01:40:12 GMT Message-ID: <6j899s$g46$2@newsfep4.sprintmail.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <stevenj-ya02408000R1105982106140001@news.mit.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: stevenj@alum.mit.edu > For goodness sakes, don't make this kind of decision until a couple of days > have passed, and we hear from more WWDC sessions. Actually, all of the > available evidence at this point indicates that Yellow is alive and > well--for example, just look at all of the WWDC sessions on Yellow Box > developement. It sounds like MacOS X *is* Rhapsody, with the full Yellow > APIs and development environment, but in addition supporting a subset of > the MacOS Toolbox API (Carbon) for legacy apps. As S. J. said, Rhapsody > was a great start, it just didn't go far enough in supporting people's > current investment in the MacOS [these are very close to his exact words]. But why did SteveJ say this? Isnt it true that what he says is not true, to wit, without this diverting attention from rhapsody, he could of just stressed metrowerks lattitude? Just keeping rhapsody and reminding people they can use lattitude to port their old macos apps does about the same thing, doesnt it? -- Running on Openstep 4.2, the best os in the world, and about to get better! Mac/openstep qa work desired, email me for resume NeXTMail and MIME OK!
From: Mark Bessey <"mark_bessey"@apple.com (no spam, please)> Newsgroups: comp.sys.next.programmer Subject: Re: NSConnection-getRootProxy only 256 calls possible? Date: Mon, 11 May 1998 16:14:00 -0700 Organization: Apple Computer, Inc. Message-ID: <6j80mf$fqe$1@news.apple.com> References: <3556C07E.296300A0@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit CC: mlerutte@square.nl There are two problems you're likely encountering here: (1) The BSD sockets code imposes a limit of 256 open sockets at a time. (2) Some versions of D.O. don't ever free up their sockets. Due to #1, you won't be able to support more than 256 connections at once. As for #2, I believe all the known resource leaks in D.O. have been fixed in Rhapsody DR2, which should be available real soon now... -Mark Maurice le Rutte wrote: > > [In an earlier post I asked why distant objects fail with only a > light usage] > > I have found out when the programs start to recieve nil as an > answer. > After only a mere 256 calls the program seems to block and will not > respond anymore. Does anybody know how I can prevent this from > happening. I can't sell my customer that he should quit the server > after 256 client connections!! > > Maurice le Rutte. > -- > +-----------------------------+---------------------------------+ > | OpenStep & Java developer | mailto:mlerutte@square.nl | > | Square B.V. | http://www.square.nl | > | Buitenop 5 | voice: (0475) 355 100 | > | 6041 LA ROERMOND | fax: (0475) 355 199 | > +-----------------------------+---------------------------------+ > > Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" > and "to trick into believing or accepting as genuine something > false and often preposterous"
From: M Rassbach <mark@milestonerdl.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 11 May 1998 20:02:45 -0500 Organization: Inc.Net http://www.inc.net Message-ID: <35579FB5.5452B326@milestonerdl.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Donald Brown wrote: > Of course, last year's WWDC, the three big things were CHRP, > Newton/eMate, and Rhapsody. The first two are dead and the third is > AWOL. And the above is a fine example of why Apple is a company not to be trusted.
From: M Rassbach <mark@milestonerdl.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 11 May 1998 20:04:41 -0500 Organization: Inc.Net http://www.inc.net Message-ID: <3557A029.11F01866@milestonerdl.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <stevenj-ya02408000R1105982106140001@news.mit.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Steven G. Johnson wrote: > don.brown@cesoft.com (Donald Brown) wrote: > >If you're not already working with Rhapsody/OpenStep/Yellow Box/etc., > >hold off until we hear more. The word I thought I heard this morning is > >that those are not the future, at least for client software. > > For goodness sakes, don't make this kind of decision until a couple of days > have passed, and we hear from more WWDC sessions. Internet discussion like this ARE useful....hopefully Apple/Attendees will ask EXACTLY these types of questions can get some direction of where Apple is going. > Sigh...I wish Jobs had been clearer on this point; it would have prevented > 24 hours of hysteria. (Which is, I think, about as long as these worries > will last.) You can only hope.....
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 11 May 1998 22:38:57 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6j8co1$11$1@crib.bevc.blacksburg.va.us> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <6j84dp$vln$1@crib.bevc.blacksburg.va.us> <MPG.fc15f0828d669629896aa@news.supernews.com> In article <MPG.fc15f0828d669629896aa@news.supernews.com>, don.brown@cesoft.com (Donald Brown) wrote: > In article <6j84dp$vln$1@crib.bevc.blacksburg.va.us>, > nurban@crib.bevc.blacksburg.va.us says... > > Better advice: if you're writing a new app, forget the Mac API. > > OpenStep is infinitely better. Might as well develop good habits from > > the beginning, and not waste your time. > Depends on what you're doing, and who your target market is. > The word from this morning is, there is a future in programming for the > Mac APIs. There also is a present. Even when Rhapsody ships, I have > strong doubts how widely it will be accepted. Not certainties, doubts. Rhapsody's acceptance has nothing to do with developing to the Yellow APIs. They run on more than just Rhapsody, you know. > > > If you're not already working with Rhapsody/OpenStep/Yellow Box/etc., > > > hold off until we hear more. The word I thought I heard this morning is > > > that those are not the future, at least for client software. > > I didn't hear anything of the sort. Want to provide a quote? I listened > > to the whole thing. > Nope, no exact quote. To rephrase myself, I don't think it will be a > focus of Apple in the near future. It depends. Apple clearly regards the development environment as "important" to the future of MacOS X and their OS strategy. To what degree they emphasize this to developers is uncertain, but I really don't see them stopping development on the Yellow APIs. > The reason to learn Carbon is that, with few changes, that code will also > work with OS 8 or OS 7.x. Is Apple reneging on its promise to bring Yellow Box to MacOS? I haven't heard that.. on the other hand, I haven't heard anything else about it either. > > > Of course, last year's WWDC, the three big things were CHRP, > > > Newton/eMate, and Rhapsody. The first two are dead and the third is > > > AWOL. > > AWOL?? RDR1 already in the hands of developers, RDR2 available now, > > RCR1 available in 3rd quarter. > In the list of the "three things most important to Apple", Rhapsody > didn't place. It fell below Java! In the overall scheme of things, it isn't important, because Rhapsody itself as an individual product going to cease to exist next year! It's going to be called "Mac OS", as you'll recall. That phrase refers to a lot of technologies. > I call that AWOL. May show up later in the WWDC, but AWOL during > the keynotes. Mention of Yellow Box as it pertains to Mac OS X was AWOL in the keynote. Lack of mention of "Rhapsody" doesn't matter. We'll likely see tomorrow where Apple sees Yellow Box fitting in.
From: don.brown@cesoft.com (Donald Brown) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 11 May 1998 22:21:52 -0500 Organization: Prairie Group Message-ID: <MPG.fc17c4e2c1ad5439896ab@news.supernews.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <1d8waq9.1t0soxr6xzkjtN@rhrz-isdn3-p44.rhrz.uni-bonn.de> In article <1d8waq9.1t0soxr6xzkjtN@rhrz-isdn3-p44.rhrz.uni-bonn.de>, schuerig@acm.org says... > Donald Brown <don.brown@cesoft.com> wrote: > > > In your apps, avoid strange tricks between applications. Do >NOT< send > > an Apple Event from one app to another, pass the address of a buffer, and > > have the second app write into the buffer in the first app's heap. Avoid > > the system heap. > > From <http://developer.apple.com/macosx/pdf/CarbonPaper.pdf> page 46: > > <EXCERPT> > Carbon will fully support the Apple Event Manager. Standard Apple events > previously defined by Apple--for example, the required suite of Apple > events discussed in Inside Macintosh: Interapplication > Communication--will be supported in Mac OS X and play the same roles as > they do in Mac OS 8. See also "Open Scripting Architecture" (page 66) > for information on scripting support. > </EXCERPT> > > And page 66: > > <EXCERPT> > Carbon will fully support the Open Scripting Architecture (OSA), > including AppleScript-specific interfaces. See "Apple Event Manager" > (page 46) for related information. > </EXCERPT> > > So if there's any safe means of IAC, the it's AppleEvents. I didn't state my warning clearly. The good and will continue to be supported way of sending a hunk of data from one app to another is to put that hunk into an Apple Event and send it. The bad way that will break under protected memory is if you keep the data in yourself, and you just send the 4 byte address of your block, and that other block takes that address, and starts writing. Apple Events will be supported, if you don't break the rules. That's what I meant to say. Donald
From: don.brown@cesoft.com (Donald Brown) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 11 May 1998 22:30:55 -0500 Organization: Prairie Group Message-ID: <MPG.fc17e6f707c72619896ac@news.supernews.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <stevenj-ya02408000R1105982106140001@news.mit.edu> <6j899s$g46$2@newsfep4.sprintmail.com> In article <6j899s$g46$2@newsfep4.sprintmail.com>, macghod@concentric.net says... > > But why did SteveJ say this? Isnt it true that what he says is not true, to > wit, without this diverting attention from rhapsody, he could of just > stressed metrowerks lattitude? Just keeping rhapsody and reminding people > they can use lattitude to port their old macos apps does about the same > thing, doesnt it? > The thing was, to be a good Rhapsody friendly app, you had to do major rewrites, their estimate was 1-2 man years for a large app. And, there was reluctance to do this. The new Carbon system means that you can use almost the same old API and get all the advantages of the new OS, and their estimate was 5 days to do this to a large app. I'm unclear about whether MacOS X will run openstep programs. That was NOT the impression I got from the keynote, but others have had different impressions. Presuming I'm wrong and Mac OS X will run both Carbon-based apps using most of the classic Mac API and will run OpenStep based apps, it will be much easier to take an existing program and make it fully support the new OS via Cargon instead of OpenStep. So, the whole point I was trying to make is...your original post said you thought the Mac API was dead and so there's no purpose in learning it. With the announcements made today, the Mac API is not dead and there is some value in learning it. Whether there will be other APIs/development environments that will work on the Mac OS X machines or not...is beyond the scope of this message. Because I just don't know yet. Donald
From: don.brown@cesoft.com (Donald Brown) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 11 May 1998 22:32:17 -0500 Organization: Prairie Group Message-ID: <MPG.fc17eba5ac7c7be9896ad@news.supernews.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <1d8waq9.1t0soxr6xzkjtN@rhrz-isdn3-p44.rhrz.uni-bonn.de> <correia-1105982314010001@206.119.231.106> In article <correia-1105982314010001@206.119.231.106>, correia@barebones.com says... > > So if there's any safe means of IAC, the it's AppleEvents. > > The wording in the original post could have been more clear. The point > was don't pass a long across an apple event that is really a pointer to a > buffer and expect the other app to be able to write into or read from this > buffer. "Normal" apple events should be ok. > Yeah. What Jim said. Donald
From: correia@barebones.com (Jim Correia) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 11 May 1998 23:14:01 -0400 Organization: Bare Bones Software, Inc. Message-ID: <correia-1105982314010001@206.119.231.106> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <1d8waq9.1t0soxr6xzkjtN@rhrz-isdn3-p44.rhrz.uni-bonn.de> 4Gsb1P\V@,Y=`_HPtpn/WCgz}lmj0}1&r> In article <1d8waq9.1t0soxr6xzkjtN@rhrz-isdn3-p44.rhrz.uni-bonn.de>, schuerig@acm.org (Michael Schuerig) wrote: > > In your apps, avoid strange tricks between applications. Do >NOT< send > > an Apple Event from one app to another, pass the address of a buffer, and > > have the second app write into the buffer in the first app's heap. Avoid > > the system heap. [SNIP] > So if there's any safe means of IAC, the it's AppleEvents. The wording in the original post could have been more clear. The point was don't pass a long across an apple event that is really a pointer to a buffer and expect the other app to be able to write into or read from this buffer. "Normal" apple events should be ok. -- Jim Correia Bare Bones Software, Inc. correia@barebones.com <http://www.barebones.com>
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 12 May 1998 00:14:11 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6j8iaj$be$1@crib.bevc.blacksburg.va.us> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <stevenj-ya02408000R1105982106140001@news.mit.edu> <6j899s$g46$2@newsfep4.sprintmail.com> <MPG.fc17e6f707c72619896ac@news.supernews.com> In article <MPG.fc17e6f707c72619896ac@news.supernews.com>, don.brown@cesoft.com (Donald Brown) wrote: > I'm unclear about whether MacOS X will run openstep programs. That was > NOT the impression I got from the keynote, It definitely will. See the Stepwise coverage I mentioned in another article. > So, the whole point I was trying to make is...your original post said you > thought the Mac API was dead and so there's no purpose in learning it. > With the announcements made today, the Mac API is not dead and there is > some value in learning it. I don't think anyone is arguing that the Mac API is dead -- it's certainly going to be used for a long time to come in the existing Mac software shops. It's just that for someone starting out _new_ without any legacy code, it's better to pay attention to Yellow since the APIs are a lot better.
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 11 May 98 15:35:49 Organization: Is a sign of weakness Message-ID: <SCOTT.98May11153549@slave.doubleu.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> In-reply-to: M Rassbach's message of Mon, 11 May 1998 15:20:57 -0500 In article <35575DA8.B7A1148F@milestonerdl.com>, M Rassbach <mark@milestonerdl.com> writes: Nathan Urban wrote: > In article <6j7me9$so3$1@newsfep4.sprintmail.com>, > macghod@concentric.net wrote: > > With todays "carbon" anouncement I am VERY confused ?? Join the club. > I think we need to wait a bit longer and learn more about Apple's > plans. A day? A week? A month? or 3 years? > but it doesn't sound like Yellow Box is being dropped. But there is no conformation of the status of YellowBox staying either. NeXT and Apple have always been known for their abilities in promulgating a very simple message that confuses the hell out of everyone. That said, I can't believe they'd have such a high percentage of the conference devoted to Rhapsody sessions if they intended to dump Rhapsody in a year. I mean, all of the NeXTSTEP/OpenStep/Rhapsody vendors _together_ don't add up to most of the single MacOS vendors. I think that why Rhapsody got little mention in the keynote was simply to keep the MacOS contingent calm. My biggest non-hint is that at no point did they mention a version of Carbon for Windows - which is a tremendously obvious thing to not mention at such an event. My suspicion is that Carbon is essentially a hugely improvd BlueBox, and that something AppKit based is still where the medium to long term future is. The big thing I think has been left up in the air is whether the AppKit will be an Objective-C thing or a Java thing. They've essentially reset the countdown on Rhapsody (who's going to deploy on a system which was pre-announced as having a one-year lifespan?). Given that extra year, it's hard to argue Java versus Objective-C, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 11 May 98 21:23:26 Organization: Is a sign of weakness Message-ID: <SCOTT.98May11212326@slave.doubleu.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> In-reply-to: nurban@crib.bevc.blacksburg.va.us's message of 11 May 1998 18:17:43 -0400 In article <6j7te7$vah$1@crib.bevc.blacksburg.va.us>, nurban@crib.bevc.blacksburg.va.us (Nathan Urban) writes: In article <35575DA8.B7A1148F@milestonerdl.com>, mark@milestonerdl.com wrote: > Nathan Urban wrote: > > but it doesn't sound like Yellow Box is being dropped. > > But there is no conformation of the status of YellowBox staying > either. Scott Anguish says otherwise.. http://www.stepwise.com/SpecialCoverage/WWDC98/index.html though I don't know where he got his information from. The "Rhapsody and Yellow Box in 1998" session seemed to very much lay it out - in the sense of decades, Yellow Box is the future. In the near and medium term, the former "blue box" (without the "box") is provided and use is encouraged in order to get the most people over to a more capable system with the minimum amount of pain. And for companies with larger MacOS investments, it might make sense to continue using Carbon for quite some time. But for fresh starts, Yellow is the way to go, since it will be more portable, and in my experience easier to code for to boot, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: see-below@not-my-address.com (Matthew Vaughan) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 11 May 1998 20:14:15 -0700 Message-ID: <see-below-1105982014160001@209.24.240.244> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <1d8waq9.1t0soxr6xzkjtN@rhrz-isdn3-p44.rhrz.uni-bonn.de> In article <1d8waq9.1t0soxr6xzkjtN@rhrz-isdn3-p44.rhrz.uni-bonn.de>, schuerig@acm.org (Michael Schuerig) wrote: > Donald Brown <don.brown@cesoft.com> wrote: > > > In your apps, avoid strange tricks between applications. Do >NOT< send > > an Apple Event from one app to another, pass the address of a buffer, and > > have the second app write into the buffer in the first app's heap. Avoid > > the system heap. > > From <http://developer.apple.com/macosx/pdf/CarbonPaper.pdf> page 46: > > <EXCERPT> > Carbon will fully support the Apple Event Manager. Standard Apple events > previously defined by Apple--for example, the required suite of Apple > events discussed in Inside Macintosh: Interapplication > Communication--will be supported in Mac OS X and play the same roles as > they do in Mac OS 8. See also "Open Scripting Architecture" (page 66) > for information on scripting support. > </EXCERPT> I don't think he was saying not to use Apple Events, though it does sound like it at first. He was referring to one app telling another to access the first app's memory space directly. .................................................... MATTHEW VAUGHAN matthewv at best dot com (damn spammers...) http://www.best.com/~matthewv/ ....................................................
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 12 May 1998 00:48:58 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6j8kbq$el$1@crib.bevc.blacksburg.va.us> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <SCOTT.98May11153549@slave.doubleu.com> In article <SCOTT.98May11153549@slave.doubleu.com>, scott@doubleu.com (Scott Hess) wrote: > The big thing I think has been left up in the air is whether the > AppKit will be an Objective-C thing or a Java thing. They have mentioned that they've been making some modifications in the Obj-C++ language (I think to cater to the C++ crowd), so that sounds like they're not intending to drop it.
From: szallies@energotec.de (Constantin Szallies) Newsgroups: comp.sys.next.programmer Subject: Saving NSTableView column width and order in the default database Date: 12 May 1998 07:23:08 GMT Organization: Customer of Technet Message-ID: <6j8tcs$hd4$1@oxygen.technet.net> Hi everybody, I was wondering if somebody has already implemented the above functionality. Because I'm a lazy programmer, I like to reuse existing code instead of reinventing the wheel :-) By by and thanx -- # Constantin Szallies, Energotec GmbH # szallies@energotec.de # http://www.energotec.de/~szallies/ # 49211-9144018
From: macghod@concentric.net Newsgroups: comp.sys.next.programmer Subject: the M.O.S.T, but for openstep? Date: 12 May 1998 05:48:31 GMT Message-ID: <6j8nrf$ps6$2@newsfep3.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I would like to know if their is something like the MOST, yet for learning openstep/rhapsody rather than macos programming? FOr those who are not familiar with it, the MOST is a great resource for those wanting to learn to program for the mac (www.themost.org): "The MOST web site and services are the gifts of altruistic Mac OS professionals and amateurs to anyone - high school/college student, developers for other platforms, or current Mac OS developer - who wants access to net-based resources which will help them produce high quality Mac OS applications. The MOST provides an umbrella non-organization which facilitates opportunities for its volunteer members to work together to provide each other with free access to: a data base of resources for Mac OS developers self-study net based courses based on public domain and/or commercial textbooks or products opportunities to participate in group-study development projects group mentoring in Mac OS technologies and development via focused email discussion lists" -- Running on Openstep 4.2, the best os in the world, and about to get better! Mac/openstep qa work desired, email me for resume NeXTMail and MIME OK!
From: Pierre Schwarz <peschwar@diapool.informatik.uni-stuttgart.de> Newsgroups: comp.sys.next.programmer Subject: New Smalltalk/X for Next (68k) Date: Tue, 12 May 1998 11:51:40 +0200 Organization: Informatik, Uni Stuttgart, Germany Message-ID: <35581BAC.5C3C@diapool.informatik.uni-stuttgart.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit New Smalltalk/X for Next (68k) ------------------------------ The latest demo release of Smalltalk/X is available from ftp.informatik.uni-stuttgart.de/pub/stx. Before downloading read carefully README.DEMO_DISTRIB, README.NEXT and README. Thanks, Pierre. peschwar@diapool.informatik.uni-stuttgart.de
From: see-below@not-my-address.com (Matthew Vaughan) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Tue, 12 May 1998 04:03:06 -0700 Message-ID: <see-below-1205980403060001@209.24.240.56> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <SCOTT.98May11153549@slave.doubleu.com> In article <SCOTT.98May11153549@slave.doubleu.com>, scott@doubleu.com (Scott Hess) wrote: > I think that why Rhapsody got little mention in the keynote was simply > to keep the MacOS contingent calm. My biggest non-hint is that at no > point did they mention a version of Carbon for Windows - which is a > tremendously obvious thing to not mention at such an event. My > suspicion is that Carbon is essentially a hugely improvd BlueBox, and > that something AppKit based is still where the medium to long term > future is. While I don't think it's directly related to Blue Box, it is certianly a replacement for it, and as such I don't see any more reason it would be available on Intel than the Blue Box was supposed to be. I assume that Yellow Box will still be the only cross-platform API. .................................................... MATTHEW VAUGHAN matthewv at best dot com (damn spammers...) http://www.best.com/~matthewv/ ....................................................
From: "Mike Lockwood" <lockwood@metrowerks.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Tue, 12 May 1998 09:05:05 -0400 Organization: Shore.Net/Eco Software, Inc; (info@shore.net) Message-ID: <6j9h1h$evm@fridge.shore.net> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> I'm not at WWDC, but I heard a bit of the keynote over the web. It sounds like carbon is merely a crutch to help people port their existing MacOS apps to run outside of the Blue Box. If it is a subset of the existing Mac APIs, I can't view it as a good long term solution. The existing Mac APIs are inherently not thread safe. For example, you can't have two threads drawing at the same time as long as the graphics API has functions like GetPort and SetPort. Similarly, the memory and resource managers have MemError and ResError. This can not be fixed in the OS without moving to a new API. So if we are ever to really take advantage of multitasking on MacOS, we need a new API. In the past, Apple claimed that "Gershwin" (which was supposed to come after Copland) would provide a new threadsafe API that would allow us to take full advantage of preemptive multithreading. My assumption is that the yellow box still the next generation API that we should use if we want to take full advantage of a modern OS. Does anyone know if this is true? Is the yellow box threadsafe at the UI level? Mike Nathan Urban wrote in message <6j7te7$vah$1@crib.bevc.blacksburg.va.us>... >In article <35575DA8.B7A1148F@milestonerdl.com>, mark@milestonerdl.com wrote: > >> Nathan Urban wrote: > >> > I think we need to wait a bit longer and learn more about Apple's plans. > >> A day? A week? A month? or 3 years? > >How about a day, when they have the talk on Rhapsody and Yellow Box? > >> > but it doesn't sound like Yellow Box is being dropped. > >> But there is no conformation of the status of YellowBox staying either. > >Scott Anguish says otherwise.. > > http://www.stepwise.com/SpecialCoverage/WWDC98/index.html > >though I don't know where he got his information from.
From: don.brown@cesoft.com (Donald Brown) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Tue, 12 May 1998 08:59:58 -0500 Organization: Prairie Group Message-ID: <MPG.fc211dc557e46b59896b4@news.supernews.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <stevenj-ya02408000R1105982106140001@news.mit.edu> <6j899s$g46$2@newsfep4.sprintmail.com> <MPG.fc17e6f707c72619896ac@news.supernews.com> <6j8iaj$be$1@crib.bevc.blacksburg.va.us> In article <6j8iaj$be$1@crib.bevc.blacksburg.va.us>, nurban@crib.bevc.blacksburg.va.us says... > I don't think anyone is arguing that the Mac API is dead -- it's > certainly going to be used for a long time to come in the existing Mac > software shops. It's just that for someone starting out _new_ without > any legacy code, it's better to pay attention to Yellow since the APIs > are a lot better. > The only reason I would push Mac API now instead of Yellow, is because right now, you can do a Mac API app and have a bunch of people use it. Frankly, the API is secondary--give me a programmer with a good sense of interface and good error control and such, and I can train him on a new API. The only way to learn a good sense of interface and error control is through failure. Doing things, letting people use them, hear their complaints and see what went wrong, repeat many times. Or, as Jobs once said, "Artists ship." Donald
From: jason@jhste1.dyn.ml.org (Jason S.) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 12 May 1998 15:23:32 GMT Organization: I don't think so Message-ID: <slrn6lgqih.6a2.jason@jhste1.dyn.ml.org> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <stevenj-ya02408000R1105982106140001@news.mit.edu> <6j899s$g46$2@newsfep4.sprintmail.com> <MPG.fc17e6f707c72619896ac@news.supernews.com> <6j8iaj$be$1@crib.bevc.blacksburg.va.us> <MPG.fc211dc557e46b59896b4@news.supernews.com> Donald Brown wrote: >> I don't think anyone is arguing that the Mac API is dead -- it's >> certainly going to be used for a long time to come in the existing Mac >> software shops. It's just that for someone starting out _new_ without >> any legacy code, it's better to pay attention to Yellow since the APIs >> are a lot better. >The only reason I would push Mac API now instead of Yellow, is because >right now, you can do a Mac API app and have a bunch of people use it. >Frankly, the API is secondary--give me a programmer with a good sense of >interface and good error control and such, and I can train him on a new >API. >The only way to learn a good sense of interface and error control is >through failure. Doing things, letting people use them, hear their >complaints and see what went wrong, repeat many times. >Or, as Jobs once said, "Artists ship." Or was that "Artists ship, con men keep pushing back the ship date by 18 months..." -- If FreeBSD actually did that, I would concede that FreeBSD was any more "correct" than Linux is, but not even the FreeBSD people can justify that kind of performance loss. -- Linus Torvalds on comp.unix.advocacy
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.next.advocacy,comp.sys.mac.advocacy,comp.sys.next.programmer Subject: WWDC: Rhapsody BoF Date: 12 May 1998 15:09:02 GMT Organization: P & L Systems Message-ID: <6j9ome$1cc$10@ironhorse.plsys.co.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit There will be a Rhapsody BoF (*) meeting on Wednesday evening in room J2, inside the conference centre, organised (if I understand correctly) by Ernie Prabhakar (Rhapsody Product Marketing Manager), Jordan Dea-Mattson (of Apple whom you should all know from his contributions to the lists), and Paul Lynch (P & L Systems' MD). Watch out for more details later and at the show. Best wishes, mmalc. P & L Systems -- developers of Mesa http://www.plsys.co.uk/plsys/ Tel: +44 1494 432422 Fax: +44 1494 432478 (* Birds of a Feather)
From: M Rassbach <mark@milestonerdl.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Tue, 12 May 1998 10:38:54 -0500 Organization: Inc.Net http://www.inc.net Message-ID: <35586D0E.732E4BB3@milestonerdl.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <SCOTT.98May11153549@slave.doubleu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Scott Hess wrote: > But there is no conformation of the status of YellowBox staying > either. > And I'm still waiting for that conformation. > NeXT and Apple have always been known for their abilities in > promulgating a very simple message that confuses the hell out of > everyone. And they are doing it again. :-) > That said, I can't believe they'd have such a high > percentage of the conference devoted to Rhapsody sessions if they > intended to dump Rhapsody in a year. What if this whole Carbon thing was thought up at the 'last minute'?
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 12 May 1998 12:49:19 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6j9uif$1op$1@crib.bevc.blacksburg.va.us> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> In article <6j9h1h$evm@fridge.shore.net>, "Mike Lockwood" <lockwood@metrowerks.com> wrote: > So if we are ever to really take advantage of multitasking on MacOS, we need > a new API. In the past, Apple claimed that "Gershwin" (which was supposed > to come after Copland) would provide a new threadsafe API that would allow > us to take full advantage of preemptive multithreading. My assumption is > that the yellow box still the next generation API that we should use if we > want to take full advantage of a modern OS. There are plenty of reasons to use the Yellow API other than threading. But as to threading specifically: the Foundation Kit is threadsafe. The AppKit, including the UI stuff, is not. (However, you can often use distributed objects to make the necessary mutexing transparent.) I've heard rumors that they may try to make it threadsafe, but so far nothing official.
From: andyba@corp.webtv.net (Andy Bates) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Tue, 12 May 1998 09:16:21 -0700 Organization: WebTV Networks Message-ID: <andyba-ya02408000R1205980916210001@news> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <6j7sdr$4gn$1@newsfep4.sprintmail.com> <6j7tij$vb7$1@crib.bevc.blacksburg.va.us> <stevenj-ya02408000R1105982039000001@news.mit.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <stevenj-ya02408000R1105982039000001@news.mit.edu>, stevenj@alum.mit.edu (Steven G. Johnson) wrote: > nurban@vt.edu wrote: > >macghod@concentric.net wrote: > >> Another thing, will carbon based apps run on rhapsody for intel? > > > >Good question. My guess would be "no", but there might be a MacOS X for > >Intel down the road.. > > My guess would be "yes." Think of MacOS X as Rhapsody with Latitude++ > (Carbon). Both Rhapsody and Latitude *already* support multiple processors > natively. That is too much of an advantage for Apple to throw it away for > no reason. No, they already said that carbon-based apps would NOT run on Rhapsody for Intel. Specifically, the APIs are still PowerPC-only, and hence not runnable on Intel. Andy Bates.
From: macghod@concentric.net Newsgroups: comp.sys.next.programmer Subject: events Date: 12 May 1998 17:01:49 GMT Message-ID: <6j9v9t$l39$2@newsfep4.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Does yellowbox programming have events? I have only done one app, basically the same thing as currency converter, and I didnt have to use a event loop and was amazed that I didnt! While it was just a simple app for help with taxes (5 fields to total income taxes, a add button, and a total field) with the mac it would take: a bunch of lines to initialize the toolbox. a event loop fuctions to take care of mouse clicks functions to take care of keyboard clicks Plus other functions. BUt with OS, I didnt have to do any of that!! And I could copy and paste, even tho I didnt have any code for it! Could even paste using the command keys -- Running on Openstep 4.2, the best os in the world, and about to get better! Mac/openstep qa work desired, email me for resume NeXTMail and MIME OK!
From: M Rassbach <mark@milestonerdl.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Tue, 12 May 1998 11:08:16 -0500 Organization: Inc.Net http://www.inc.net Message-ID: <355873F0.6B86BC5C@milestonerdl.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <6j7sdr$4gn$1@newsfep4.sprintmail.com> <6j7tij$vb7$1@crib.bevc.blacksburg.va.us> <stevenj-ya02408000R1105982039000001@news.mit.edu> <andyba-ya02408000R1205980916210001@news> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Andy Bates wrote: > No, they already said that carbon-based apps would NOT run on Rhapsody for > Intel. Specifically, the APIs are still PowerPC-only, and hence not > runnable on Intel. Is there a TECHNICAL reason, or just political?
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 12 May 1998 09:20:01 -0700 Organization: Primenet Services for the Internet Message-ID: <B17DC4C3-33EF6@206.165.43.129> References: <6j8iaj$be$1@crib.bevc.blacksburg.va.us> nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Nathan Urban <nurban@crib.bevc.blacksburg.va.us> said: > It's just that for someone starting out _new_ without >any legacy code, it's better to pay attention to Yellow since the APIs >are a lot better. Except for graphics (unless Mike & company show something really spiffy, of course). ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: scott@eviews.com (Scott Ellsworth) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Tue, 12 May 98 17:04:57 GMT Organization: QMS Message-ID: <6j9ve8$g6d$1@news01.deltanet.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <B17CCA7D-359B7@206.165.43.153> In article <B17CCA7D-359B7@206.165.43.153>, "Lawson English" <english@primenet.com> wrote: >It's an interesting issue, this fragmentation of the MacOS market into ever >smaller pieces. But it was in the name of consolidation <irony...> >68K/PPC System 7.1 & later. >68040/PPC MacOS 8.0 >PPC MacOS 8.1 >PPC MacOS 8.2 and better (Carbon-based) >Yellow Box (Rhapsody, 8.x + libraries (?), WIntel 95/98/NT and MacOS X). In theory, carbon-based apps will have a shared lib that will let them run correctly on PPC 8.0 and 8.1. Check tidbits though, I may have misread it. There were also indications that non Carbon 8.x apps would run on Rhapsody, just in a Blue Box, so it looks like we really have three markets: Unsupported OS/machines that cannot run Carbon apps Supported MacOS only apps that are rewritten for Carbon, running on any PPC with system 8 or higher, as well as Rhapsody Forward looking (irony) yellow box apps for Rhapsody and Intel, and OS 10 someday in the distant future. In all fairness, it appears that they have about as good a plan as can be created, given that they MUST fiddle with the OS on a deep level, and the YB RAD tools are just not really going to play nice with the old MacOS apps. It would be very difficult to reimplement the old event/windowing model API under the YB, and get anything a developer would want. Unfortunately, there are important apps that are not going to be rewritten anytime soon, so they need some kind of way to get MacOS apps under Rhapsody as soon as possible, but in a maintainable way. As an example, QMS is doing diddly squat with thier Mac version, so I can say with absolute certainty that there will not be a YB implementation of our app, unless YB becomes a better looking tool for Windows development than VC++, and that is not likely. (It is not, after all from the Evil Empire.) I expected that to mean that our product would just die when the time came, unless the Blue Box was really good, and even then, it would be enough of a second class citizen that few would buy it. Now, with Carbon, it is possible for me to turn it into a functioning app on my own time. From where I sit, this may well be a good thing. My primary focus for new development on the MacOS is going to be in Java, using pieces of the YB as appropriate, since I cannot just develop for the MacOS world, but that is the world I prefer to work in. >In my opinion, for all but a tiny fraction of Apple's developers, this is >the most schizophrenic scheme possible. Everyone but the upper-crust, >high-end developers who target those who purchase the latest and greatest >of everything are going to be scratching their heads on this one. Not really, imho. Those with an installed base or development meant to terminate in a product anytime soon are going to be using Carbon APIs, those with a substantial 68K market are going to be targetting system 7, and those with the luxury of a distant ship date, or the dual platform requirement are going to be using the YB. This is not a new decision for many Mac developers. We agonized over dropping system 6 support, and finally did it. We were not happy, but it became needed for some of our tools to work right, and it would cost more than we wnated to spend to keep it. Scott Scott Ellsworth scott@eviews.com "When a great many people are unable to find work, unemployment results" - Calvin Coolidge, (Stanley Walker, City Editor, p. 131 (1934)) "The barbarian is thwarted at the moat." - Scott Adams
From: jim@ergotech.com Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Tue, 12 May 1998 17:12:34 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6j9vu2$dqg$1@nnrp1.dejanews.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j87v9$sam$1@news.xmission.com> In article <6j87v9$sam$1@news.xmission.com>, don@misckit.com wrote: > I do. Let's just say someone at Apple in a position to know specifically > *asked* Scott to post that comment. That's all I can say. > I had hoped for a little better marketing strategy from Apple than we'd ever seen with NeXT. (Mind you, why? is a good question). I'll stack this away with the announcement of the shipping of a product (EOF) that required the signing of an NDA in order to find out what it. Whatever the real strategy is, the marketing message is a mess. Plus c'est la même chose. Jim PS This is not meant to belittle the efforts of Scott or stepwise who appear at this point to be _the_ definitive source of information at the moment. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.next.programmer Subject: Re: events Date: 12 May 1998 17:33:23 GMT Organization: MiscKit Development Message-ID: <6ja153$fcj$1@news.xmission.com> References: <6j9v9t$l39$2@newsfep4.sprintmail.com> macghod@concentric.net wrote: > Does yellowbox programming have events? Emphatically yes. Take a look at the AppKit's NSEvent object. The NSApplication, NSResponder, and NSView classes are also all pertinent. > I have only done one app, basically the same thing as > currency converter, and I didnt have to use a event loop > and was amazed that I didnt! > While it was just a simple app for help with taxes (5 fields > to total taxes, a add button, and a total field) with the > mac it would take: > a bunch of lines to initialize the toolbox. > a event loop > fuctions to take care of mouse clicks > functions to take care of keyboard clicks > > Plus other functions. But with OS, I didnt have to do any of that!! > And I could copy and paste, even tho I didnt have any code for it! Could > even paste using the command keys Yep. The YellowBox does _all_ this for you. It can do a lot more for you, too, which is why the "NeXT old guard" keeps insisting that YB is so darned cool. Believe me, you have only barely scratched the surface! There are events, and there even is an event loop. As you dig deeper, you'll find that you can hook into that whole process and even modify it, if you need to. But, and this is consistent throughout the YB environment, you don't _have_ to bother with those details if you don't want to. You only need to get your hands as dirty as the task at hand requires. Another way to put it is that you focus on your app and what makes it special, and you don't bother with the trivial stuff that can be confusing, lead to errors, etc. An added side benefit is that the AppKit helps your apps comply with the GUI guidelines, since it follows them itself. In order to deviate from the GUI guidelines, you have to actively override things--which should in theory help make YB apps much more consistent amongst each other. -- Later, -Don Yacktman don@misckit.com <a href="http://www.misckit.com/don.html">My home page</a>
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 12 May 1998 17:43:56 GMT Organization: MiscKit Development Message-ID: <6ja1os$fcj$2@news.xmission.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <6j9uif$1op$1@crib.bevc.blacksburg.va.us> nurban@crib.bevc.blacksburg.va.us (Nathan Urban) wrote: > There are plenty of reasons to use the Yellow API other than threading. > But as to threading specifically: the Foundation Kit is threadsafe. Yes. > The AppKit, including the UI stuff, is not. (However, you can often > use distributed objects to make the necessary mutexing transparent.) > I've heard rumors that they may try to make it threadsafe, but so far > nothing official. From http://www.stepwise.com/SpecialCoverage/WWDC98/Monday.html ] The OS improvements in DR2 include [...]. ] The changes to the Yellow Box frameworks include: [...] ] Multi-threading Support -- AppKit is finally fully multi-threaded So there you have it. Time to update our knowledge bases, folks. Thread time is here *now* in full force with DR2! And I'll just bet you that this has been done as a prelude to SMP support... -- Later, -Don Yacktman don@misckit.com <a href="http://www.misckit.com/don.html">My home page</a>
From: John Jensen <jjens@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 12 May 1998 11:07:01 -0700 Organization: Primenet (602)416-7000 Message-ID: <6ja345$i2l@nntp02.primenet.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <SCOTT.98May11153549@slave.doubleu.com> <35586D0E.732E4BB3@milestonerdl.com> M Rassbach <mark@milestonerdl.com> wrote: : > That said, I can't believe they'd have such a high : > percentage of the conference devoted to Rhapsody sessions if they : > intended to dump Rhapsody in a year. : What if this whole Carbon thing was thought up at the 'last minute'? Go to the Apple website and scan the developer newsletters for the last six months. Compare the levels of development in Mac API and Yellow Box. John
From: float@interport.net Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Yellowbox programming, and ragosta Followup-To: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Date: 12 May 1998 14:43:49 -0400 Organization: Interport Communications Corp. Message-ID: <6ja595$49m$1@interport.net> References: <6j2qnb$r2p$2@newsfep1.sprintmail.com> <1d8q5gj.17qnilxg9rrdzN@pppsl552.chicagonet.net> Why not just run OPENSTEP on a non-broken architecture? I haven't seen an operating system that runs really well on a PC, because the PC is a piece of shite. Edwin E. Thorne (ethorne@chicagonet.net) wrote: : 1. Raise your PC above your head with both hands. : 2. Throw it down against a cement floor as hard as you can. : 3. Sweep up the pieces. : 4. Obtain another PC and resume with step # 1. : 5 Repeat steps 1 through 4 until Openstep works as desired, or until : all available PCs are exhausted, whichever comes first. -- Ben <Just Another System Administrator>
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 12 May 1998 15:21:01 -0700 Organization: Primenet Services for the Internet Message-ID: <B17E1956-A8944@206.165.43.53> References: <6j9uif$1op$1@crib.bevc.blacksburg.va.us> nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Nathan Urban <nurban@crib.bevc.blacksburg.va.us> said: >There are plenty of reasons to use the Yellow API other than threading. >But as to threading specifically: the Foundation Kit is threadsafe. >The AppKit, including the UI stuff, is not. (However, you can often >use distributed objects to make the necessary mutexing transparent.) >I've heard rumors that they may try to make it threadsafe, but so far >nothing official. I've heard that Sun's implementation of OpenStep is threadsafe but it isn't very pretty. ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: SnowSim@halcyon.com (Simulated Snow) Newsgroups: comp.sys.next.programmer Subject: Re: what exactly is Smalltalk ? Date: Tue, 12 May 1998 14:18:23 -0700 Organization: Baloney Research Institute Message-ID: <SnowSim-1205981418230001@blv-ux100-ip15.nwnexus.net> References: <1998050822010800.SAA10901@ladder01.news.aol.com> In article <1998050822010800.SAA10901@ladder01.news.aol.com>, rafkah@aol.com (Rafkah) wrote: > I've been talking about Smalltalk a lot. what is it exactly ? what are his > advantages against oc, c++ or java ? the "root" of all that is good in computer software...{< : only slightly kidding -- "I know of no safe depository of the ultimate powers of the society but the people themselves; and if we think them not enlightened enough to exercise their control with a wholesome discretion, the remedy is not to take it from them but to inform their discretion." - Thomas Jefferson
From: float@interport.net Newsgroups: comp.sys.next.programmer,comp.sys.next.advocacy Subject: Re: whats up with zoneraiders?!?/! Followup-To: comp.sys.next.programmer,comp.sys.next.advocacy Date: 12 May 1998 18:33:40 -0400 Organization: Interport Communications Corp. Message-ID: <6jaio4$lon$1@interport.net> References: <6j04ej$nvc$1@newsfep4.sprintmail.com> macghod@concentric.net wrote: : Zoneraiders is out, and it is for rhapsody for ppc ONLY. I thought : something written for rhapsody would work on BOTH rhapsodies? And openstep : as well? Why in gods name would zoneraiders be ppc only? OPENSTEP apps are supposed to work on Rhapsody, but I haven't heard anything suggesting the reverse is true. : A while ago on a thread on the developer program, I said I thought : programmers would need rhapsody to develop for rhapsody, and the openstep : programmers wondered why everyone would think apps have to be written : differently to work on rhapsody. Do they or dont they? The official word is that OPENSTEP apps will run on Rhapsody. I don't know whether that's source compatibility or binary compatibility. -- Ben <Just Another System Administrator>
Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer From: spambait@seemysig.com (Mike Cohen) Subject: Re: future of mac programming Message-ID: <B17E5CD896688D35A@0.0.0.0> Organization: ISIS International References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> Date: Wed, 13 May 1998 00:08:58 GMT NNTP-Posting-Date: Tue, 12 May 1998 20:08:58 EST In article <MPG.fc14db42395aa139896a2@news.supernews.com>, don.brown@cesoft.com (Donald Brown) wrote: | Of course, last year's WWDC, the three big things were CHRP, | Newton/eMate, and Rhapsody. The first two are dead and the third is | AWOL. So, you need to be flexible. It seems that for the last 2-3 WWDCs, everything Apple proclaimed hot died before the year was out. I hope Carbon will be different. -- Mike Cohen - mike_cohen (at) pobox (dot) com - http://pobox.com/~macguru Sound is the same for all the world - Youssou N'dour, "Eyes Open"
From: macghod@concentric.net Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Rhapsody dead?? Date: 13 May 1998 00:35:38 GMT Message-ID: <6japsq$nt1$1@newsfep3.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NO! As I see it their are 2 possibilities 1) Apple has big plans for yellowbox, things are all steam ahead for yellowbox, and jobs said what he did merely to appease mac programmers. BUT YELLOWBOX *IS* THE FUTURE 2) This one disturbs me. Apple is hedging its bets. Steve wants to see how both will do, and whichever gives apple more money will not be "steved", but the yellowbox may be steved depending how things go. People are saying how apple yellowbox engineers are stressing #1, the only problem is what these engineers say isnt binding. The opendoc engineers said the same thing about opendoc, and look what happened to it. The engineers want yellowbox to be the future, but that does not prevent them from getting steved. I am hoping that this move isnt Steve deemphasising Rhapsody, so if it doesnt have instant success he can "steve" it. What did steve say in the wired article? About how he would make decissions only on how much money they make? -- Running on Openstep 4.2, the best os in the world, and about to get better! Mac/openstep qa work desired, email me for resume NeXTMail and MIME OK!
Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer From: spambait@seemysig.com (Mike Cohen) Subject: Re: future of mac programming Message-ID: <B17E5CDC96688D440@0.0.0.0> Organization: ISIS International References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <1d8waq9.1t0soxr6xzkjtN@rhrz-isdn3-p44.rhrz.uni-bonn.de> Date: Wed, 13 May 1998 00:09:04 GMT NNTP-Posting-Date: Tue, 12 May 1998 20:09:04 EST In article <1d8waq9.1t0soxr6xzkjtN@rhrz-isdn3-p44.rhrz.uni-bonn.de>, schuerig@acm.org (Michael Schuerig) wrote: | Carbon will fully support the Apple Event Manager. Standard Apple events | previously defined by Apple--for example, the required suite of Apple | events discussed in Inside Macintosh: Interapplication | Communication--will be supported in Mac OS X and play the same roles as | they do in Mac OS 8. See also "Open Scripting Architecture" (page 66) | for information on scripting support. Only if you pass the ACTUAL data in the AppleEvent, rather than doing something like passing pointers to data in an AppleEvent (I've seen some applications which do that - it's really gross). On a similar note, they also say you shouldn't use Gestalt to pass pointers between applications. -- Mike Cohen - mike_cohen (at) pobox (dot) com - http://pobox.com/~macguru Sound is the same for all the world - Youssou N'dour, "Eyes Open"
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer,comp.sys.next.advocacy Subject: Re: whats up with zoneraiders?!?/! Date: 12 May 1998 20:24:25 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6jap7p$2hb$1@crib.bevc.blacksburg.va.us> References: <6j04ej$nvc$1@newsfep4.sprintmail.com> <6jaio4$lon$1@interport.net> In article <6jaio4$lon$1@interport.net>, float@interport.net wrote: > OPENSTEP apps are supposed to work on Rhapsody, but I haven't heard > anything suggesting the reverse is true. Correct; it isn't. > The official word is that OPENSTEP apps will run on Rhapsody. I don't > know whether that's source compatibility or binary compatibility. Source.
From: Pascal Bourguignon <pjb@imaginet.fr> Newsgroups: comp.sys.next.programmer Subject: Re: Strange event Date: 13 May 1998 01:16:12 GMT Organization: Deficiente Message-ID: <6jas8s$erk$3@news.imaginet.fr> References: <6jamu4$7n9$1@Masala.CC.UH.EDU> mezzino@gauss.cl.uh.edu (Michael Mezzino) wrote: >Can anyone tell me how to find the cause of the following runtime error: > >May 12 18:18:40 [1670] Uncaught exception: NSInvalidArgumentException. >return code = 100003. *** NSInlineCString doesNotRecognizeSelector: >-string > >I do not have any string methods in my app, so I assume that NSString is >somehow causing this. It occurs when I double click in an NXTableView. >Also, if I monitor the event flow through the sendEvent:(NXEvent *)event >method, the error occurs between the single click and the double click. I >have tried to use gdb setting breakpoints around the critical section of >code. The error does not occur in single step mode but does occur in run >mode. Has anyone ever experienced anything like this? Thanks so much. > >Mike Mezzino >mezzino@gauss.cl.uh.edu It should be +string. So, I guess, may be you sent a string allocator (stringWithFormat:, ...) to a string _object_ (self) instead of a string _class_. -- __Pascal Bourguignon__ | The box said 'Requires Windows 95, mailto:pjb@imaginet.fr | or better.' So I bought a Macintosh.
From: mezzino@gauss.cl.uh.edu (Michael Mezzino) Newsgroups: comp.sys.next.programmer Subject: Strange event Date: 12 May 1998 23:45:08 GMT Organization: University of Houston Message-ID: <6jamu4$7n9$1@Masala.CC.UH.EDU> Can anyone tell me how to find the cause of the following runtime error: May 12 18:18:40 [1670] Uncaught exception: NSInvalidArgumentException. return code = 100003. *** NSInlineCString doesNotRecognizeSelector: -string I do not have any string methods in my app, so I assume that NSString is somehow causing this. It occurs when I double click in an NXTableView. Also, if I monitor the event flow through the sendEvent:(NXEvent *)event method, the error occurs between the single click and the double click. I have tried to use gdb setting breakpoints around the critical section of code. The error does not occur in single step mode but does occur in run mode. Has anyone ever experienced anything like this? Thanks so much. Mike Mezzino mezzino@gauss.cl.uh.edu
From: macghod@concentric.net Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Macos rumors quote Date: 13 May 1998 04:08:28 GMT Message-ID: <6jb6bs$nag$1@newsfep4.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I was disturbed by this quote on macos rumors. Of course since the apple employees (on irc) are not identified, I guess it could just be BS? Here is what they said: MOSR: That's one big worry that people have -- the perceived "death" of Rhapsody, and more specifically, the Yellow Box. What is Apple going to do with all that good technology? Tron: Well, Steve touched on that yesterday. Rhapsody 1.0 (aka CR1) ships sometime this fall, and then all the Next'ers move to the X effort. Rhapsody just didn't make sense for us. It was good technology, but it wasn't Apple, and making it Apple-ified would have taken a heck of a lot longer than OS X is going to take. I think that Sark has more to say about the YB.... -- Running on Openstep 4.2, the best os in the world, and about to get better! Mac/openstep qa work desired, email me for resume NeXTMail and MIME OK!
From: M Rassbach <mark@milestonerdl.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Macos rumors quote Date: Tue, 12 May 1998 22:32:16 -0500 Organization: Inc.Net http://www.inc.net Message-ID: <3559143F.159F8947@milestonerdl.com> References: <6jb6bs$nag$1@newsfep4.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit And as a counter quote: Rhapsody / Yellow Box is not dead, dying or withering. No matter what the instant analysis says. And, a lot of people missed some major points on this. There are other shoes to drop and we're going to let them drop. Unless we get in a pique and release that Report RFI sent out today. http://www.pelagius.com/AppleRecon/ And loyal Scott Anguish says all is well. When will Apple publish its own position on this matter? macghod@concentric.net wrote: > I was disturbed by this quote on macos rumors. Of course since the apple > employees (on irc) are not identified, I guess it could just be BS? Here is > what they said: > > MOSR: That's one big worry that people have -- the perceived "death" of > Rhapsody, and more specifically, the Yellow Box. What is Apple going to do > with all that good technology? > > Tron: Well, Steve touched on that yesterday. Rhapsody 1.0 (aka CR1) ships > sometime this fall, and then all the Next'ers move to the X effort. Rhapsody > just didn't make sense for us. It was good technology, but it wasn't Apple, > and making it Apple-ified would have taken a heck of a lot longer than OS X > is going to take. I think that Sark has more to say about the YB.... > > -- > Running on Openstep 4.2, the best os in the world, and about to get better! > Mac/openstep qa work desired, email me for resume > NeXTMail and MIME OK!
From: rbrezden@japan.ml.com (Richard Brezden) Newsgroups: comp.sys.next.programmer Subject: MiscTableScroll (OpenStep 4.2 NT) "class not loaded" error Date: Wed, 13 May 1998 04:38:18 GMT Organization: Merrill Lynch Message-ID: <6jb83u$757$1@news.ml.com> I'm new to OpenStep and have been battling this error for too long now: warning: May 13 13:34:15 ScrollTest[214] *** class error for 'MiscTableScroll': class not loadedd I get this error when I try to load a nib that contains a MiscTableScroll object dragged from the MiscTableScroll pallete. When debugging, the MiscTableScroll.dll does not appear in the list of dlls loaded. I have the Frameworks added to my project and the example ScrollDir program included in the MiscTableScroll source works flawlessly. Any clues what I'm missing? Thanks, Richard Brezden
Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.nsc.32k Subject: cmsg cancel <6jba7m$ler$435@kemer.cc.metu.edu.tr> ignore no reply Control: cancel <6jba7m$ler$435@kemer.cc.metu.edu.tr> Message-ID: <cancel.6jba7m$ler$435@kemer.cc.metu.edu.tr> Date: Wed, 13 May 1998 05:44:25 +0000 Sender: pcd@north-cyprus.com From: andrew@erlenstar.demon.co.uk Organization: Annihilator v0.3 Spam (EMP) cancelled - multiposted binary files BI=1179.5/1 SPAM ID=cSEEPtkOIrp0mAJ88cXl8A==
Message-ID: <35592E26.6D14BA54@unet.univie.ac.at> Date: Wed, 13 May 1998 07:22:47 +0200 From: Christian Benesch <a9226931@unet.univie.ac.at> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer,comp.sys.next.advocacy Subject: Re: whats up with zoneraiders?!?/! References: <6j04ej$nvc$1@newsfep4.sprintmail.com> <6jaio4$lon$1@interport.net> <6jap7p$2hb$1@crib.bevc.blacksburg.va.us> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Due to the OpenStep standard there should also be a backward combatibility of source code from Rhapsody to OS4.x.,I should think. Christian Benesch Nathan Urban wrote: > In article <6jaio4$lon$1@interport.net>, float@interport.net wrote: > > > OPENSTEP apps are supposed to work on Rhapsody, but I haven't heard > > anything suggesting the reverse is true. > > Correct; it isn't. > > > The official word is that OPENSTEP apps will run on Rhapsody. I don't > > know whether that's source compatibility or binary compatibility. > > Source.
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: 12 May 98 23:14:00 Organization: Is a sign of weakness Message-ID: <SCOTT.98May12231400@slave.doubleu.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <6j7sdr$4gn$1@newsfep4.sprintmail.com> <6j7tij$vb7$1@crib.bevc.blacksburg.va.us> <stevenj-ya02408000R1105982039000001@news.mit.edu> <andyba-ya02408000R1205980916210001@news> <355873F0.6B86BC5C@milestonerdl.com> In-reply-to: M Rassbach's message of Tue, 12 May 1998 11:08:16 -0500 In article <355873F0.6B86BC5C@milestonerdl.com>, M Rassbach <mark@milestonerdl.com> writes: Andy Bates wrote: > No, they already said that carbon-based apps would NOT run on > Rhapsody for Intel. Specifically, the APIs are still > PowerPC-only, and hence not runnable on Intel. Is there a TECHNICAL reason, or just political? Most likely, the Carbon API implementation will share with sections of the bluebox, which is to say it will be the same code as used in current MacOS, compiled with the same compilers, etc. Things will just be hooked together differently (I'm imagining various hardware shims, like used for the current bluebox). Meaning that they could IN THEORY port it to, say, Windows. But this would invoke a _huge_ round of SQA. Getting a demo running might only take a couple months, but getting a fully tested usable version would take years. It took NeXT a year and a half to go from the NS/FIP demo to NS3.2 for Intel Processors, and that was for a system who's core (Mach/BSD) was already extremely portable. It may feel like a political problem, but it's not. There literally aren't enough engineers to make Carbon run on Windows in any reasonable timeframe. [Carbon for Windows in 2001 is _not_ a reasonable timeframe,] -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 12 May 98 22:54:03 Organization: Is a sign of weakness Message-ID: <SCOTT.98May12225403@slave.doubleu.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <stevenj-ya02408000R1105982106140001@news.mit.edu> In-reply-to: stevenj@alum.mit.edu's message of Mon, 11 May 1998 21:06:14 -0400 In article <stevenj-ya02408000R1105982106140001@news.mit.edu>, stevenj@alum.mit.edu (Steven G. Johnson) writes: Sigh...I wish Jobs had been clearer on this point; it would have prevented 24 hours of hysteria. (Which is, I think, about as long as these worries will last.) _Any_ positive mention of Rhapsody in the keynote would have brought on tremendous hysteria! A couple hundred thousand MacOS developers have a _much_ louder voice than the couple hundred remaining NeXTSTEP developers. Jobs gave the MacOS people what they wanted/needed, and I'm willing to forgive his snubbing of the couple percent of NS/OS/Rhap developers, for now. [So long as the long term pans out,] -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 12 May 98 22:51:16 Organization: Is a sign of weakness Message-ID: <SCOTT.98May12225116@slave.doubleu.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <6j84dp$vln$1@crib.bevc.blacksburg.va.us> <MPG.fc15f0828d669629896aa@news.supernews.com> In-reply-to: don.brown@cesoft.com's message of Mon, 11 May 1998 20:17:04 -0500 In article <MPG.fc15f0828d669629896aa@news.supernews.com>, don.brown@cesoft.com (Donald Brown) writes: In article <6j84dp$vln$1@crib.bevc.blacksburg.va.us>, nurban@crib.bevc.blacksburg.va.us says... > Better advice: if you're writing a new app, forget the Mac API. > OpenStep is infinitely better. Might as well develop good habits > from the beginning, and not waste your time. Depends on what you're doing, and who your target market is. The word from this morning is, there is a future in programming for the Mac APIs. There also is a present. Even when Rhapsody ships, I have strong doubts how widely it will be accepted. Not certainties, doubts. What, pray tell, do you think MacOS X is? [To be fair, I'll tell you what _I_ think MacOS X is, coming from a NeXTSTEP background. I think MacOS X is Rhapsody with the "BlueBox without the Box" highly integrated. They've spun it out as "MacOS with Rhapsody technologies", but IMHO what we're hearing in the YellowBox and Rhapsody sessions boils down to "Rhapsody with MacOS technologies." From what is being said right now, if they ever ship MacOS X and Rhapsody in parallel, the differences will be entirely in the packaging,] -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer,comp.sys.next.advocacy Subject: Re: whats up with zoneraiders?!?/! Date: 13 May 1998 02:58:40 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6jbgb0$3g3$1@crib.bevc.blacksburg.va.us> References: <6j04ej$nvc$1@newsfep4.sprintmail.com> <6jaio4$lon$1@interport.net> <6jap7p$2hb$1@crib.bevc.blacksburg.va.us> <35592E26.6D14BA54@unet.univie.ac.at> In article <35592E26.6D14BA54@unet.univie.ac.at>, Christian Benesch <a9226931@unet.univie.ac.at> wrote: > Due to the OpenStep standard there should also be a backward combatibility > of source code from Rhapsody to OS4.x.,I should think. Things were added to the Yellow API that aren't in OpenStep (or even the 4.x enhanced OpenStep), and a few things were changed. So you'd have to be very careful of what you use.
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 13 May 1998 01:46:01 -0700 Organization: Primenet Services for the Internet Message-ID: <B17EABCF-8F758@206.165.43.165> References: <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Garry Roseman <memphis@macconnect.com> said: >My implementation required every object that was subclassed from >GRAPHIC, a generic graphic object, to have its own set of "QuickDraw >globals". A call to SetPort set the object's port, as calls to set the >pen width, color, pattern, and mode set the object's graphic state, not >any global state. When it came time for the underlying graphics driver >to actually write to the hardware the _driver_ compared the state of the >hardware to the state of the object and re-wrote the graphics registers >if, and only if, necessary. Of course, this is what a retained -mode API engine like GX is supposed to do (whether it does so or not is another matter entirely). As GX stands right now, every thread in a pre-emptive system would need to have its own GX graphics heap. The only problems that arise, I believe, involve specifying which heap is active and how to coordinate clipping between threads. Taligent graphics (which Apple holds a perpetual license to, I believe) delved into these matters deeper than anyone else, AFAIK, and if Apple really DOES decide to withdraw from DPS as the main graphics engine for MacOS X+, I think that everyone should examine what a combination of Taligent, GX and Rhapsody graphics could do. My little GXFCN external for HyperCard will supposedly be available in the earliest alpha version this Friday. I encourage everyone to download it and play with it. I'm hoping to get a protoyping GUI tool working in HyperCard ASAP after that. People should seriously consider Eric King's original suggestion that GX would make an ideal basis for a GUI and consider how a retained mode graphics system could be designed to be thread-safe, interupt safe, compatible with Carbon and YB needs and doable. Unless, of course, Mike and company have already done all of that and announce it on Thursday... ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: William Edward Woody <woody@alumni.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 13 May 1998 02:17:04 -0700 Organization: In Phase Consulting Message-ID: <35596510.F5627440@alumni.caltech.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <B17CCA7D-359B7@206.165.43.153> <6j9ve8$g6d$1@news01.deltanet.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Scott Ellsworth wrote: > In theory, carbon-based apps will have a shared lib that will let them > run correctly on PPC 8.0 and 8.1. Check tidbits though, I may have > misread it. There were also indications that non Carbon 8.x apps > would run on Rhapsody, just in a Blue Box, so it looks like we really > have three markets: > > Unsupported OS/machines that cannot run Carbon apps > > Supported MacOS only apps that are rewritten for Carbon, running on > any PPC with system 8 or higher, as well as Rhapsody > > Forward looking (irony) yellow box apps for Rhapsody and Intel, and > OS 10 someday in the distant future. But... If you write for Carbon, it's a recompile for MacOS 8. So for systems that cannot run Carbon apps, just follow the guidelines and recompile. Heck; from Apple's web site, it sounds like some services (like the Navigation services) will be backwards compatable (when compiled that way) all the way back to System 7.5.5. Okay, okay; so that means your application may likely ship with three different executables: a 68K executable, a PPC for MacOS8 (and earlier) executable, and a Carbon executable. But if you write the code correctly, all of these executables are built from the same source kit. It even sounds like it's possible that the Yellow Box API (which is really more of a framework than an API) may allow you to write the three different application executables above, as well as recompile for the Rhapsody YB environment and for Windows. (Assuming Apple doesn't drop the YB framework API.) So in the end--yes, there is significant fragmentation going on. But you'll be able to generate all of the necessary executables from the same source kit. Sounds to me like the only thing we'll have to worry about is being a little more careful with writing portable code, spend a little more time debugging the source kit on different platforms (like we do now with debugging both the 68K and PPC versions of our code), and add a couple of extra configuration steps to the Installer program. Beats the hell out of having to rewrite everything for the OS API of the week... - Bill -- William Edward Woody | In Phase Consulting woody@alumni.caltech.edu | Macintosh & MS Windows Development http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/
From: William Edward Woody <woody@alumni.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 13 May 1998 02:21:50 -0700 Organization: In Phase Consulting Message-ID: <3559662E.E37ABBE0@alumni.caltech.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <1d8waq9.1t0soxr6xzkjtN@rhrz-isdn3-p44.rhrz.uni-bonn.de> <B17E5CDC96688D440@0.0.0.0> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Mike Cohen wrote: > Only if you pass the ACTUAL data in the AppleEvent, rather than doing > something like passing pointers to data in an AppleEvent (I've seen some > applications which do that - it's really gross). On a similar note, they > also say you shouldn't use Gestalt to pass pointers between applications. Well, hell's bells; Apple has been telling us not to do this sort of grossness since Day One. If the reason why programmers have to modify 10% of thier existing applications is because of stuff like *this*, I dunno if there's anything to say to those programmers other than "I told you so," and "Better off doing it right the first time than having to redo it later." - Bill -- William Edward Woody | In Phase Consulting woody@alumni.caltech.edu | Macintosh & MS Windows Development http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: 13 May 1998 10:23:50 GMT Organization: Idiom Communications Message-ID: <6jbsbm$eua$12@news.idiom.com> References: <6jbldv$eua$4@news.idiom.com> <6jbpfl$rji$1@pump1.york.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: pete@ohm.york.ac.uk -bat. may or may not have said: --> If you really want to see a confusing method of addressing memory go take -> a look at transputers sometime with a signed address space - and it's -> relationship to OCCAM. Which is where a lot of our NeXTStep code originated. Really? That's news to me. I've never heard of any relationship between NeXTSTEP and OCCAM. Transputers were something I never really investigated: I was more into array processors, and the Novix FORTH engine at the time. -jcr
From: pete@ohm.york.ac.uk (-bat.) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: 13 May 1998 12:16:38 GMT Organization: The University of York, UK Sender: pcf1@york.ac.uk Message-ID: <6jc2v6$osr$1@pump1.york.ac.uk> References: <6jbsbm$eua$12@news.idiom.com> jcr.remove@this.phrase.idiom.com (John C. Randolph) writes: > Really? That's news to me. I've never heard of any relationship between > NeXTSTEP and OCCAM. Transputers were something I never really investigated: Uh, when I said "our NeXT code" I mean "our" as in myself and the people I work with, not "our" in the sense of the wider NeXT community. I got involved in NeXTs as we had an OCCAM2 application which did all it's own mouse, windows gui etc on custom hardware. I was working with a bloke called $an at the time (who used to hang about on here) and he one day said "oh, you could do all that gui stuff in an afternoon on a NeXT" - a sentance with lasting repercussiongs for the subsequent 6 years. The gui code got ditched, the rest of the OCCAM2 was re-written under UNIX and a front-end was duly knocked up. Took more than an afternoon, but it's proved to be a huge success and we've been working with the code ever since. Theres not a lot of evidence of the OCCAM2 left now - except for the fact that all my C code is written using a folding editor :-) The backend still parallelises quite nicely and will run across a network of NeXTs though. > I was more into array processors, and the Novix FORTH engine at the time. I never got to play witha Novix, but it's nice to think that features of FORTH are still going strong in the shape of PostScript. transputers are pretty much dead as general purpose computing devices these days I guess. Though my colleagues across the lake in computer science have a system which translates java byte code to transputer byte code and apparently they are an excellent processor for running java for embedded controllers on. -bat.
From: "Mike Lockwood" <lockwood@metrowerks.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 13 May 1998 08:58:14 -0400 Organization: Shore.Net/Eco Software, Inc; (info@shore.net) Message-ID: <6jc50h$eh@fridge.shore.net> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> Garry Roseman wrote in message <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com>... >Mike Lockwood <lockwood@metrowerks.com> wrote: > >> It sounds like carbon is merely a crutch to help people port their existing >> MacOS apps to run outside of the Blue Box. If it is a subset of the >> existing Mac APIs, I can't view it as a good long term solution. >> >> The existing Mac APIs are inherently not thread safe. For example, you >> can't have two threads drawing at the same time as long as the graphics API >> has functions like GetPort and SetPort. Similarly, the memory and resource >> managers have MemError and ResError. This can not be fixed in the OS >> without moving to a new API. > >This notion that the MacOS programming interface is inherently >incompatible with multitasking sounds authoritative as stated by Mike; >but I know by experience that he's wrong. I see no great problem in the >"API". The problem with the old Mac toolbox is in the implementation, >and implementations can be changed. That is what Apple has stated to be >their goal --- to reimplement the Mac toolbox using the same calling >conventions but with reentrant implentations. I don't see how the Mac API could be made thread safe. Suppose you have two threads, thread A, and thread B. Thread A calls SetPort(), and then starts drawing. At about the same time, thread B calls SetPort to a different port, and then starts drawing. The problem is a global current port, so both threads will end up drawing into the same port. With the Win32 API, on the other hand, all drawing is done via a device context. Two threads can draw simultaneously into different device contexts without interfering with each other. (I'm not trying to start a debate over the advantages or disadvantages of the MacOS vs. Win32 APIs. I'm just pointing out that the API would need to change in order to have thread safe graphics). There are other more serious examples of non-threadsafeness in the Mac APIs. Think about all the places in the API that use a handle. If one thread tries to allocate memory and causes a heap compaction while another thread is doing something with a handle in the same heap, you are in big trouble. I guess you could fix that without changing the API by making all handles permenantly locked, but of course that defeats the purpose of having handles in the first place. Mike
From: "Mike Lockwood" <lockwood@metrowerks.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 13 May 1998 09:04:13 -0400 Organization: Shore.Net/Eco Software, Inc; (info@shore.net) Message-ID: <6jc5br$so@fridge.shore.net> References: <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <B17EABCF-8F758@206.165.43.165> Lawson English wrote in message ... >My little GXFCN external for HyperCard will supposedly be available in the >earliest alpha version this Friday. I encourage everyone to download it and >play with it. I'm hoping to get a protoyping GUI tool working in HyperCard >ASAP after that. People should seriously consider Eric King's original >suggestion that GX would make an ideal basis for a GUI and consider how a >retained mode graphics system could be designed to be thread-safe, interupt >safe, compatible with Carbon and YB needs and doable. > >Unless, of course, Mike and company have already done all of that and >announce it on Thursday... Nope, not me. I'm not announcing anything on Thursday. And I haven't written a HyperCard XCMD since 1988... Mike
From: craig@mpv.com (Craig Halley) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Macos rumors quote Date: Wed, 13 May 1998 05:11:35 GMT Organization: MPV, Inc. Message-ID: <6jba3v$sbu$1@nnrp1> References: <6jb6bs$nag$1@newsfep4.sprintmail.com> >Tron: Well, Steve touched on that yesterday. Rhapsody 1.0 (aka CR1) ships >sometime this fall, and then all the Next'ers move to the X effort. Rhapsody >just didn't make sense for us. It was good technology, but it wasn't Apple, >and making it Apple-ified would have taken a heck of a lot longer than OS X >is going to take. I think that Sark has more to say about the YB.... If I remember correctly, what Steve said was more like this: "Rhapsody is good technology, great technology, but it didn't go far enough. So we're going to take it farther." Just the fact that Mac OS X will be based on Mach, with Terminal.app if you want it, is enough proof for me. Hooray for the command line :). I don't know if the Mach thing is first release or what, but that was definitely part of the "Rhapsody and Yellow Box in 1998" presentation that's available on the Apple WebTheater. I think the change is pretty simple. Instead of integrating Mac OS into Rhapsody as announced last year, they're going to integrate Rhapsody into the Mac OS. Pretty subtle... not that big of a deal for the NeXT people, since Yellow Box doesn't change, but it's a bit easier for the original Mac people, who don't have to go the Yellow API's right away. Also, the "box" of the Blue Box will be going away, so you'll be able to run Rhapsody apps, current Mac apps, and Carbon-based Mac apps all under Mac OS X, without any emulation or virtual environments. Sounds good to me! Craig Halley
From: memphis@macconnect.com (Garry Roseman) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 13 May 1998 00:30:41 -0500 Organization: Writer & Freelance Programmer Sender: macconr@pppa40-memphis9-2r431.saturn.bbn.com Message-ID: <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> Mike Lockwood <lockwood@metrowerks.com> wrote: > It sounds like carbon is merely a crutch to help people port their existing > MacOS apps to run outside of the Blue Box. If it is a subset of the > existing Mac APIs, I can't view it as a good long term solution. > > The existing Mac APIs are inherently not thread safe. For example, you > can't have two threads drawing at the same time as long as the graphics API > has functions like GetPort and SetPort. Similarly, the memory and resource > managers have MemError and ResError. This can not be fixed in the OS > without moving to a new API. This notion that the MacOS programming interface is inherently incompatible with multitasking sounds authoritative as stated by Mike; but I know by experience that he's wrong. I see no great problem in the "API". The problem with the old Mac toolbox is in the implementation, and implementations can be changed. That is what Apple has stated to be their goal --- to reimplement the Mac toolbox using the same calling conventions but with reentrant implentations. Let me give you an example. I implemented a graphics package for an embedded control system that used a Matrox graphics board on Intel hardware. I copied most of the interface from QuickDraw (!), with the coordinate transformations like QD GX. For example, I had QD-like SetPort and GetPort functions, but my implentation was for a fast multitasking control system with multiple threads within a process writing to the display simultaneously. Why not?? In fact, I allowed every object to execute within its own "task" and multiple objects could draw themselves simultaneously. When you do work like that you just decide on the design, such as a thread-safe QD look-alike, and then do whatever is necessary underneath to "make it so!" My implementation required every object that was subclassed from GRAPHIC, a generic graphic object, to have its own set of "QuickDraw globals". A call to SetPort set the object's port, as calls to set the pen width, color, pattern, and mode set the object's graphic state, not any global state. When it came time for the underlying graphics driver to actually write to the hardware the _driver_ compared the state of the hardware to the state of the object and re-wrote the graphics registers if, and only if, necessary. The low-level driver functions had critical sections which had to be non-divisible, of course. All multithreaded systems on conventional hardware have to use locks/semaphores and such to control access to the (non-reentrant) hardware. It's just a question of doing it. Some of the re-implementations have ripple effects that cause a lot of code to change at once but I see no reason not to believe that Apple can do it, especially since they're realistically committing to only PowerPC. If they do succeed, as I think they will, Carbon will be much more than a crutch. -- Garry Roseman <mailto:memphis@macconnect.com> Custom Programming & Scripting for Macintosh Memphis TN USA
From: mpaque@wco.com (Mike Paquette) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Macos rumors quote Date: Tue, 12 May 1998 23:12:51 -0700 Organization: Electronics Service Unit No. 16 Message-ID: <1d8xtgy.nhbnkgr2hlp0N@cetus181.wco.com> References: <6jb6bs$nag$1@newsfep4.sprintmail.com> <macghod@concentric.net> wrote: > I was disturbed by this quote on macos rumors. Of course since the apple > employees (on irc) are not identified, I guess it could just be BS? It's BS. And how do you verify that a random IRC yakkerr is an Apple Employee? Sniff for IP addresses of the form 17.*.*.*? -- Mike Paquette mpaque@wco.com "Troubled Apple Computer" and the "Troubled Apple" logo are trade and service marks of Apple Computer, Inc.
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 12 May 98 23:08:11 Organization: Is a sign of weakness Message-ID: <SCOTT.98May12230811@slave.doubleu.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> In-reply-to: "Mike Lockwood"'s message of Tue, 12 May 1998 09:05:05 -0400 In article <6j9h1h$evm@fridge.shore.net>, "Mike Lockwood" <lockwood@metrowerks.com> writes: I'm not at WWDC, but I heard a bit of the keynote over the web. <...> Does anyone know if this is true? Is the yellow box threadsafe at the UI level? The impression I'm getting is "sort of". They've stated that you can draw in multiple threads, but that you need seperate contexts. That's no problem. There's also been some back-and-forth on FoundationKit and AppKit being thread-safe. What it sounds like to me (I'm analyzing, here, not quoting anyone) is that they are for the most part thread-safe, except for documented sections that might not be. I get the impression that there are some areas which are considered part of the various kits, but just aren't sensibly multi-threaded. I can't offhand recall whether they remarked on specific areas - though they did remark on a number of specific types of things that _are_ thread-safe. [Above, where I say "context", I mean "connection to the windowserver." Seperate windows aren't sufficient, you have to have seperate connections. To a great extent this is an Adobe DPS problem, not an Apple problem. I recall seeing a method to cause automagic provision of an extra context for a thread you're launching off into the blue...] -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 12 May 98 23:28:55 Organization: Is a sign of weakness Message-ID: <SCOTT.98May12232855@slave.doubleu.com> References: <6j8iaj$be$1@crib.bevc.blacksburg.va.us> <B17DC4C3-33EF6@206.165.43.129> In-reply-to: "Lawson English"'s message of 12 May 1998 09:20:01 -0700 In article <B17DC4C3-33EF6@206.165.43.129>, "Lawson English" <english@primenet.com> writes: Nathan Urban <nurban@crib.bevc.blacksburg.va.us> said: >It's just that for someone starting out _new_ without any legacy >code, it's better to pay attention to Yellow since the APIs are a >lot better. Except for graphics (unless Mike & company show something really spiffy, of course). Just to throw you a bone ... in one session someone did mention QuickDraw as a possible future replacement for DPS. They phrased it as something like retaining the PostScript imagining model, but using QuickDraw as the renderer. Obviously, there wasn't a lot of info to be had on this (it sounded like a real blue-sky comment). Now surprise me - instead of jumping off on an "I told you so" rampage, why not present a reasonable prediction of various ways that you could more or less seamlessly use QuickDraw as the drawing layer for AppKit - which is to say using Apple's design patterns in Objective-C, as opposed to C-based APIs. And remember, they have the source, so they can rearrange the code and how it relates to itself arbitrarily, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: events Date: 12 May 98 23:24:22 Organization: Is a sign of weakness Message-ID: <SCOTT.98May12232422@slave.doubleu.com> References: <6j9v9t$l39$2@newsfep4.sprintmail.com> In-reply-to: macghod@concentric.net's message of 12 May 1998 17:01:49 GMT In article <6j9v9t$l39$2@newsfep4.sprintmail.com>, macghod@concentric.net writes: Does yellowbox programming have events? I have only done one app, basically the same thing as currency converter, and I didnt have to use a event loop and was amazed that I didnt! "Event" is such a broad term, which has been broadly bastardized. I generally think of "event" as something like a keydown or mousedown, but in some areas it's been stretched to also mean things like button pressed and slider moved. Though those are _literal_ events, it's useful to distinguish between composite or meta-level events and hard events like mouse clicks. That said, what does a target/action connection in InterfaceBuilder break down to? In many frameworks out there, a mouse click gets fed into the event loop, which feeds it to the view/window it happened in, which might turn it into a button-pressed event, which goes back to the same event loop but (hopefully) feeds out a different piece of the switch statement. Most attempt to cover this sequence with various sugar, but it's still ugly. Meanwhile, on YellowBox, the button distributes the button-pressed event as a message to the "appropriate" object. Essentially the event loop's giant switch statement is pushed down into the Objective-C runtime. This might seem somewhat restrictive - but, in like 99% of the time, there is a specific object that obviously should receive the message anyhow. In the other 1%, target/action is easy enough to handle by making the target's action a method which determines where to forward to. [Actually, I don't think I've _ever_ had such a method just forward the message. Invariably, that last 1% tends to be actions which simply require more glue than you can reasonably encode into a control-drag in InterfaceBuilder :-).] -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 12 May 98 22:44:38 Organization: Is a sign of weakness Message-ID: <SCOTT.98May12224438@slave.doubleu.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <B17CCA7D-359B7@206.165.43.153> In-reply-to: "Lawson English"'s message of 11 May 1998 15:32:01 -0700 Let's say that Apple went with a totally MacOS, no Rhapsody solution. Unless they just continued shipping MacOS8.1 right on into the grave, there'd be a new version. There would still be your 7.1, 8.0, 8.1, and Carbon (Copland, Taligent, whatever) breakout. You'd still have 4 "markets". But, beyond that, you've only listed three "markets", 7.1, Carbon, and Yellow box. And Carbon can also run on MacOS X, if you like, --- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed> In article <B17CCA7D-359B7@206.165.43.153>, "Lawson English" <english@primenet.com> writes: It's an interesting issue, this fragmentation of the MacOS market into ever smaller pieces. Here is what I see: For MacOS developers, there are now at least 5 markets: 68K/PPC System 7.1 & later. 68040/PPC MacOS 8.0 PPC MacOS 8.1 PPC MacOS 8.2 and better (Carbon-based) Yellow Box (Rhapsody, 8.x + libraries (?), WIntel 95/98/NT and MacOS X). Each developer is going to have to look at their installed base, their potential revenue from upgrades and new users, etc., and decide which market to target. In my opinion, for all but a tiny fraction of Apple's developers, this is the most schizophrenic scheme possible. Everyone but the upper-crust, high-end developers who target those who purchase the latest and greatest of everything are going to be scratching their heads on this one. Nothing new for Jobs. -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: Ken MacLeod <ken@bitsko.slc.ut.us> Newsgroups: comp.sys.next.programmer Subject: Java and distributed objects Date: 13 May 1998 09:41:02 -0500 Organization: PSINet Message-ID: <m3vhraqlmp.fsf@biff.bitsko.slc.ut.us> In the WWDC98 Tuesday Report on Stepwise <http://www.stepwise.com/SpecialCoverage/WWDC98/Tuesday-AppKit.html>, Scott Anguish wrote: > There are some APIs that are not exposed because they are already > available in Java, for example the Distributed Objects classes [...] Does this mean 1) that Java has classes for supporting Apple's (nee NeXT's) Distributed Objects, or 2) that Java has it's own way of doing this (like RMI) and doesn't need to use DO? I'm working on an internal protocol that's patterned after DO and having additional implementations like DO would help a lot. -- Ken MacLeod ken@bitsko.slc.ut.us
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 13 May 1998 10:47:00 -0700 Organization: Primenet Services for the Internet Message-ID: <B17F2AC3-51C24@206.165.43.172> References: <6jc50h$eh@fridge.shore.net> nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Mike Lockwood <lockwood@metrowerks.com> said: >I don't see how the Mac API could be made thread safe. Suppose you have >two >threads, thread A, and thread B. Thread A calls SetPort(), and then starts >drawing. At about the same time, thread B calls SetPort to a different >port, and then starts drawing. The problem is a global current port, so >both threads will end up drawing into the same port. With the Win32 API, >on >the other hand, all drawing is done via a device context. Two threads can >draw simultaneously into different device contexts without interfering >with >each other. See my comments about GX and Garry Roseman's (<memphis@macconnect.com>) comments about his QD-like library. You need to specify which thread is active (in the case of GX, you could modify the API slightly to associate a given heap with a specific thread) and coordinate clipping. The API would have to change to accomidate specifying (or somehow obtaining) the current thread/heap ID, but it would be doable. Memory management is likely no more tricky than that. Presumeably the Carbon APIs take care of everything for you. NOte that the Standard File Package is no more because apparently that was a lost cause, PMT-wise. Don't any of you recall Eric King's point, made several years ago, that it was the windows manager and QuickDraw that was holding up PMT? Everything else was fixable. Rhapsody provides a pre-emptive windows manager and apparently they've either fixed QD or have found a decent way of emulating it and hence have implemented Eric King's suggestion 2 years late. If you want more discussion on this, subscribe to the AIMED-talk mailing list, where various ways of implementing this kind of thing on MacOS have been going on for years. ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: Sterling <Sterling_Augustine@stanford.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 13 May 1998 10:50:02 -0700 Organization: Stanford CS Message-ID: <3559DD44.3914@stanford.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Mike Lockwood wrote: > > Garry Roseman wrote in message > <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com>... > >Mike Lockwood <lockwood@metrowerks.com> wrote: > > > >> It sounds like carbon is merely a crutch to help people port their > existing > >> MacOS apps to run outside of the Blue Box. If it is a subset of the > >> existing Mac APIs, I can't view it as a good long term solution. > >> > >> The existing Mac APIs are inherently not thread safe. For example, you > >> can't have two threads drawing at the same time as long as the graphics > API > >> has functions like GetPort and SetPort. Similarly, the memory and > resource > >> managers have MemError and ResError. This can not be fixed in the OS > >> without moving to a new API. > > > >This notion that the MacOS programming interface is inherently > >incompatible with multitasking sounds authoritative as stated by Mike; > >but I know by experience that he's wrong. I see no great problem in the > >"API". The problem with the old Mac toolbox is in the implementation, > >and implementations can be changed. That is what Apple has stated to be > >their goal --- to reimplement the Mac toolbox using the same calling > >conventions but with reentrant implentations. > > I don't see how the Mac API could be made thread safe. Suppose you have two > threads, thread A, and thread B. Thread A calls SetPort(), and then starts > drawing. At about the same time, thread B calls SetPort to a different > port, and then starts drawing. The problem is a global current port, so > both threads will end up drawing into the same port. So give each thread its own copy of the QT globals. Viola, the problem goes away. This will solve most of the "inherently" nonthread safe problems, including in the resource manager. > If one thread > tries to allocate memory and causes a heap compaction while another thread > is doing something with a handle in the same heap, you are in big trouble. > I guess you could fix that without changing the API by making all handles > permenantly locked, but of course that defeats the purpose of having handles > in the first place. Handles should eventually go away, because a modern memory manager should manage memory in such a way that compaction as it is understood in the Mac universe never occurs. Modern OS's don't have to worry about heap compaction because they solve the fragmentation problems through virtual address spaces mapped into physical pages. It requires hardware assistance, though. So the MacOS's reliance on them is an artifact of the original 68000 not having a hardware MMU. This technique is one of the first things you learn in an undergraduate operating systems course. So permanently locking handles would be no big deal. The function they used to perform _should_ be moved out of the programmers' domain into the OS where it belongs. Permanantly locking them would be a good step in that direction. You wouldn't even need to permanantly set the locked bit. You could just say that if you are running under an OS version greater than 10, handles will not be relocated. Sterling
From: simpson@nospam.cts.com (Michael Simpson) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Yellowbox programming, and ragosta Date: Wed, 13 May 1998 17:57:13 GMT Organization: QUALCOMM, Incorporated; San Diego, CA, USA Message-ID: <3559deb7.90979081@news> References: <6j2qnb$r2p$2@newsfep1.sprintmail.com> <6j487r$ibo$1@bignews.shef.ac.uk> On 10 May 1998 12:57:31 GMT, mmalcolm crawford <m.crawford@shef.ac.uk> wrote: >I am quite sure that you have some sort of installation problem. Did you >install all the relevant binaries? You should only be clicking boxes for >binaries you have installed -- My thought is: Why does the interface allow you to click on a check box for a product you do not have installed? >if you're trying to compile for Sparc, for >example, on Prelude to Rhapsody, and don't have the Sparc binaries installed, >you'll have problems. Similarly, if you're running RDR, as far as I could >tell from demos of Project Builder, it still has buttons for compiling on >NeXT and Sparc, and yet as far as I'm aware versions of RDR were never >produced for those platforms. So again, if you tried compiling for either of >those platforms, you'll be sunk.
From: macghod@concentric.net Newsgroups: comp.sys.next.programmer Subject: C++ and rhapsody Date: 13 May 1998 20:28:38 GMT Message-ID: <6jcvpm$25r$1@newsfep1.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I was looking over the programers tutorial for rhapsody (discovering rhapsody, a developers tuturial) and it said , IN EFFECT, "you can use c++, java, and object c, in WHOLE (my emphases) or in part. From what I have heard this is incorrect. THe correct verbiage would be "you can use java and object c, in whole or in part, and c++ in part". Is the manual incorrect, or will one be able to use only c++ with ib/pb in rhapsody? (I know c++, and object c is such a pain! I expect the object part to be different, but little things irritate me, like how you use : for arguments, when in c you use (). This isnt a object oriented feature, so why change it from how it was in c? -- Running on Openstep 4.2, the best os in the world, and about to get better! Mac/openstep qa work desired, email me for resume NeXTMail and MIME OK!
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: C++ and rhapsody Date: 13 May 1998 18:38:35 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6jd7db$555$1@crib.bevc.blacksburg.va.us> References: <6jcvpm$25r$1@newsfep1.sprintmail.com> In article <6jcvpm$25r$1@newsfep1.sprintmail.com>, macghod@concentric.net wrote: > I was looking over the programers tutorial for rhapsody (discovering > rhapsody, a developers tuturial) and it said , IN EFFECT, "you can use c++, > java, and object c, in WHOLE (my emphases) or in part. From what I have > heard this is incorrect. THe correct verbiage would be "you can use java and > object c, in whole or in part, and c++ in part". Is the manual incorrect, or > will one be able to use only c++ with ib/pb in rhapsody? You can use C++ in whole, but you can't access any of the OpenStep APIs. If you're interested in doing non-OpenStep, pure C++ development on Rhapsody, I would recommend g++ or egcs over the native compiler. > (I know c++, and object c is such a pain! Get used to it, it's better. > I expect the object part to be > different, but little things irritate me, like how you use : for arguments, > when in c you use (). This isnt a object oriented feature, so why change it > from how it was in c? Because messaging an object is NOT the same thing as calling a function in C. This is not like C++. There is a runtime messaging system. Because the Objective-C messaging model was based on Smalltalk, which has the : syntax. Because the : syntax is _superior_ to the () syntax. Why are you complaining about syntax just because it's _different_? It's trivial to use. Syntax should be judged on its own merits. And the : syntax is better because it lets you intersperse the method name with the arguments. Which is easier to discern the meaning of, when reading someone's code (_especially_ when the methods being called are unfamiliar to you): obj->plot(10,10,1.0,1.0,0.0) obj->plotAtXandYwithRGB(10,10,1.0,1.0,0.0) vs. [obj plotPointAtX:10 andY:10 withRed:1.0 Blue:1.0 Green:0.0] Or how about (taken mostly at random from OmniAppKit), with a variation on the indentation and whitespace conventions: obj->startDrag(aView,anEntry,view,image,location,event,pasteboardHelper) vs. [obj outlineView: aView startDragOnEntry: anEntry fromView: view image: image atPoint: location event: event pasteboardHelper: pasteboardHelper] With the Obj-C examples, you can pretty much tell what's going on just by reading this code (or at least, a lot better than you could with the C++ variant); no need to look up the methods in the API. And even the long-name variants of the C++ methods just look ugly, which encourages short, undescriptive names. Plus, it saves you from making parameter mistakes. Suppose you have a "move" method that takes source and destination parameters. Is it obj->move(src,dest) or obj->move(dest,src) ? I'm forgetful, and might have to look it up. Plus, if I did happen to make a mistake and swapped the two parameters, then it might not be too obvious. In Obj-C, you could just write [obj moveFrom:src to:dest]. If you forgot exactly how it went and tried to write [obj moveTo:dest from:src], the compiler would catch it. And you aren't likely to accidentally get careless and switch the arguments in [obj moveFrom:dest to:src]. There are other examples of mixing up parameters. Real life example: I was using a C conversion function that would take a number and convert it to a string, placing the result in a buffer being passed. I had an 10-digit number, the prototype was something akin to Convert(char*,int,int) -- the code was obj->Convert(buf,num,10). (Actually, it was a C function, but I'm just writing it like a method for illustration purposes.) Well, it turned out at some point that an 11-digit number was necessary after all in some extreme cases, so I simply went into the code (which I hadn't written) and just changed it to obj->Convert(num,buf,11). Except, as I found out, that last parameter _wasn't_ the size of the buffer to keep it from overflowing. It was the radix of the conversion, and I was getting base-11 numbers, leading to an odd bug. If the original code had been [obj convertNumber:num toString:buf usingRadix:10], that would never have happened. And in C++, no one would write obj->convertNumberToStringUsingRadix(num,buf,11), that's just illegible. Which brings up another point -- C++ syntax discourages descriptive methods names in favor of briefer ones. This can have an actual impact on the API design and program! Not only is it less descriptive, but it encourages the API developer to use _multiple_ methods each with short names taking a few parameters -- purely so that the cumulative effect of having multiple short method names becomes sufficiently descriptive and you don't have methods with lots of undescribed parameters -- where Obj-C would use _one_ method that takes lots of parameters. You actually see this occurring in the Java APIs a lot. Which makes you write more lines of code and have the overhead of more method calls, and makes things less atomic. Once you start using Obj-C syntax a lot, the readability/maintainability advantages become such that I don't want to switch (to, say, Java) largely because of the syntax! But anyway, if you don't like Obj-C syntax in Rhapsody, just use Java. They've got new bindings for the method names and C++ syntax.
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 13 May 1998 18:46:01 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6jd7r9$56b$1@crib.bevc.blacksburg.va.us> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <3559EA34.B89B62D7@alumni.caltech.edu> <3559F204.3352@stanford.edu> <Pine.NXT.3.96.980513175733.12677B-100000@pathos> In article <Pine.NXT.3.96.980513175733.12677B-100000@pathos>, Bill Bumgarner <bbum@codefab.com> wrote: > (3) Just because malloc() is effecient about dealing with memory, does not > justify one being a lazy programmer! Understanding how and when memory is > allocated, what allocations are for long-term and which are temporary, and > optimizing accordingly can greatly improve performance. Indeed. I've been wanting to learn more about the effective use of virtual memory. Does anyone have any tips, or can point me to something to look at? I hear that understanding your app's page access profile and rearranging how thins are stored can lead to significant optimization opportunities in some cases. Setting things up so that paging is minimized, how long things stay around and how costly it is to allocate more memory, things like that.. I do have a general understanding of VM, but I'm sure there are some obvious tricks extending from that sitting under my nose which just haven't occurred to me. > [Threads] typically buy you little in the way of performance and there are > generally elegant ways to avoid threading that can greatly improve the > flexibility and maintenance costs of a particular solution. How would you handle updating a couple of live views and maybe a progress bar, all in the same window? > [BTW: I have been programming NEXTSTEP/NeXTSTEP/NeXTstep/OpenStep/Rhapsody > since 1989... with a number of years of mac programming experience > intermixed. You mac 'gammers are gonna LOVE this stuff... imagine an OS > that DOESN'T crash. Not ever. CodeFab's primary server runs OpenStep 4.2-- > it had an uptime of 88 days until we took it down yesterday to add drive > space. I have friends that have had machines with up times of UP TO A > YEAR!] My OpenStep 4.2 box locked up tight as a drum just today, screen froze and couldn't even telnet in.. had to hard reboot. :) (But he's right, OpenStep really does not crash all that much.)
Date: Wed, 13 May 1998 16:02:24 -0700 From: stevehix@safemail.com (Steve Hix) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <stevehix-1305981602240001@ip53.safemail.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <stevenj-ya02408000R1105982106140001@news.mit.edu> <6j899s$g46$2@newsfep4.sprintmail.com> <MPG.fc17e6f707c72619896ac@news.supernews.com> In article <MPG.fc17e6f707c72619896ac@news.supernews.com>, don.brown@cesoft.com (Donald Brown) wrote: > The new Carbon system means that you can use almost the same old API and > get all the advantages of the new OS, and their estimate was 5 days to do > this to a large app. > > I'm unclear about whether MacOS X will run openstep programs. Would it be sufficient for MacOS X to support the full YB API for this to happen?
Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer From: spambait@seemysig.com (Mike Cohen) Subject: Re: rhapsody for intel and "carbon" Message-ID: <B17FB1DD9668AC557@0.0.0.0> Organization: ISIS International References: <6jbldv$eua$4@news.idiom.com> <6jbpfl$rji$1@pump1.york.ac.uk> Date: Thu, 14 May 1998 00:24:00 GMT NNTP-Posting-Date: Wed, 13 May 1998 20:24:00 EST In article <6jbpfl$rji$1@pump1.york.ac.uk>, pete@ohm.york.ac.uk (-bat.) wrote: | > Yeah. Intel processors suck. They keep the bytes in a word in the wrong | | Yes they suck - but not because of byte ordering, which is fairly irrelevent to | decently written code. In fact there are a lot of good reasons for making | processors little-endian. Intel's processors are incredibly ugly & baroque. All of them carry around lots of excess baggage for backward compatibility. Even the Merde^h^hced is just as ugly. They wouldn't know an elegant design if it jumped up and bit them. -- Mike Cohen - mike_cohen (at) pobox (dot) com - http://pobox.com/~macguru Sound is the same for all the world - Youssou N'dour, "Eyes Open"
From: macghod@concentric.net Newsgroups: comp.sys.next.programmer Subject: newby questions Date: 14 May 1998 01:17:09 GMT Message-ID: <6jdgml$jdq$2@newsfep4.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I am having problems with the travel advisor tutorial in the openstep developer tutorial. I am at #3, resuse the converter class (p69) it says to select the converter class from the currency converter nib and then copy, then to select NSObject in travelAdvisor.nib, then select paste. I do so and I get Unexpected Error Error: *** -[IBClassDescriptor initWithDictionary:]; selector not recognized Also, can developer be installed onto windows 95? Can openstep apps be compiled so they run on windows 95? How do I get a file to run on win nt? -- Running on Openstep 4.2, the best os in the world, and about to get better! Mac/openstep qa work desired, email me for resume NeXTMail and MIME OK!
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: Re: NSConnection-getRootProxy only 256 calls possible? Date: Wed, 13 May 1998 10:04:45 +0200 Organization: Square B.V. Message-ID: <3559541D.D17F5B1E@Square.nl> References: <3556C07E.296300A0@Square.nl> <6j80mf$fqe$1@news.apple.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Mark Bessey <mark_bessey@apple.com> I guess it's a combination of those two, but mainly the second point. I stopped passing NSHost objects (by reference?) and added the host address to a dictionary. The dictionary gets copied totally (it contains only NSStrings) and does not cause any problems. Maurice. > There are two problems you're likely encountering here: > (1) The BSD sockets code imposes a limit of 256 open sockets at a time. > (2) Some versions of D.O. don't ever free up their sockets. > > Due to #1, you won't be able to support more than 256 connections at > once. As for #2, I believe all the known resource leaks in D.O. have > been fixed in Rhapsody DR2, which should be available real soon now... > > -Mark -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto:mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+ Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" and "to trick into believing or accepting as genuine something false and often preposterous"
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: 13 May 1998 08:25:35 GMT Organization: Idiom Communications Message-ID: <6jbldv$eua$4@news.idiom.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <6j7sdr$4gn$1@newsfep4.sprintmail.com> <6j7tij$vb7$1@crib.bevc.blacksburg.va.us> <stevenj-ya02408000R1105982039000001@news.mit.edu> <andyba-ya02408000R1205980916210001@news> <355873F0.6B86BC5C@milestonerdl.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: mark@milestonerdl.com M Rassbach may or may not have said: -> -> -> Andy Bates wrote: -> -> > No, they already said that carbon-based apps would NOT run on Rhapsody for -> > Intel. Specifically, the APIs are still PowerPC-only, and hence not -> > runnable on Intel. -> -> Is there a TECHNICAL reason, or just political? Yeah. Intel processors suck. They keep the bytes in a word in the wrong order, and NeXT really had to pull teeth to make NeXTSTEP cope with that. The carbon code is probably riddled with byte-order dependencies, which aren't actually a problem when you're running on a processor that doesn't suck. -jcr
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 13 May 1998 01:37:00 -0700 Organization: Primenet Services for the Internet Message-ID: <B17EA9AE-87744@206.165.43.165> References: <SCOTT.98May12232855@slave.doubleu.com> nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Scott Hess <scott@doubleu.com> said: >Now surprise me - instead of jumping off on an "I told you so" >rampage, why not present a reasonable prediction of various ways that >you could more or less seamlessly use QuickDraw as the drawing layer >for AppKit - which is to say using Apple's design patterns in >Objective-C, as opposed to C-based APIs. And remember, they have the >source, so they can rearrange the code and how it relates to itself >arbitrarily, I'm hoping that they meant "use the QuickDraw GX algorithms" (as opposed to GX itself). AFAIK, the 32-bit QuickDraw algorithms, while fast, aren't really designed for that kind of thing (resolution-independent graphics) while GX algorithms were (that's why they invented them). Also, there's no provision for any color-space besides RGB in 32-bit QUickDraw, while GX handles RGB, RGBA, etc with a lot of overlap with DPS's color spaces (although DPS uses an alpha channel for all color spaces while GX only uses an alpha channel for RGBA, ARGB and GrayA). If they just grabbed the various GX algorithms and re-implemented them as the rasterizer for GhostScript, it would be relatively doable, I'd think, but I can't imagine using 32-bit QuickDraw algorithms for that purpose. They're pretty much designed for 72 DPI imaging, although you're supposed to be able to specify a different device resolution at a low level. On the other hand, perhaps the fact that the GX algorithms are designed around a fixed-point coordinate system makes them less attractive than using the [apparently] integer-coordinate 32-bit QD algorithms? If you'll recall, I predicted that it would take at least 2-3 years to enhance GX sufficiently to take over for DPS in Rhapsody and that I thought that the best bet would be to abstract all DPS dependencies in the Yellow Box so that GX could take over once it had been made robust enough to do so? I have ALWAYS maintained that it was insane for APPLE to depend on a 3rd party for something as fundamental as its graphics engine, even if it made sense for NeXT to do so. [yes, I couldn't resist an "I told you so" -although maybe not the one that you expected] I just don't see how Apple can do it in any reasonable time-frame as long as there is the need to have DPS used in the framework and I don't see GX as being ready to do so (and I don't see 32-bit QuickDraw as being CAPABLE of doing so) before the year 2000 unless they can simply grab the appropriate primitives & convert them to floating point with minimal bother. Besides, doesn't Apple own the DPS algorithms that they've used to implement DPS on NeXT? How can they have a new Carbon-based graphics engine [as hinted at by Mike Paquette] if they're not simply using the graphics primitives of DPS, but called directly rather than via DPS? You don't need DPS to be able to draw resolution graphics (as GX demonstrates) but you do need drawing routines owned by Apple, and I believe that Apple DOES own the source to the primitives that Rhaposdy DPS uses, right Mike? Why not take a discussion of what should be in the graphics engine of Carbon/Yellow Box into another thread or even to GX-talk? I think that we need a superset of GX and the Rhaposdy graphics objects that can be called using structured code ala GX in Carbon but that can be used and sub-classed in the standard YB way under the Yellow Box. That would be the best of all possible worlds, no? It would allow some ability to handle GX graphics in Carbon and YB graphics, while still retaining the full OOPness in YB. [actually, I'm pretty sure that a structured, non-extensible version of the various Rhaposdy graphics classes is what will be announced on Thursday, unless they're providing a complete OOP library in Carbon, which doesn't fit with the rest of the design of the MacOS APIs -I just want backwards compatibility with GX shapes AND a more GX-like feel to the whole thing -I'm betting that we only get an immediate-mode API in Carbon] ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: Mark Bessey <"mark_bessey"@apple.com (no spam, please)> Newsgroups: comp.sys.next.programmer Subject: Re: Strange event Date: Wed, 13 May 1998 02:40:00 -0700 Organization: Apple Computer, Inc. Message-ID: <6jbpo8$7qa$1@news.apple.com> References: <6jamu4$7n9$1@Masala.CC.UH.EDU> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Some suggestions: 0. Have someone else look at the code in question. Better yet, try to explain how it is supposed to work to them. You'd be surprised how often that works. 1. You aren't REALLY using NXTableView, are you? You meant NSTableView, right? If not, you really ought to consider upgrading to OPENSTEP... 2. Run the application under gdb, and set a breakpoint on [NSException raise:] That should let you see if your code is actually sending the string message somewhere, or if it's happening in the AppKit somewhere. 3. If that doesn't work, you may be keeping a reference to a released object. Use gdb's set env command to enable "zombies" for your application. set env NSZombieEnabled YES This ensures that any object you allocate is given a unique address, which makes it easier to track down which variable is being mistakenly released. You can read all about NSZombieEnabled in the "Debugging" chapter of the online "DevEnvGuide" documentation. I hope one or more of these suggestions helps. -Mark Michael Mezzino wrote: > > Can anyone tell me how to find the cause of the following runtime error: > > May 12 18:18:40 [1670] Uncaught exception: NSInvalidArgumentException. > return code = 100003. *** NSInlineCString doesNotRecognizeSelector: > -string > > I do not have any string methods in my app, so I assume that NSString is > somehow causing this. It occurs when I double click in an NXTableView. > Also, if I monitor the event flow through the sendEvent:(NXEvent *)event > method, the error occurs between the single click and the double click. I > have tried to use gdb setting breakpoints around the critical section of > code. The error does not occur in single step mode but does occur in run > mode. Has anyone ever experienced anything like this? Thanks so much. > > Mike Mezzino > mezzino@gauss.cl.uh.edu
From: pete@ohm.york.ac.uk (-bat.) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: 13 May 1998 09:34:45 GMT Organization: The University of York, UK Sender: pcf1@york.ac.uk Message-ID: <6jbpfl$rji$1@pump1.york.ac.uk> References: <6jbldv$eua$4@news.idiom.com> jcr.remove@this.phrase.idiom.com (John C. Randolph) writes: > Yeah. Intel processors suck. They keep the bytes in a word in the wrong Yes they suck - but not because of byte ordering, which is fairly irrelevent to decently written code. In fact there are a lot of good reasons for making processors little-endian. > order, and NeXT really had to pull teeth to make NeXTSTEP cope with that. They did ? That surprises me greatly - the underlying Mach kernel was portable, and high level code shouldn't contain many byte order dependencies. I suspect that what caused far more of a problem was making the OS work with a rather horrid PC hardware architechture. > The carbon code is probably riddled with byte-order dependencies, which > aren't actually a problem when you're running on a processor that doesn't > suck. Uh, no - what you mean is "byte order dependencies aren't a problem as long as you keep running the code on the processor for which they were written". I don't see how you manage to riddle code with byte order dependencies without coding in a very careless manner. In the NeXT world we've been very used to writing Apps which run on both m68k and i386 architectures for some time. 99.9% of the time you just compile it on the different architecture and away it goes quite happily. We have an app here of about 70 thousand lines which uses raw 16 bit data from the disc in fix endian format. In all of that source there are only 37 lines of code to deal with endian dependencies. And thats only that long because I was feeling lazy when I wrote those functions. Yeah, I prefer big endian machines as well, but only because it's easier to read the hex when you dump out the bytes. When actually working with one of those processors - especially when trying to wire the damn thing up - it's marginally easier to work with little endian. So my opinion as to why I prefer depends on whether I'm programming or soldering. Out of interest what is your opinion on processors which can change their endianness to either way - are they crap in one mode and good in the other mode ? does it really make that much difference to programming the things ? If you really want to see a confusing method of addressing memory go take a look at transputers sometime with a signed address space - and it's relationship to OCCAM. Which is where a lot of our NeXTStep code originated. ..and if you want to see an example of where people who stick to big endian format religious is a right pain in the arse then try programming with sockets and remembering when you need to call htons() and htonl(). yuk. -bat.
From: don.brown@cesoft.com (Donald Brown) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 13 May 1998 08:41:02 -0500 Organization: Prairie Group Message-ID: <MPG.fc35eecf63b73e69896bb@news.supernews.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <1d8waq9.1t0soxr6xzkjtN@rhrz-isdn3-p44.rhrz.uni-bonn.de> <B17E5CDC96688D440@0.0.0.0> <3559662E.E37ABBE0@alumni.caltech.edu> In article <3559662E.E37ABBE0@alumni.caltech.edu>, woody@alumni.caltech.edu says... > Mike Cohen wrote: > > Only if you pass the ACTUAL data in the AppleEvent, rather than doing > > something like passing pointers to data in an AppleEvent (I've seen some > > applications which do that - it's really gross). On a similar note, they > > also say you shouldn't use Gestalt to pass pointers between applications. > > Well, hell's bells; Apple has been telling us not to do this sort of > grossness since Day One. > > If the reason why programmers have to modify 10% of thier existing > applications is because of stuff like *this*, I dunno if there's > anything to say to those programmers other than "I told you so," and > "Better off doing it right the first time than having to redo it > later." Absolutely! Many things that you can't do under Carbon, were things you were told you shouldn't do under MacOS. But, until Carbon, you could get away with them (and sometimes have some pretty impressive advantages from them). The only trick I'm worried that won't be replaceable are the trick's I've played with jGNEFilter...I use it for having hot keys (and customers do like hot keys, at least as an option), I also use it for faceless apps to pop up dialogs in other applications, and to have a non-modal window floating over all apps using TSM. My customers have come to expect these features, so I'm hoping there will be some API that lets me do them. Donald
From: Mark Frank <mark.s.frank@tek.com> Newsgroups: comp.sys.next.programmer Subject: OS version of DPSTimedEntry? Date: Wed, 13 May 1998 09:28:16 -0700 Organization: Tektronix Message-ID: <3559CA20.D82@tek.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Is there an open step version of DPSTimedEntry? I couldn't find one in the online documentation. If not, what I want to do is start a task running when a button is pressed, and then be able to respond to a stop button. I tried using NSThread, but this didn't work, since the AppKit can not be used by more than one thread at a time. (I read this in the documentation after already trying it). Any help would be appreciated. Please respond by email since the news server here at Tek seems to throw away about half the posts! (at least from the Next news groups). - Thanks, Mark mark.s.frank@tek.com
From: jim@ergotech.com Newsgroups: comp.sys.next.programmer Subject: Re: NSConnection-getRootProxy only 256 calls possible? Date: Wed, 13 May 1998 16:24:49 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6jchgh$g1g$1@nnrp1.dejanews.com> References: <3556C07E.296300A0@Square.nl> <6j80mf$fqe$1@news.apple.com> Since there was mention of customers here, any fix in DR2 is probably useless. If the normal license applies it will not be possible for OpenStep developers to ship product with DR2. Since Rhaposody is not planned for shipment until the fall of this year that will be something more than a year with developers shipping on product with known bugs. Does anyone know if there are any forcoming solutions to this problem? Jim In article <6j80mf$fqe$1@news.apple.com>, Mark Bessey <"mark_bessey"@apple.com (no spam, please)> wrote: > > There are two problems you're likely encountering here: > (1) The BSD sockets code imposes a limit of 256 open sockets at a time. > (2) Some versions of D.O. don't ever free up their sockets. > > Due to #1, you won't be able to support more than 256 connections at > once. As for #2, I believe all the known resource leaks in D.O. have > been fixed in Rhapsody DR2, which should be available real soon now... > > -Mark > > Maurice le Rutte wrote: > > > > I can't sell my customer that he should quit the server > > after 256 client connections!! > > -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: Pascal Bourguignon <pjb@imaginet.fr> Newsgroups: comp.sys.next.programmer Subject: Re: C++ and rhapsody Date: 14 May 1998 03:31:33 GMT Organization: Deficiente Message-ID: <6jdoil$76r$2@news.imaginet.fr> References: <6jcvpm$25r$1@newsfep1.sprintmail.com> <6jd7db$555$1@crib.bevc.blacksburg.va.us> nurban@crib.bevc.blacksburg.va.us (Nathan Urban) wrote: >[...] >Well, it turned out at some point that an 11-digit number was necessary >after all in some extreme cases, so I simply went into the code (which I >hadn't written) and just changed it to obj->Convert(num,buf,11). Except, >as I found out, that last parameter _wasn't_ the size of the buffer to >keep it from overflowing. It was the radix of the conversion, and I >was getting base-11 numbers, leading to an odd bug. If the original >code had been [obj convertNumber:num toString:buf usingRadix:10], >that would never have happened. And in C++, no one would write >obj->convertNumberToStringUsingRadix(num,buf,11), that's just illegible. Well, being the fool on the hill that I am, I do, at least since I leaned Objective-C. But I always liked long descriptive names. So far, my longest one is 97 characters. -- __Pascal Bourguignon__ | The box said 'Requires Windows 95, mailto:pjb@imaginet.fr | or better.' So I bought a Macintosh.
From: see-below@not-my-address.com (Matthew Vaughan) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 13 May 1998 21:38:06 -0700 Message-ID: <see-below-1305982138070001@209.24.241.190> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> In article <6jc50h$eh@fridge.shore.net>, "Mike Lockwood" <lockwood@metrowerks.com> wrote: > I don't see how the Mac API could be made thread safe. Suppose you have two > threads, thread A, and thread B. Thread A calls SetPort(), and then starts > drawing. At about the same time, thread B calls SetPort to a different > port, and then starts drawing. The problem is a global current port, so > both threads will end up drawing into the same port. With the Win32 API, on > the other hand, all drawing is done via a device context. Two threads can > draw simultaneously into different device contexts without interfering with > each other. > > (I'm not trying to start a debate over the advantages or disadvantages of > the MacOS vs. Win32 APIs. I'm just pointing out that the API would need to > change in order to have thread safe graphics). > > There are other more serious examples of non-threadsafeness in the Mac APIs. > Think about all the places in the API that use a handle. If one thread > tries to allocate memory and causes a heap compaction while another thread > is doing something with a handle in the same heap, you are in big trouble. > I guess you could fix that without changing the API by making all handles > permenantly locked, but of course that defeats the purpose of having handles > in the first place. I'm not sure if they're trying to make Carbon thread-safe within an application (though perhaps they are). I think their main concern is between applications. The handle problem shouldn't apply, since they'll each have their own virtual memory space. .................................................... MATTHEW VAUGHAN matthewv at best dot com (damn spammers...) http://www.best.com/~matthewv/ ....................................................
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 13 May 1998 10:48:00 -0700 Organization: Primenet Services for the Internet Message-ID: <B17F2B02-52ACD@206.165.43.172> References: <6jc5br$so@fridge.shore.net> nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Mike Lockwood <lockwood@metrowerks.com> said: >>Unless, of course, Mike and company have already done all of that and >>announce it on Thursday... > >Nope, not me. I'm not announcing anything on Thursday. And I haven't >written a HyperCard XCMD since 1988... > That's Mike Paquette, and the reference was to whatever graphics will be made available in YB and Carbon. ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: OS version of DPSTimedEntry? Date: 14 May 1998 01:04:25 GMT Organization: Idiom Communications Message-ID: <6jdfup$jpv$1@news.idiom.com> References: <3559CA20.D82@tek.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: mark.s.frank@tek.com Mark Frank may or may not have said: -> Is there an open step version of DPSTimedEntry? I couldn't find -> one in the online documentation. If not, what I want to do is -> start a task running when a button is pressed, and then be able -> to respond to a stop button. I tried using NSThread, but this -> didn't work, since the AppKit can not be used by more than one -> thread at a time. (I read this in the documentation after already -> trying it). -> Any help would be appreciated. Please respond by email since -> the news server here at Tek seems to throw away about half the -> posts! (at least from the Next news groups). -> - Thanks, Mark -> mark.s.frank@tek.com What you want is NSTimer. -jcr
From: Sterling <Sterling_Augustine@stanford.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 13 May 1998 12:18:40 -0700 Organization: Stanford CS Message-ID: <3559F204.3352@stanford.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559EA34.B89B62D7@alumni.caltech.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit William Edward Woody wrote: > > Of course this process would be helped if Apple added a 'QDFlush()' > call, but if you use a timer which expired every 1/60th of a second, > you wouldn't need that flush routine. > > This also doesn't fix the problem of what if Thread A deletes a port > being drawn to by Thread B; however, it does solve the bigger problem > of multi-threaded access. (In the case of one thread deleting another > thread's drawing port, well--Don't Do It!) > When two threads share memory, within a process, it is the programmer's responsibility to ensure that one doesn't destroy memory that the other is using. Carbon will separate apps into their own address space, so one process won't be able to trash memory that the other is using. > In the case of memory, I would say that a thread-safe memory manager > should serialize access to the heap. It sounds like Carbon uses a global > heap (no more separate Application heaps), so access would have to be > serialized anyways. > No, it doesn't use a global heap. Every process will have it's own memory protected heap. That makes all the threading issues moot. > That means that while the heap is compacting, another thread who asks > for a new handle would just have to wait. See my previous post on why there should never be a need to compact the heap in a modern OS. I have a hard time believing this fix won't be part of Carbon. Sterling
From: Pascal Bourguignon <pjb@imaginet.fr> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 14 May 1998 03:36:48 GMT Organization: Deficiente Message-ID: <6jdosg$76r$3@news.imaginet.fr> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <6j84dp$vln$1@crib.bevc.blacksburg.va.us> <MPG.fc15f0828d669629896aa@news.supernews.com> <SCOTT.98May12225116@slave.doubleu.com> scott@doubleu.com (Scott Hess) wrote: >[...] >What, pray tell, do you think MacOS X is? > >[To be fair, I'll tell you what _I_ think MacOS X is, coming from a >NeXTSTEP background. I think MacOS X is Rhapsody with the "BlueBox >without the Box" highly integrated. They've spun it out as "MacOS >with Rhapsody technologies", but IMHO what we're hearing in the >YellowBox and Rhapsody sessions boils down to "Rhapsody with MacOS >technologies." From what is being said right now, if they ever ship >MacOS X and Rhapsody in parallel, the differences will be entirely in >the packaging,] >-- >scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ ><Favorite unused computer book title: The Compleat Demystified Idiots > Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed> They've done that in A/UX already. You have in it an API to the Macintosh toolbox, directly from the UNIX part of it. (It integrated also a BlueBox to let you run MacOS applications). -- __Pascal Bourguignon__ | The box said 'Requires Windows 95, mailto:pjb@imaginet.fr | or better.' So I bought a Macintosh.
From: Sterling <Sterling_Augustine@stanford.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 13 May 1998 21:36:52 -0700 Organization: Stanford CS Message-ID: <355A74AC.107@stanford.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: William Edward Woody <woody@alumni.caltech.edu> William Edward Woody wrote: > [stuff about the benefits of handles deleted] So what your saying is that the following OS's memory management systems are terribly flawed: -UNIX -VMS -OS/2 -Windows 95/NT -Rhapsody/NextStep Sorry, I just don't buy it, but perhaps you could convince me. Name one operating system generally considered modern and equipped with virtual memory that uses handles as its primary method of memory allocation. >Say I have two pointers, A and B. And their addresses are: > > A = 0x00010000 > B = 0x00010100 > >Q: How many bytes are between A and B? > >A: There are 256 bytes between A and B. If both these pointers are allocated via malloc, then you shouldn't be relying on their relative locations anyway. If you need to grow one beyond the available bounds, then copying the data is really the only good way. Handles do exactly that as well. But with the pointer method, you only lose performance only when you absolutely have to. With handles you live with dual indirection on every memory access and memory safe calls. And if it's absolutely vital for you to keep your own 4 gigabyte address space unfragmented, then you can implement them yourself quite trivially. Yes, keeping a large working set will make your performance fall (in particular when it exceeds the amount of physical memory) but intelligent paging algorithms like those found in Mach do a fair amount to minimize that. Plus it's not too hard to localize your memory access. Talking about turning VM on and off shows a very home-computer orientation. Industrial strength operating systems don't allow it. >The reason why you don't have to worry about fragmentation in most >systems with virtual addressing is because the address space is so large, >for the most part when the currently allocated heap gets too fragmented, >you can just allocate more virtual memory. But this doesn't allow you >to magically defragment the heap. It would be a pathological program indeed where compaction would be enough better than allocating additional virtual pages to make it worth it. You are talking about fragmenting every page in your working set such that addtional allocations won't fit inside the free blocks. Can you name one that wasn't designed to do it that does? As another poster pointed out, memory allocators are extremely complicated and go to great lengths to prevent the kind of fragmentation handles are designed to avoid. They manage to do it quite well in other operating systems. The MacOS needs to catch up. Name one > And apparently someone wasn't paying attention during class. Sorry, got an A-. No mention of handles, even in the section on memory management. At any rate, I would prefer to keep this to the level of civil discussion. How about you? Sterling
From: William Edward Woody <woody@alumni.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 13 May 1998 23:17:47 -0700 Organization: In Phase Consulting Message-ID: <355A8C8B.3C7D9DAB@alumni.caltech.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <355A74AC.107@stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit First, let me apologize for the tone of my last note. It was a bad day, and after reading the same thing over and over again, I had just had it. Sterling wrote: > William Edward Woody wrote: > > > [stuff about the benefits of handles deleted] > > So what your saying is that the following OS's memory management systems > are terribly flawed: > > -UNIX > -VMS > -OS/2 > -Windows 95/NT > -Rhapsody/NextStep > > Sorry, I just don't buy it, but perhaps you could convince me. Name one > operating system generally considered modern and equipped with virtual > memory that uses handles as its primary method of memory allocation. Not terribly flawed. Just missing a powerful tool for heap management. Oh, and GlobalAlloc() still exists on Windows 95/NT, though for backwards compatability. (And much of the original functionality of Window's memory handles has been dumped.) > >Say I have two pointers, A and B. And their addresses are: > > > > A = 0x00010000 > > B = 0x00010100 > > > >Q: How many bytes are between A and B? > > > >A: There are 256 bytes between A and B. > > If both these pointers are allocated via malloc, then you shouldn't be > relying on their relative locations anyway. If you need to grow one > beyond the available bounds, then copying the data is really the only > good way. Handles do exactly that as well. ... That wasn't my point. My point was what you just noted: that when you have to grow a chunk of allocated memory, something's location has to move. And that was the thing that griped me: in your original post, you suggested that VM solves this problem by being able to change the physical address of a pointer. But as heap fragmentation happens in virtual address space, no amount of remapping of physical addresses will change things. > ... But with the pointer method, > you only lose performance only when you absolutely have to. With handles > you live with dual indirection on every memory access and memory safe > calls. ... You don't have to have dual indirection on every memory access; it's possible to lock handles temporarly when you are doing stuff to them. And in the Win16/Win32 API, you *had* to lock a handle in order to access stuff in it; there is no (legal) mechanism of dereferencing a memory handle. > Talking about turning VM on and off shows a very home-computer > orientation. Industrial strength operating systems don't allow it. No; talking about turning on and off VM helps differentiate the difference between having VM and not having VM. Unfortunately, I've encountered so many comments about the benefits of VM which seem to show a lack of understanding of what VM really provides, it's easier to explain when you describe it as turned off. <sarcasm> Don't worry; I've played with an "industrial strength OS" once or twice in my distant pass, so I didn't miss out in my miserable youth. </sarcasm> > >The reason why you don't have to worry about fragmentation in most > >systems with virtual addressing is because the address space is so large, > >for the most part when the currently allocated heap gets too fragmented, > >you can just allocate more virtual memory. But this doesn't allow you > >to magically defragment the heap. > > It would be a pathological program indeed where compaction would be > enough better than allocating additional virtual pages to make it worth > it. You are talking about fragmenting every page in your working set > such that addtional allocations won't fit inside the free blocks. Can > you name one that wasn't designed to do it that does? First, I've encountered a few daemons on the Sun workstation which had such pathological problems. (Long story short: as a sysman of a couple of Unix boxes, I had a couple which would lose performance and eventually croak. It was because one of the daemons on the system (I forget which now) would grow it's working set to the point where the system would bog down page thrashing. A patch from Sun fixed it.) I've also encountered fragmentation problems with graphics rendering systems written for VMS; the original writer would malloc() and free() individual objects as they were processed during a rendering session; after a few hundred frames the process would thrash like crazy, due to fragmentation caused by mallocing and freeing such small objects. Yes, I've seen fragmentation inside of every page in a working set. > As another poster pointed out, memory allocators are extremely > complicated and go to great lengths to prevent the kind of fragmentation > handles are designed to avoid. They manage to do it quite well in other > operating systems. The MacOS needs to catch up. Name one They manage to do it "quite well" only because they are able to grow the working set when all else fails. And far from suggesting the Macintosh OS needs to "catch up", it's worth acknowledging that the MacOS's memory manager is designed to operate quite well inside a small heap. And it's also important to recognize that those same tools would also help developers keep the working set small in a world of virtual addressing. If you don't want to use handles, don't use 'em. But don't throw away the tool just because you don't understanding it. -- William Edward Woody | In Phase Consulting woody@alumni.caltech.edu | Macintosh & MS Windows Development http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/
From: William Edward Woody <woody@alumni.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 13 May 1998 23:35:34 -0700 Organization: In Phase Consulting Message-ID: <355A90B6.6386212@alumni.caltech.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559EA34.B89B62D7@alumni.caltech.edu> <3559F204.3352@stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sterling wrote: > William Edward Woody wrote: > > Of course this process would be helped if Apple added a 'QDFlush()' > > call, but if you use a timer which expired every 1/60th of a second, > > you wouldn't need that flush routine. > > > > This also doesn't fix the problem of what if Thread A deletes a port > > being drawn to by Thread B; however, it does solve the bigger problem > > of multi-threaded access. (In the case of one thread deleting another > > thread's drawing port, well--Don't Do It!) > > > When two threads share memory, within a process, it is the programmer's > responsibility to ensure that one doesn't destroy memory that the other > is using. Carbon will separate apps into their own address space, so one > process won't be able to trash memory that the other is using. When you run a new PalmPilot applet, the previously active applet actually exits. This means that while all of the applets may be in memory, only one applet is running at any one time, even though it seems like they are all running simultaneously. Of course I have no idea what my comment, or your comment above, has to do with my original post about making QuickDraw thread-safe. > > In the case of memory, I would say that a thread-safe memory manager > > should serialize access to the heap. It sounds like Carbon uses a global > > heap (no more separate Application heaps), so access would have to be > > serialized anyways. > > > > No, it doesn't use a global heap. Every process will have it's own > memory protected heap. That makes all the threading issues moot. Sorry; was talking physical memory, not virtual memory. Me fault for not being clear. > > That means that while the heap is compacting, another thread who asks > > for a new handle would just have to wait. > > See my previous post on why there should never be a need to compact the > heap in a modern OS. I have a hard time believing this fix won't be part > of Carbon. Fix? Well, according to the Carbon docs thus far released, the Memory Manager remains unchanged. Remember: the goal of Carbon is not just to create a new modern OS. It's to create a new modern OS which is source-level compatable with most existing Macintosh applications. And whether you like it or not, Handles is a very important part of that API. - Bill (Who again notes that just because you don't like a tool doesn't mean it should go the way of the Dodo.) -- William Edward Woody | In Phase Consulting woody@alumni.caltech.edu | Macintosh & MS Windows Development http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/
From: William Edward Woody <woody@alumni.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 13 May 1998 23:29:31 -0700 Organization: In Phase Consulting Message-ID: <355A8F4B.17463CF@alumni.caltech.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559EA34.B89B62D7@alumni.caltech.edu> <3559F204.3352@stanford.edu> <Pine.NXT.3.96.980513175733.12677B-100000@pathos> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Bill Bumgarner wrote: > A few FACTS about programming under OSX/Rhapsody/<<insert modern OS here>> > and their memory models: ... Not to rain on anyone's parade, but it's clear from the Carbon docs on the Apple developer web site that the Memory Manager will be mostly unaffected, aside from the fact that every app has it's own address space. Of course the situation may be like that with Windows 95/Windows NT; routines like 'GlobalAlloc()' still exist, though their functionality is seriously curtailed. (For those who don't know, GlobalAlloc() et al were the Win16 mechanism for allocating memory handles.) > (2) Handles go away. The "memory manager"-- comprised of a whopping two > calls, malloc() and free() [or various cousins; their are other APIs for > effeciently managing large hunks of memory]-- automatically "compacts" > memory as it is allocated. > > Actually, what REALLY happens is that everytime malloc() is asked to > allocate some memory, it automatically tries to grab a piece that is in > between other pieces of allocated memory. If it can't find a hunk large > enough, it allocates a new hunk at the "end" of the apps current address > space. If needed, the kernel will create more virtual pages of memory for > the app. Ummmm, on every Unix flavor I played with, the only (and I mean _only_) memory manager routine provided is some routine or another to grow the page set. The rest of the memory management functionality was done in a library linked with your application; thus, malloc() and free() aren't actually parts of the operating system. This has the advantage of allowing completely different memory management systems to be tried out, as you aren't locked by the operating system to a 'alloc on demand/release when requested' mechanism such as provided by malloc() and free(). > Threads suck. *shrug* I kinda like 'em. - Bill -- William Edward Woody | In Phase Consulting woody@alumni.caltech.edu | Macintosh & MS Windows Development http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/
From: Bill Bumgarner <bbum@codefab.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 13 May 1998 18:14:50 -0400 Organization: Spacelab.net Internet Access Message-ID: <Pine.NXT.3.96.980513175733.12677B-100000@pathos> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559EA34.B89B62D7@alumni.caltech.edu> <3559F204.3352@stanford.edu> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII In-Reply-To: <3559F204.3352@stanford.edu> A few FACTS about programming under OSX/Rhapsody/<<insert modern OS here>> and their memory models: (1) Every application has a virtual memory space that is distinct from all others. For all intents and purposes, it is as if every app has it's own distinct hunk of memory-- addressing may or may not be the same as any other app's space. (2) Handles go away. The "memory manager"-- comprised of a whopping two calls, malloc() and free() [or various cousins; their are other APIs for effeciently managing large hunks of memory]-- automatically "compacts" memory as it is allocated. Actually, what REALLY happens is that everytime malloc() is asked to allocate some memory, it automatically tries to grab a piece that is in between other pieces of allocated memory. If it can't find a hunk large enough, it allocates a new hunk at the "end" of the apps current address space. If needed, the kernel will create more virtual pages of memory for the app. This is a grossly simplified view of the algorithms involved. For an example, read the source for the gnu malloc package. It is interesting. See #1-- every app has a distinct memory space... that means that hunks of memory returned by malloc WILL NOT BE interleaved with other applications [called "tasks"]. (3) Just because malloc() is effecient about dealing with memory, does not justify one being a lazy programmer! Understanding how and when memory is allocated, what allocations are for long-term and which are temporary, and optimizing accordingly can greatly improve performance. (4) Multithreading is just as hard... the threads of any given task all run in the same memory space. So, all the wonderful synchronization issues normally associated with threads are fully present. However, because (a) new tasks are easy to create/launch (b) interprocess communication under YB OR Mach is relatively trivial and (c) mach can handle many, many processes (my rhapsody machine is currently running 100 simultaneous processes-- that's under non-heavy usage) effeciently, you are FAR better off avoiding threaded designs. They typically buy you little in the way of performance and there are generally elegant ways to avoid threading that can greatly improve the flexibility and maintenance costs of a particular solution. Threads suck. (5) Mach's virtual memory manager is very good. It uses an excellent paging algorithm and allows for things like shared memory and other niceities with little pain. It features "copy on write" page sharing-- which means that when app A shares a page with app B with the intent that both apps will scribble on, but not share the changes, then mach doesn't actually create a COPY of any of the pages of memory until one of the apps changes the shared memory. Sharing pages of memory where both apps can write is a bit more difficult. (6) Protected memory is your friend. Not only does it mean that when an app crashes the ONLY tasks on the machine [other apps, system tasks, etc...] that are affected will be the crashing app and any tasks that it launched. I.e. no more need to reboot after an app crashed "just to be sure". Not ever. No need. But it also means that you can actually have a secure system. I.e. if a process is running as root, there is NO WAY that a user level process can scribble on the root owned process's memory. [BTW: I have been programming NEXTSTEP/NeXTSTEP/NeXTstep/OpenStep/Rhapsody since 1989... with a number of years of mac programming experience intermixed. You mac 'gammers are gonna LOVE this stuff... imagine an OS that DOESN'T crash. Not ever. CodeFab's primary server runs OpenStep 4.2-- it had an uptime of 88 days until we took it down yesterday to add drive space. I have friends that have had machines with up times of UP TO A YEAR!] enjoy, b.bum
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 14 May 1998 07:23:55 GMT Organization: Idiom Communications Message-ID: <6je66b$pcj$1@news.idiom.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <355A74AC.107@stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: Sterling_Augustine@stanford.edu Sterling may or may not have said: -> William Edward Woody wrote: -> > -> [stuff about the benefits of handles deleted] -> -> So what your saying is that the following OS's memory management systems -> are terribly flawed: -> -> -UNIX -> -VMS -> -OS/2 -> -Windows 95/NT -> -Rhapsody/NextStep -> -> Sorry, I just don't buy it, but perhaps you could convince me. Name one -> operating system generally considered modern and equipped with virtual -> memory that uses handles as its primary method of memory allocation. I don't know about OS's, but don't Postscript and Smalltalk both use double-indirection to facilitate garbage collection? -jcr
From: William Edward Woody <woody@alumni.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 13 May 1998 11:45:08 -0700 Organization: In Phase Consulting Message-ID: <3559EA34.B89B62D7@alumni.caltech.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Mike Lockwood wrote: > I don't see how the Mac API could be made thread safe. Suppose you have two > threads, thread A, and thread B. Thread A calls SetPort(), and then starts > drawing. At about the same time, thread B calls SetPort to a different > port, and then starts drawing. The problem is a global current port, so > both threads will end up drawing into the same port. With the Win32 API, on > the other hand, all drawing is done via a device context. Two threads can > draw simultaneously into different device contexts without interfering with > each other. > > (I'm not trying to start a debate over the advantages or disadvantages of > the MacOS vs. Win32 APIs. I'm just pointing out that the API would need to > change in order to have thread safe graphics). > > There are other more serious examples of non-threadsafeness in the Mac APIs. > Think about all the places in the API that use a handle. If one thread > tries to allocate memory and causes a heap compaction while another thread > is doing something with a handle in the same heap, you are in big trouble. > I guess you could fix that without changing the API by making all handles > permenantly locked, but of course that defeats the purpose of having handles > in the first place. Actually, in both of these cases, you can start to fix the problem by including 'thread globals'; globals which are local to a single thread. In the case of QuickDraw, SetPort() et. al. would affect the thread global state. Other QuickDraw calls would be stored in a thread global cache. Then, when SetPort() is called again, or when the buffer fills, or when a certain amount of time has elapsed, the buffer would be sent to a separate 'screen manager' process which would then actually execute the drawing code. (This 'screen manager' process would mediate access to the screen, and would be written to be multi-thread and multi-process safe.) Of course this process would be helped if Apple added a 'QDFlush()' call, but if you use a timer which expired every 1/60th of a second, you wouldn't need that flush routine. This also doesn't fix the problem of what if Thread A deletes a port being drawn to by Thread B; however, it does solve the bigger problem of multi-threaded access. (In the case of one thread deleting another thread's drawing port, well--Don't Do It!) In the case of memory, I would say that a thread-safe memory manager should serialize access to the heap. It sounds like Carbon uses a global heap (no more separate Application heaps), so access would have to be serialized anyways. That means that while the heap is compacting, another thread who asks for a new handle would just have to wait. Unlike the 'screen manager', the memory manager wouldn't have to be a separate process. It would just have to be written to be thread safe. (You have to serialize this access anyways--after all, it takes several steps to mark a chunk of free memory in use, and a race condition when two threads both call NewPtr() at the same time could result in a trashed heap. It's just harder to provide thread safe handles because there are more entry points into the Memory Manager that have to be properly protected with semaphores.) Most of the MacOS API doesn't have to be redesigned to make it thread safe. You just need to add the concept of 'thread globals' to store the QD global states. Make those 'thread globals' accessable to the programmer (through additions to the Thread Manager or Multiprocessor APIs), and they would allow programmers who also rely on global states a way to make their application work in a multi-threaded environment. - Bill (Who notes such thread globals are already available under the Win32 API, and are used by Microsoft to make the C library thread safe. Where do you think 'errno' goes in the MT safe MS C library?) -- William Edward Woody | In Phase Consulting woody@alumni.caltech.edu | Macintosh & MS Windows Development http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/
From: William Edward Woody <woody@alumni.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 13 May 1998 12:03:41 -0700 Organization: In Phase Consulting Message-ID: <3559EE8D.DA8F1525@alumni.caltech.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sterling wrote: > Handles should eventually go away, because a modern memory manager > should manage memory in such a way that compaction as it is understood > in the Mac universe never occurs. Modern OS's don't have to worry about > heap compaction because they solve the fragmentation problems through > virtual address spaces mapped into physical pages. It requires hardware > assistance, though. So the MacOS's reliance on them is an artifact of > the original 68000 not having a hardware MMU. I've read this now about a dozen times, and everytime I read this, I want to scream. Here's why. Say I have two pointers, A and B. And their addresses are: A = 0x00010000 B = 0x00010100 Q: How many bytes are between A and B? A: There are 256 bytes between A and B. Now if we turn on VM and use virtual address spaces mapped onto physical pages. Q: How many bytes are between A and B (in my virtual address space, which is the only address space my application can access)? A: There are 256 bytes between A and B. Q: Say I want to grow the size of the pointer pointed to by A to 512 bytes, and VM (and virtual adress mapping) is turned off. How can I grow pointer A? A: Either move the chunk of memory pointed to by A somewhere else, or move the chunk of memory pointed to by B somewhere else. Q: Now say I turn VM (and virtual address mapping) on. Can I rely on the OS to move A or B around to a different phyiscal address, and thus grow the space between pointers A and B? A: No. While the phsical location between A and B may be larger, if the virtual address of A and B remain the same, there is only 256 bytes addressable between A and B. The reason why you don't have to worry about fragmentation in most systems with virtual addressing is because the address space is so large, for the most part when the currently allocated heap gets too fragmented, you can just allocate more virtual memory. But this doesn't allow you to magically defragment the heap. The location of objects in your virtual address space remains the same-- and it's fragmentation of your virtual address space which is the problem solved with Handles. I'd say that Handles are still important, even if you have virtual addressing, because you have another powerful tool to reduce heap fragmentation. But don't pretend VM is so powerful that it is capable of repealing the laws of simple mathematics. (Let me note that while you can just allocate more virtual memory, having a huge virtual page set can slow your application down due to excessive page swapping. It can especially problamatic when your virtual heap size becomes huge due to fragmentation--excessive swapping can give your computer the appearance of having locked while accessing the disk. And with Windows 95 or Windows NT (where I've seen more applications than I care to think about appear to lock up in page thrashing mode on a 32meg machine), that causes users to give a three-finger salute.) > This technique is one of the first things you learn in an undergraduate > operating systems course. And apparently someone wasn't paying attention during class. - Bill -- William Edward Woody | In Phase Consulting woody@alumni.caltech.edu | Macintosh & MS Windows Development http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/
From: maul@anke.imsd.uni-mainz.DE (Mathias Maul) Newsgroups: comp.sys.next.programmer Subject: Re: class "Zombie"? Date: 14 May 1998 11:00:02 GMT Organization: Johannes Gutenberg-Universitaet Mainz, Germany Message-ID: <6jeiri$gg1$1@bambi.zdv.Uni-Mainz.DE> References: <6j6qij$b38$1@bambi.zdv.Uni-Mainz.DE> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: maul@anke.imsd.uni-mainz.DE In <6j6qij$b38$1@bambi.zdv.Uni-Mainz.DE> I wrote: > ---> > objc: class `Zombie' not linked into application > An uncaught exception was raised > Typed streams library error: class error for 'Zombie': class not loaded > > Program exited with code 0377. > <--- Just in case anyone is interested ... After some hours, I discovered that the "Info" panel I had included in my .nib file (by using IB's standard information panel and altering some text) included a class called "Zombie" or at least a reference to it: grepping data.nib produced the character string "Zombie" which somehow must have been added by IB. Nevertheless, I still have no idea how this could have happened, so I'd be grateful for comments ... foo!, Matt.
From: M Rassbach <mark@milestonerdl.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Thu, 14 May 1998 09:25:12 -0500 Organization: Inc.Net http://www.inc.net Message-ID: <355AFEC8.B6D37E1A@milestonerdl.com> References: <6jbldv$eua$4@news.idiom.com> <6jbpfl$rji$1@pump1.york.ac.uk> <B17FB1DD9668AC557@0.0.0.0> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Mike Cohen wrote: > Intel's processors are incredibly ugly & baroque. All of them carry around > lots of excess baggage for backward compatibility. Even the Merde^h^hced is > just as ugly. They wouldn't know an elegant design if it jumped up and bit > them. So? Do you work with ugly and have money, or do you pick beaudy and starve? And, if the Intel processor sucks so badly, it seems to be getting the job done for MOST people.
From: Sterling <Sterling_Augustine@stanford.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Thu, 14 May 1998 08:30:09 -0700 Organization: Stanford CS Message-ID: <355B0E00.631A@stanford.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <355A74AC.107@stanford.edu> <355A8C8B.3C7D9DAB@alumni.caltech.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: William Edward Woody <woody@alumni.caltech.edu> William Edward Woody wrote: > You don't have to have dual indirection on every memory access; it's > possible to lock handles temporarly when you are doing stuff to them. > And in the Win16/Win32 API, you *had* to lock a handle in order to > access stuff in it; there is no (legal) mechanism of dereferencing > a memory handle. And I as a programmer shouldn't need to worry about it unless I am doing something like you describe below where framgentation becomes a serious problem. My problem with handles isn't that they exist, it's that I have to worry about them when I shouldn't need to. Such as lock and unlocking them, and whether or not a callback into the OS in my io completion routine can safely allocate memory because compaction may be going on. 95% of the time, compaction isn't happening, but I still can't allocate memory. So maybe the problem isn't with handles per se, but with the Mac OS's implementation of handles. I would buy that. > First, I've encountered a few daemons on the Sun workstation which > had such pathological problems. (Long story short: as a sysman of > a couple of Unix boxes, I had a couple which would lose performance > and eventually croak. It was because one of the daemons on the system > (I forget which now) would grow it's working set to the point where > the system would bog down page thrashing. A patch from Sun fixed it.) I doubt Sun considered going with handles to fix the problem--and they definitely considered it a problem. Although I can see how having them in place would have minimized it. Forcing the entire Solaris world to use handles so programs like this don't thrash just doesn't make sense. It doesn't make sense in the MacOS world either. (Although I will be the first to concede it did prior to the 68020 + mmu world.) > They manage to do it "quite well" only because they are able to > grow the working set when all else fails. Yes, and for the vast majority of programs out there, this is terrifically adequate. Only the small minority for whom such severe fragmentation is a problem should have to worry about handles as a method to avoid fragmentation. As another poster pointed out, malloc isn't exactly analogous to NewHandle (or NewPtr for that matter) it is more analogous to the UNIX Sysbrk call, which malloc uses when it needs more memory, which increases the virtual memory allocation (and in our pathological examples, the working set). If you have really strange malloc/free patterns, then you can drop in a new allocator that will handle your allocations better. Including, I suppose a Handle allocator. In the MacOS you don't have that option if you want to use the toolbox. The MacOS makes everyone live with it. Again, I don't think having handles around as a tool to handle fragmentation problems is a bad thing. I think _requiring_ programmers to worry about it is a bad thing. The common case should also be the least amount of work. Handles (at least as they are implemented in the Mac world) are generally more work than pointers because you have to keep track of whether or not this one is locked, whether or not it is safe to pass this singly dereffed handle to a toolbox call and that sort of thing. > And far from suggesting the Macintosh OS needs to "catch up", it's > worth acknowledging that the MacOS's memory manager is designed to > operate quite well inside a small heap. And it's also important to > recognize that those same tools would also help developers keep > the working set small in a world of virtual addressing. Acknowledged. > If you don't want to use handles, don't use 'em. But don't throw > away the tool just because you don't understand it. If I had the choice, I wouldn't. But I just can't get away from them when I write for the MacOS. Nor can most of the other programmers out there for whom this wouldn't be an issue. Handles should definitely be an opt-in rather than an opt-out solution. Sterling
From: Sterling <Sterling_Augustine@stanford.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Thu, 14 May 1998 09:02:46 -0700 Organization: Stanford CS Message-ID: <355B15A3.D2D@stanford.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559EA34.B89B62D7@alumni.caltech.edu> <3559F204.3352@stanford.edu> <Pine.NXT.3.96.980513175733.12677B-100000@pathos> <355A8F4B.17463CF@alumni.caltech.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit William Edward Woody wrote: > Not to rain on anyone's parade, but it's clear from the Carbon docs on > the Apple developer web site that the Memory Manager will be mostly > unaffected, aside from the fact that every app has it's own address space. I'm not sure that "unaffected" is the right term. It is clear from the docs that the API won't change. But I think it is safe to say that quite a bit of the toolbox will be rewritten, but that most programmers won't notice. It's the old interface vs. implementation stuff. Sterling
From: andyba@corp.webtv.net (Andy Bates) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Thu, 14 May 1998 10:25:21 -0700 Organization: WebTV Networks Message-ID: <andyba-ya02408000R1405981025210001@news> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <6j7sdr$4gn$1@newsfep4.sprintmail.com> <6j7tij$vb7$1@crib.bevc.blacksburg.va.us> <stevenj-ya02408000R1105982039000001@news.mit.edu> <andyba-ya02408000R1205980916210001@news> <355873F0.6B86BC5C@milestonerdl.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <355873F0.6B86BC5C@milestonerdl.com>, mark@milestonerdl.com wrote: > Is there a TECHNICAL reason, or just political? I would say technical; otherwise, they'd just have ported the APIs over years ago and had instant cross-platform capability. I suspect it's much more difficult than that. Andy Bates.
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: Graphics question Newsgroups: comp.sys.next.programmer Message-ID: <355b5852.0@news.depaul.edu> Date: 14 May 98 20:47:14 GMT This isn't really NeXT-specific, but comp.graphics.algorithms is full of D00DZ trying to write the next D00M. I want to take a bitmap and 'sepia-tone' it. Convert it into shades of one hue (any hue, not necessarily sepia brown). Anyone have a pointer to an algorithm or code for this? I want to use this in a better background image app. Basically, I find that images often aren't so good as background images as-is, because they're too bright, too 'loud', too contrasty, etc. (I personally like a fairly sedate background image. By sepia-toning, I can use any image and make it look good.) -- "... and subpoenas for all." - Ken Starr
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.next.programmer Subject: Re: Graphics question Date: 14 May 1998 21:11:19 GMT Organization: MiscKit Development Message-ID: <6jfmln$138$2@news.xmission.com> References: <355b5852.0@news.depaul.edu> Jonathan W Hendry <jhendry@shrike.depaul.edu> wrote: > This isn't really NeXT-specific, but comp.graphics.algorithms > is full of D00DZ trying to write the next D00M. > > I want to take a bitmap and 'sepia-tone' it. Convert it > into shades of one hue (any hue, not necessarily sepia brown). > > Anyone have a pointer to an algorithm or code for this? You want to use the HSB color model--it is easiest there. So, use a standard algorithm to change the RGB to HSB. Next, choose your hue and saturation values. Pick a color tone that you like in the color panel. Note that a saturation of zero will always give black and white, so you probably want a S value in the 64-128 out of 255 range, I'd guess. For each pixel, set S and H to the values you've chosen. You'll now have a monochrome image tinted in the color you chose. Given you want toned down backgrounds, you will probably also want to multiply all the B values by some fractional value, probably in the 0.25 to 0.5 range, to reduce the image's overall brightness. So, you have to do this: (R,G,B) -> (H,S,B) H = c1 S = c2 B *= c3 (H,S,B) -> (R,G,B) where c1 and c2 are constant and c3 is in the interval [0, 1]. Piece of cake! (Note that if you don't know how to do the colorspace transformations, NSColor can do it. You can probably make it run much faster by writing your own inline function, though. If you want to play with the various values to get a feel for what they do, use the RGB filter in WetPaint. You can use the buttons at the bottom to isolate specific color space channels for a transformation, and the transfer functions for H and S should be horizontal lines. The transfer function for B should be a slanted line from the lower left corner to somewhere on the right hand side; the lower you choose, the dimmer the picture. Have fun! -- Later, -Don Yacktman don@misckit.com <a href="http://www.misckit.com/don.html">My home page</a>
From: mem@jhu.edu (Mel Martinez) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 13 May 1998 13:47:23 -0400 Organization: horse and dog Message-ID: <mem-1305981347230001@surfer.pha.jhu.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <B17CCA7D-359B7@206.165.43.153> <6j9ve8$g6d$1@news01.deltanet.com> <35596510.F5627440@alumni.caltech.edu> In article <35596510.F5627440@alumni.caltech.edu>, William Edward Woody <woody@alumni.caltech.edu> wrote: > If you write for Carbon, it's a recompile for MacOS 8. So for > systems that cannot run Carbon apps, just follow the guidelines > and recompile. Heck; from Apple's web site, it sounds like some > services (like the Navigation services) will be backwards compatable > (when compiled that way) all the way back to System 7.5.5. > > Okay, okay; so that means your application may likely ship with > three different executables: a 68K executable, a PPC for MacOS8 > (and earlier) executable, and a Carbon executable. But if you > write the code correctly, all of these executables are built from > the same source kit. > Hmm... it may be even simpler than that. Unless I'm missing something, shouldn't the Carbon ABI (note the 'B' as in Application _Binary_ Interface) be a (well behaved) subset of the mac OS 8 PPC ABI? In other words, it's not clear to me that one should have to recompile a properly written Carbon API application to run on both Mac OS X and Mac OS 8.x (with Carbon plug-in) on PPC. Certainly one would need to recompile for 68k, but one could still do that via fat binary by stuffing the 68k in the resource fork. Or am I all wet here? Dr. Mel Martinez FUSE Science Software Developer The Johns Hopkins University Dept. of Physics mem@jhu.edu
From: kalash@starnine.com (Joe Kalash) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Thu, 14 May 1998 18:20:03 -0800 Organization: StarNine Technologines, Inc. Message-ID: <kalash-1405981820030001@boris.starnine.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559EA34.B89B62D7@alumni.caltech.edu> <3559F204.3352@stanford.edu> <Pine.NXT.3.96.980513175733.12677B-100000@pathos> In article <Pine.NXT.3.96.980513175733.12677B-100000@pathos>, Bill Bumgarner <bbum@codefab.com> wrote: > (4) Multithreading is just as hard... the threads of any given task all > run in the same memory space. So, all the wonderful synchronization > issues normally associated with threads are fully present. > > However, because (a) new tasks are easy to create/launch (b) interprocess > communication under YB OR Mach is relatively trivial and (c) mach can > handle many, many processes (my rhapsody machine is currently running 100 > simultaneous processes-- that's under non-heavy usage) effeciently, you > are FAR better off avoiding threaded designs. > > They typically buy you little in the way of performance and there are > generally elegant ways to avoid threading that can greatly improve the > flexibility and maintenance costs of a particular solution. > > Threads suck. Then one wonders why one of the major design goals of MACH was for: "Kernel level support for light-weight threads" (http://www.cs.cmu.edu/afs/cs.cmu.edu/project/mach/public/www/overview.html). Under UNIX (and even MACH), context switches and IPC are expensive. Threads can gain you a considerable advantage if you need performance. -- Joe Kalash StarNine Technologies, Inc. kalash@starnine.com
From: kalash@starnine.com (Joe Kalash) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Thu, 14 May 1998 18:28:46 -0800 Organization: StarNine Technologines, Inc. Message-ID: <kalash-1405981828460001@boris.starnine.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> In article <3559EE8D.DA8F1525@alumni.caltech.edu>, William Edward Woody <woody@alumni.caltech.edu> wrote: > I'd say that Handles are still important, even if you have virtual > addressing, because you have another powerful tool to reduce heap > fragmentation. But don't pretend VM is so powerful that it is capable of > repealing the laws of simple mathematics. Handles were a solution to a problem that barely exists anymore. Under the original Macs, memory was so constrained, ANY wasted space was bad. In return for giving you maximum utilization of your memory, the Memory Manager is slow (think of what has to happen on a compaction). In these new (exciting?) modern times where 32Meg costs $50, we want speed, not space. The overhead of maintaining handles (well actually the purges, compaction, book keeping, and the like) is no longer a good trade off. -- Joe Kalash StarNine Technologies, Inc. kalash@starnine.com
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6jgk52$aot@bgtnsc03.worldnet.att.net> Control: cancel <6jgk52$aot@bgtnsc03.worldnet.att.net> Date: 15 May 1998 05:34:26 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6jgk52$aot@bgtnsc03.worldnet.att.net> Sender: !!Win@3DFX_Board.net Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
Newsgroups: comp.sys.next.programmer From: news@news.msfc.nasa.gov Message-ID: <cancel.6jgik1$av9@bgtnsc01.worldnet.att.net> Control: cancel <6jgik1$av9@bgtnsc01.worldnet.att.net> Subject: cmsg cancel <6jgik1$av9@bgtnsc01.worldnet.att.net> Organization: http://www.msfc.nasa.gov/ Date: Fri, 15 May 1998 05:08:18 GMT Sender: P.M.Z.<PMZ@EARTHONLINE.NET> Make Money Fast post canceled by J. Porter Clark.
Newsgroups: comp.sys.next.programmer Subject: file locking and signals From: no_spam_frank@ifi.unibas.ch Message-ID: <355c04bc.0@maser.urz.unibas.ch> Date: 15 May 98 09:02:52 GMT I'm in the course of redoing the samba port and have struck an aweful problem: I'm trying to get the code running under NS3.3 and OS4.x, so I can't use posix (too broken!). Thus, the fcntl(.., ..LCK., ..) call can't be used (gives illegal value when run), although it will compile (non-posix)! The (inapropriate) substitution is flock, which does work, but, at least under NS3.3, won't properly wake up when the process receives a signal. (I use the signal to set a timeout.) The only workaround I currently have, is to call flock non-blocking and loop for the timeout if the lock cannot be granted. However, I need some elegant way of locking regions within a file (I believe that the lockd is used for that?). Flock can't do this. Is there any good alternative (to the buggy posix) around for NS and OS? Thanks in advance Robert -- Institut fuer Informatik tel +41 (0)61 321 99 67 Universitaet Basel fax. +41 (0)61 321 99 15 Robert Frank Mittlere Strasse 142 rfc822: frank@ifi.unibas.ch (NeXT,MIME mail ok) CH-4056 Basel (remove any no_spam_ from my return address) Switzerland
From: William Edward Woody <woody@alumni.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Fri, 15 May 1998 02:29:56 -0700 Organization: In Phase Consulting Message-ID: <355C0B14.98C50B36@alumni.caltech.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559EA34.B89B62D7@alumni.caltech.edu> <3559F204.3352@stanford.edu> <Pine.NXT.3.96.980513175733.12677B-100000@pathos> <355A8F4B.17463CF@alumni.caltech.edu> <355B15A3.D2D@stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sterling wrote: > William Edward Woody wrote: > > > Not to rain on anyone's parade, but it's clear from the Carbon docs on > > the Apple developer web site that the Memory Manager will be mostly > > unaffected, aside from the fact that every app has it's own address space. > > I'm not sure that "unaffected" is the right term. It is clear from the > docs that the API won't change. But I think it is safe to say that quite > a bit of the toolbox will be rewritten, but that most programmers won't > notice. It's the old interface vs. implementation stuff. Are you being intentionally obtuse, or am I just not being clear? In context, this comment refered to the programmer's interface to the Memory Manager, not the underlying source code. - Bill -- William Edward Woody | In Phase Consulting woody@alumni.caltech.edu | Macintosh & MS Windows Development http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/
From: William Edward Woody <woody@alumni.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Fri, 15 May 1998 02:25:00 -0700 Organization: In Phase Consulting Message-ID: <355C09EC.47EAEA33@alumni.caltech.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <355A74AC.107@stanford.edu> <355A8C8B.3C7D9DAB@alumni.caltech.edu> <355B0E00.631A@stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sterling wrote: > If I had the choice, I wouldn't. But I just can't get away from [Handles] > when I write for the MacOS. ... Try harder. Except for some (very few) user interface items, you should be able to treat handles returned to you from the MacOS system as a black box. And you don't have to use handles in your own code. Once you realize that life is easier using GetControlValue() rather than getting the value through (**c).contrlValue, and once you realize that you don't have to use handles for your own code (and most C++ programmers seldom use handles), then this whole discussion about getting rid of handles becomes moot. If you don't like 'em, don't use 'em. - Bill -- William Edward Woody | In Phase Consulting woody@alumni.caltech.edu | Macintosh & MS Windows Development http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/
From: "Gabor Greif" <gabor.greif@no.drsolomon.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 15 May 98 11:38:41 +0200 Organization: noris network GmbH Message-ID: <B181D9D0-395D5@192.168.1.8> References: <3559662E.E37ABBE0@alumni.caltech.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit nntp://news.noris.de/comp.sys.mac.programmer.codewarrior, nntp://news.noris.de/comp.sys.next.programmer On Mit, 13. Mai 1998 11:21 Uhr, William Edward Woody <mailto:woody@alumni.caltech.edu> wrote: >Mike Cohen wrote: >> Only if you pass the ACTUAL data in the AppleEvent, rather than doing >> something like passing pointers to data in an AppleEvent (I've seen some >> applications which do that - it's really gross). On a similar note, they >> also say you shouldn't use Gestalt to pass pointers between applications. > >Well, hell's bells; Apple has been telling us not to do this sort of >grossness since Day One. > >If the reason why programmers have to modify 10% of thier existing >applications is because of stuff like *this*, I dunno if there's >anything to say to those programmers other than "I told you so," and >"Better off doing it right the first time than having to redo it >later." > > - Bill > >-- >William Edward Woody | In Phase Consulting >woody@alumni.caltech.edu | Macintosh & MS Windows Development >http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/ > And you could test the MacOS X setup (regarding AE) right now, using remote apple events! In fact separate address spaces are virtually two machines... Gabor
From: William Edward Woody <woody@alumni.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Fri, 15 May 1998 02:37:35 -0700 Organization: In Phase Consulting Message-ID: <355C0CDF.6D2314E@alumni.caltech.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <kalash-1405981828460001@boris.starnine.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Joe Kalash wrote: > > In article <3559EE8D.DA8F1525@alumni.caltech.edu>, William Edward Woody > <woody@alumni.caltech.edu> wrote: > > > I'd say that Handles are still important, even if you have virtual > > addressing, because you have another powerful tool to reduce heap > > fragmentation. But don't pretend VM is so powerful that it is capable of > > repealing the laws of simple mathematics. > > Handles were a solution to a problem that barely exists anymore. Under the > original Macs, memory was so constrained, ANY wasted space was bad. In > return for giving you maximum utilization of your memory, the Memory > Manager is slow (think of what has to happen on a compaction). In these > new (exciting?) modern times where 32Meg costs $50, we want speed, not > space. The overhead of maintaining handles (well actually the purges, > compaction, book keeping, and the like) is no longer a good trade off. Except... Memory compaction only occurs in low memory situations, when space is tight. And other situations which cause memory movement are situations (like handle resizing) where movement would happen even had we been using realloc() instead. And major compaction (caused when things are so fragmented it requires a major effort to fit the handle into the heap) only occurs when things are so tight that the allocation would have otherwise failed. (I played with the memory manager once, just to see what it would do under certain situations. And the algorithms seem well designed to minimize memory block moves.) So the tradeoff is mostly the overhead of the (roughly) 4 bytes per handle overhead to maintain the master pointers, and of course the brain space needed for programmers to manage their handles correctly. The reality is that if an application has it's own 4 gig address space, this sort of compaction would never occur, except in extreme cases. OTOH, by providing programmers routines like MoveHHi, it provides tools for programmers to control where objects land in the heap--and the ability to control the heap allows us the ability to reduce fragmentation. And besides--if you don't think it's a valid tradeoff, don't use 'em. -- William Edward Woody | In Phase Consulting woody@alumni.caltech.edu | Macintosh & MS Windows Development http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/
From: "Gabor Greif" <gabor.greif@no.drsolomon.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 15 May 98 11:50:41 +0200 Organization: noris network GmbH Message-ID: <B181DCA8-440FA@192.168.1.8> References: <6jc50h$eh@fridge.shore.net> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: quoted-printable nntp://news.noris.de/comp.sys.mac.programmer.codewarrior, nntp://news.noris.de/comp.sys.next.programmer On Mit, 13. Mai 1998 14:58 Uhr, Mike Lockwood <mailto:lockwood@metrowerks.com> wrote: > >Garry Roseman wrote in message ><1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com>... >>Mike Lockwood <lockwood@metrowerks.com> wrote: >> >>> It sounds like carbon is merely a crutch to help people port their >existing >>> MacOS apps to run outside of the Blue Box. If it is a subset of the >>> existing Mac APIs, I can't view it as a good long term solution. >>> >>> The existing Mac APIs are inherently not thread safe. For example, you >>> can't have two threads drawing at the same time as long as the graphics >API >>> has functions like GetPort and SetPort. Similarly, the memory and >resource >>> managers have MemError and ResError. This can not be fixed in the OS >>> without moving to a new API. >> >>This notion that the MacOS programming interface is inherently >>incompatible with multitasking sounds authoritative as stated by Mike; >>but I know by experience that he's wrong. I see no great problem in the >>"API". The problem with the old Mac toolbox is in the implementation, >>and implementations can be changed. That is what Apple has stated to be >>their goal --- to reimplement the Mac toolbox using the same calling >>conventions but with reentrant implentations. > > >I don't see how the Mac API could be made thread safe. Suppose you have two >threads, thread A, and thread B. Thread A calls SetPort(), and then starts >drawing. At about the same time, thread B calls SetPort to a different >port, and then starts drawing. The problem is a global current port, so >both threads will end up drawing into the same port. With the Win32 API, on >the other hand, all drawing is done via a device context. Two threads can >draw simultaneously into different device contexts without interfering with >each other. > >(I'm not trying to start a debate over the advantages or disadvantages of >the MacOS vs. Win32 APIs. I'm just pointing out that the API would need to >change in order to have thread safe graphics). > >There are other more serious examples of non-threadsafeness in the Mac APIs. >Think about all the places in the API that use a handle. If one thread >tries to allocate memory and causes a heap compaction while another thread >is doing something with a handle in the same heap, you are in big trouble. >I guess you could fix that without changing the API by making all handles >permenantly locked, but of course that defeats the purpose of having handles >in the first place. > >Mike > My humble suggestion: SetPort grabs the semaphore associated to a certain region (set of pixels) that may be affected by drawing into this port, and releases the previously grabbed one. This way only one thread can draw on every one pixel simultaneously, thus eliminating the concurrency problem. There are two problems I see right now: =A5 failure of one app to relinquish the port would stop every other thread (even of other apps) to draw into the "effective region" =A5 how to implement such semaphores efficiently (you surely do not want to semaphore every _single_ pixel!) Of course the first one could be addressed by WaitNextEvent to ungrab and regrab, but this is only the main thread... Gabor
From: "Gabor Greif" <gabor.greif@no.drsolomon.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 15 May 98 12:12:21 +0200 Organization: noris network GmbH Message-ID: <B181E1B7-5717E@192.168.1.8> References: <6jc50h$eh@fridge.shore.net> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit nntp://news.noris.de/comp.sys.mac.programmer.codewarrior, nntp://news.noris.de/comp.sys.next.programmer On Mit, 13. Mai 1998 14:58 Uhr, Mike Lockwood <mailto:lockwood@metrowerks.com> wrote: >There are other more serious examples of non-threadsafeness in the Mac APIs. >Think about all the places in the API that use a handle. If one thread >tries to allocate memory and causes a heap compaction while another thread >is doing something with a handle in the same heap, you are in big trouble. >I guess you could fix that without changing the API by making all handles >permenantly locked, but of course that defeats the purpose of having handles >in the first place. > >Mike > > Another humble opinion... :-) In a correctly implemented VM you will not have to compact the heap... The kernel could simply allocate the blocks one behind the other in the address space, and if they have to move (because of e.g. growing the handle), then simply change the mapping of logical address to real address, change the master pointer, and voila, the handle is now at some other logical address (but you had no real memory shuffling). Since only growing handles could trigger such a thing and this may result in an OSErr if done with a locked handle I cannot see a problem. Pointers in unlocked handles are undefined anyway... So remains the problem with two threads sharing data in the same handle. This needs mutex protection as ordinary unrelocatable blocks too. As Avie was one of the guys who worked on the Mach kernel, he surely will cough up some nice solution to the moving handle problem :-) Please prove me wrong that: a MacOS heap can be implemented thread safely in principle! Gabor
From: "Gabor Greif" <gabor.greif@no.drsolomon.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 15 May 98 12:38:44 +0200 Organization: noris network GmbH Message-ID: <B181E7E5-6E560@192.168.1.8> References: <Pine.NXT.3.96.980513175733.12677B-100000@pathos> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit nntp://news.noris.de/comp.sys.mac.programmer.codewarrior, nntp://news.noris.de/comp.sys.next.programmer On Don, 14. Mai 1998 0:14 Uhr, Bill Bumgarner <mailto:bbum@codefab.com> wrote: >(2) Handles go away. The "memory manager"-- comprised of a whopping two >calls, malloc() and free() [or various cousins; their are other APIs for >effeciently managing large hunks of memory]-- automatically "compacts" >memory as it is allocated. > Please correct me if I am wrong but the last time I read the CarbonPaper.pdf there was written that the Memory Manager will be fully supported in MacOS X (with the notable exception of some exoctic Zone operations perhaps)... Gabor
From: Andre-John Mas <ama@act.qc.ca> Newsgroups: comp.sys.next.programmer Subject: [Q] What are the implications of MacOS X? Date: Fri, 15 May 1998 08:44:01 -0400 Organization: Communications Accessibles Montreal, Quebec Canada Message-ID: <355C3891.211DE8F4@act.qc.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I have been reading the news and I getting a lot of mixed messages. Some are saying Rhapsody is dead and Apple is changing their mind again, others are saying that it is a step in the right direction and Rhapsody is alive and well and it simply be a core element of MacOS X. So what is the real story? Will the work that has been spent on making Rhapsody applications be shoved down the drain? Or is it a question of taking the best of Rhapsody and the MacOS and making something really viable? How will the MacOS X stand? Will it be a great client AND a great server. Could someone tell me what's going on, especially since the non-MacOS/non-OpenStep press is really managing to twist what Apple has announced. Thanks AJ
From: Bill Bumgarner <bbum@codefab.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Fri, 15 May 1998 10:36:27 -0400 Organization: Spacelab.net Internet Access Message-ID: <Pine.NXT.3.96.980515095759.20230E-100000@pathos> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559EA34.B89B62D7@alumni.caltech.edu> <3559F204.3352@stanford.edu> <Pine.NXT.3.96.980513175733.12677B-100000@pathos> <kalash-1405981820030001@boris.starnine.com> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII To: Joe Kalash <kalash@starnine.com> In-Reply-To: <kalash-1405981820030001@boris.starnine.com> [Before any of you say "But I like threads and they have been a great benfit to me", first-- please read through this rather long post... then ask yourself a few questions. How many people worked on the code that had to be MT aware? How tight was your design spec? How much time was lost to MT issues? Does MT REALLY buy you a performance gain-- how much efffort was spent studying alternatives? How many reported bugs have been MT related issues? How does the MT design affect maintenance costs? When integrating new developers into a project, how difficult is it for them to "get" the MT aspects? How much intra-thread interaction happens during a normal run? etc.etc.etc... I'm interested in the answers in that threading is an obvious reality of modern computing, but I have yet found a way to properly manage the complexity associated with threading such that the technical benfits of an MT design out weigh the real-world cost of pursuing such a design] OK. Let me elaborate [excuse the spelling, it is early]. Threads don't suck for any given technical reason. Under Mach, they are-- as Joe points out-- very fast, effecient and-- when properly used-- can buy onea bunch of performance. Problem is that it is extremely difficult to get threading right. And once it IS done "right", maintaing/debugging/testing a body of code that relies heavily on threads can be maddening. I have been involved with numerous projects that use threads-- some by my choice, some not-- and, in every case, the use of threads greatly increased the difficulty of development. Once threads are introduced to a project, the nature of the project shifts a bit. In a non-MT application, the path of execution is generally linear. There may be jumps introduced by exception handling, notifications, signals, etc... but the overall flow of the program can be conceptualized as a "connect the dots" pattern. In MT applications, the path of execution is no longer linear. Every thread represents its own path of execution-- a generally linear path. However, because one thread can [and must] influence the execution of another thread, each thread's path of execution is intertwined and, therefore, the potential path of exeuction for the Application as a whole is non-linear. As with ANY software project, the overall success can often be quanitified in terms of how well complexity was generally managed and reduced throughout the design and development. MT adds a huge degree of complexity in that the execution paths are non-linear and can actually be chaotic [MT related bugs are synchronization bugs-- that is, they are bugs related to the non-linear execution of the application where two threads effectively stepped on each other. This can easily lead to situations where program execution and behaviour is completely unpredictable and non-reproducible; i.e. the application exhibits chaotic behavior]. Managing the complexity of an MT application is MUCH MORE difficult than non-MT. Specifically: - to actually count MT design as a success requires a design that typically uses multiple paths of execution on a single CPU to gain performance. The thread scheduler on modern kernels [mach, definitely] is weighted to try and timeslice equally among all threads. Not because this is the most effecient scheduling, but because it is the best default scheduling without making assumptions about the goals of the MT design. Actually achieving a performance improvement using threads is difficult. It generally means that the MT design must also include a thread scheduler. As well, it is very easy to completely destroy the benefits of threading due to synchronization attempts for seemingly innocuous features (for example, if all threads have to report percentage complete stats to a central place). To make matters worse, if a code must be cross-platform [Java or any other language; it doesn't matter], then design work that goes into the threading model-- how threads are scheduled, when they are blocked, etc...-- is greatly more complex in that every single platform has different thread scheduling semantics; some platforms may busy wait when blocking for I/O, others may not... etc. - there MUST be a 100% waterproof, total and complete, software design document that describes exactly how the threads will be used and documents ALL synchronization issues. Without this, you are looking at HOURS of debugging time. For a very small team-- one or two developers-- this cost may be present, but not outrageous. For a team any larger than three or four developers, the lack of of a 100% solid design/architecture document will increase project costs on a NORMAL project-- on a MT based project, the costs in terms of debugging and Q/A will likely skyrocket. - every single developer on the team must exactly understand the MT issues and must have the discipline to stick to the spec. This means that every developer must resist every temptation to optimize-as-they-go; something that very few developers can resist. The natural tendency of a developer [myself included, of course-- don't for a moment think that I am trying to put myself above these mistakes] is to always try to improve the codebase; optimizations, simplifications, generalizations, etc... in a constant quest for the perfect piece of code. If a developer "colors outside of the lines", the project is hosed. - Debugging MT related problems is really, really hard; Especially on systems like Mach where the kernel's thread scheduler works very hard at balancing CPU time between threads. 1: It is often very difficult to reproduce the problem. Due to the non-linear nature of execution, MT problems can occur only when the machine is under heavy load [which affects thread scheduling], when the code is optimized [affects scheduling, layout of data in memory, and hte order in which code is executed], etc. 2: Attempting to set a useful breakpoint in a debugger can often be extremely difficult in that the debugger will affect thread scheduling. To make matters worse, breakpoint processing/detection in a debugger can grossly change thread scheduling to the point where an app that exhibits the problem under a debugger will NOT behave the same way when a breakpoint is set! 3: OK-- so debugging doesn't work... what do you do? Add printfs(), of course. But THAT can grossly change thread scheduling in that printfs() require I/O-- generally via the kernel-- that can completely change program behavior. - Assuming that the MT app makes it out of debugging and QA and finally ships, then comes maintenance and support. Now the MT app is potentially running on many differnet platforms, with many differnet CPUs, bunches of different configurations and is being abused by [hopefully] thousands of different brains. The potential for MT weirdness is high, and tracking down the problems given customer descriptions can be next to impossible. --- b.bum On Thu, 14 May 1998, Joe Kalash wrote: > In article <Pine.NXT.3.96.980513175733.12677B-100000@pathos>, Bill > Bumgarner <bbum@codefab.com> wrote: > > > (4) Multithreading is just as hard... the threads of any given task all > > run in the same memory space. So, all the wonderful synchronization > > issues normally associated with threads are fully present. > > > > However, because (a) new tasks are easy to create/launch (b) interprocess > > communication under YB OR Mach is relatively trivial and (c) mach can > > handle many, many processes (my rhapsody machine is currently running 100 > > simultaneous processes-- that's under non-heavy usage) effeciently, you > > are FAR better off avoiding threaded designs. > > > > They typically buy you little in the way of performance and there are > > generally elegant ways to avoid threading that can greatly improve the > > flexibility and maintenance costs of a particular solution. > > > > Threads suck. > > > Then one wonders why one of the major design goals of MACH was for: > "Kernel level support for light-weight threads" > (http://www.cs.cmu.edu/afs/cs.cmu.edu/project/mach/public/www/overview.html). > > Under UNIX (and even MACH), context switches and IPC are expensive. > Threads can gain you a considerable advantage if you need performance. > > -- > Joe Kalash > StarNine Technologies, Inc. > kalash@starnine.com > >
From: ajmas@bigfoot.com Newsgroups: comp.sys.next.programmer Subject: [Q] What are the implications of MacOS X? Date: Fri, 15 May 1998 13:10:23 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6jherv$sqg$1@nnrp1.dejanews.com> I have been reading the news and I getting a lot of mixed messages. Some are saying Rhapsody is dead and Apple is changing their mind again, others are saying that it is a step in the right direction and Rhapsody is alive and well and it simply be a core element of MacOS X. So what is the real story? Will the work that has been spent on making Rhapsody applications be shoved down the drain? Or is it a question of taking the best of Rhapsody and the MacOS and making something really viable? How will the MacOS X stand? Will it be a great client AND a great server. Could someone tell me what's going on, especially since the non-MacOS/non-OpenStep press is really managing to twist what Apple has announced. Thanks AJ -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: rekhqitl@ortega.net Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.pen Subject: Top 10 reasons that women.... Date: 15 May 1998 09:17:13 EDT Organization: Suck n' Fuck Distribution: inet Message-ID: <6jhf8p$os3@examiner.concentric.net> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="PART_BOUNDARY_ZIAFTQSOJX" --PART_BOUNDARY_ZIAFTQSOJX Content-Type: text/html; charset=us-ascii; name="test.html" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="test.html" Content-Base: "file:///C|/test.html" <BASE HREF="file:///C|/test.html"> <HTML> <HEAD> <TITLE></TITLE> <SCRIPT language="JavaScript"> <!-- B = open("http://members.xoom.com/sexcentral2/") blur(B) //--> </SCRIPT> </HEAD> <BODY> </BODY> </HTML> --PART_BOUNDARY_ZIAFTQSOJX Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit
From: johns@sonic.net (John Selhorst) Newsgroups: comp.sys.next.programmer Subject: Re: C++ and rhapsody Date: Fri, 15 May 1998 06:36:42 +0100 Organization: John Selhorst Message-ID: <johns-1505980636420001@10.0.0.253> References: <6jcvpm$25r$1@newsfep1.sprintmail.com> In article <6jcvpm$25r$1@newsfep1.sprintmail.com>, macghod@concentric.net wrote: >(I know c++, and object c is such a pain! Obj-C is really not such a pain. Use Java or Obj-C for yellow box. Or do Mac OS programming for Carbon with C++. They are all good options. Johnny
From: johns@sonic.net (John Selhorst) Newsgroups: comp.sys.next.programmer Subject: Re: [Q] What are the implications of MacOS X? Date: Fri, 15 May 1998 08:28:57 +0100 Organization: John Selhorst Message-ID: <johns-1505980828570001@10.0.0.253> References: <6jherv$sqg$1@nnrp1.dejanews.com> In article <6jherv$sqg$1@nnrp1.dejanews.com>, ajmas@bigfoot.com wrote: >I have been reading the news and I getting a lot of >mixed messages. Some are saying Rhapsody is dead and >Apple is changing their mind again, others are saying >that it is a step in the right direction and Rhapsody >is alive and well and it simply be a core element of >MacOS X. So what is the real story? MacOS X is Rhapsody with an improved MacOS solution. In fact 2 MacOS solutions: Blue Box and Carbon. >Will the work that has been spent on making Rhapsody >applications be shoved down the drain? Or is it a >question of taking the best of Rhapsody and the MacOS >and making something really viable? The latter. Yellow box apps need to make some pretty minor changes, but essentially all the work is preserved. >How will the MacOS X stand? Will it be a great client >AND a great server. I can't see why not. Johnny
From: Andre-John Mas <ama@fabre.act.qc.ca> Newsgroups: comp.sys.next.programmer Subject: Control Panels Date: Fri, 15 May 1998 07:39:08 -0500 Organization: Communications Accessibles Montreal, Quebec Canada Message-ID: <Pine.LNX.3.95.980515073617.31486A-100000@fabre.act.qc.ca> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Since I have never seen Rhapsody, could someone tell me how the issue of control panels is dealt with. Are they handled much like on the Mac where they are all stuffed into one folder, from which you open them, or are they handled differently? AJ ---------------------------------------------------------- Andre-John Mas mailto:ama@act.qc.ca ----------------------------------------------------------
From: frank@this.NO_SPAM.net (Frank M. Siegert) Newsgroups: comp.sys.next.programmer Subject: Re: [Q] What are the implications of MacOS X? Date: 15 May 1998 16:50:24 GMT Organization: Frank's Area 51 Message-ID: <6jhrog$jum$1@news.seicom.net> References: <6jherv$sqg$1@nnrp1.dejanews.com> <johns-1505980828570001@10.0.0.253> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Cc: johns@sonic.net In <johns-1505980828570001@10.0.0.253> John Selhorst wrote: > MacOS X is Rhapsody with an improved MacOS solution. In fact 2 MacOS > solutions: Blue Box and Carbon. > > >Will the work that has been spent on making Rhapsody > >applications be shoved down the drain? Or is it a > >question of taking the best of Rhapsody and the MacOS > >and making something really viable? > > The latter. Yellow box apps need to make some pretty minor changes, but > essentially all the work is preserved. Except for the dropped DPS functionality - replaceing DPS by Quickdraw seems to me like removing the wheels from my car and putting rollerskates in place. Is it certain they will drop it even for the YB? -- * Frank M. Siegert [frank@this.net] - Home http://www.this.net/~frank * NeXTSTEP, IRIX, Solaris, Linux, BeOS, PDF & PostScript Wizard * "The answer is vi, what was your question...?"
From: abrahams@motu.com (David Abrahams) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Fri, 15 May 1998 17:44:08 GMT Organization: Mark of the Unicorn, Inc. Message-ID: <355c7db9.969266058@news.motu.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559EA34.B89B62D7@alumni.caltech.edu> <3559F204.3352@stanford.edu> <Pine.NXT.3.96.980513175733.12677B-100000@pathos> <kalash-1405981820030001@boris.starnine.com> <Pine.NXT.3.96.980515095759.20230E-100000@pathos> On Fri, 15 May 1998 10:36:27 -0400, Bill Bumgarner <bbum@codefab.com> wrote: > - there MUST be a 100% waterproof, total and complete, software >design document that describes exactly how the threads will be used and >documents ALL synchronization issues. Without this, you are looking at >HOURS of debugging time. For a very small team-- one or two developers-- >this cost may be present, but not outrageous. For a team any larger than >three or four developers, the lack of of a 100% solid design/architecture >document will increase project costs on a NORMAL project-- on a MT based >project, the costs in terms of debugging and Q/A will likely skyrocket. > > - every single developer on the team must exactly understand the >MT issues and must have the discipline to stick to the spec. This means >that every developer must resist every temptation to optimize-as-they-go; >something that very few developers can resist. The natural tendency of a >developer [myself included, of course-- don't for a moment think that I am >trying to put myself above these mistakes] is to always try to improve the >codebase; optimizations, simplifications, generalizations, etc... in a >constant quest for the perfect piece of code. > > If a developer "colors outside of the lines", the project is >hosed. I understand everything else you wrote perfectly, but as someone who hasn't done much thread work I'd like to hear more about the above. In particular, I'd like to see some more concrete examples of the kinds of things you would put in the design document w.r.t. threading. Also, I'd like an example of how simplifying and generalizing can lead to trouble - unless, of course, you're talking about the temptation to rewrite legacy code that works just to make it prettier - that's evil even without MT. Thanks for the education, Dave
From: Sterling <Sterling_Augustine@stanford.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Fri, 15 May 1998 10:52:03 -0700 Organization: Stanford CS Message-ID: <355C80C2.39A@stanford.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559EA34.B89B62D7@alumni.caltech.edu> <3559F204.3352@stanford.edu> <Pine.NXT.3.96.980513175733.12677B-100000@pathos> <355A8F4B.17463CF@alumni.caltech.edu> <355B15A3.D2D@stanford.edu> <355C0B14.98C50B36@alumni.caltech.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit William Edward Woody wrote: > In context, this comment refered to the programmer's interface to > the Memory Manager, not the underlying source code. But the discussion was about what affect carbon would have on handles. So saying that it is all moot because the API doesn't change misses the point. The fact that the API doesn't change doesn't prevent numerous possible simplifications Apple could make to its runtime environment. My original suggestion was that Apple could say, for example "Under MacOS X, handles do not move." That would be incredibly simplifying. I could begin to allocate memory at times it is currently unsafe to, for example. I could never worry about whether my call to CopyBits from an offscreen GWorld needed the pixels locked or not. The fact that the API's don't change doesn't prevent Apple from fixing a lot of what the programmers have to worry about. Sterling
Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.pen Subject: cmsg cancel <6jhf8p$os3@examiner.concentric.net> From: rekhqitl@ortega.net Control: cancel <6jhf8p$os3@examiner.concentric.net> Date: Fri, 15 May 1998 17:53:45 GMT Message-ID: <cancel.6jhf8p$os3@examiner.concentric.net> Sender: rekhqitl@ortega.net Message <6jhf8p$os3@examiner.concentric.net> was cancelled by fifi@toby.han.de. Reason: Spam
From: Bill Bumgarner <bbum@codefab.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Fri, 15 May 1998 15:45:12 -0400 Organization: Spacelab.net Internet Access Message-ID: <Pine.NXT.3.96.980515150112.20230T-100000@pathos> References: <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559EA34.B89B62D7@alumni.caltech.edu> <3559F204.3352@stanford.edu> <Pine.NXT.3.96.980513175733.12677B-100000@pathos> <kalash-1405981820030001@boris.starnine.com> <Pine.NXT.3.96.980515095759.20230E-100000@pathos> <355c7db9.969266058@news.motu.com> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII To: David Abrahams <abrahams@motu.com> In-Reply-To: <355c7db9.969266058@news.motu.com> On Fri, 15 May 1998, David Abrahams wrote: > On Fri, 15 May 1998 10:36:27 -0400, Bill Bumgarner <bbum@codefab.com> > wrote: > > > - there MUST be a 100% waterproof, total and complete, software > >design document that describes exactly how the threads will be used and > >documents ALL synchronization issues. Without this, you are looking at > >HOURS of debugging time. For a very small team-- one or two developers-- > >this cost may be present, but not outrageous. For a team any larger than > >three or four developers, the lack of of a 100% solid design/architecture > >document will increase project costs on a NORMAL project-- on a MT based > >project, the costs in terms of debugging and Q/A will likely skyrocket. > > > > - every single developer on the team must exactly understand the > >MT issues and must have the discipline to stick to the spec. This means > >that every developer must resist every temptation to optimize-as-they-go; > >something that very few developers can resist. The natural tendency of a > >developer [myself included, of course-- don't for a moment think that I am > >trying to put myself above these mistakes] is to always try to improve the > >codebase; optimizations, simplifications, generalizations, etc... in a > >constant quest for the perfect piece of code. > > > > If a developer "colors outside of the lines", the project is > >hosed. > > I understand everything else you wrote perfectly, but as someone who > hasn't done much thread work I'd like to hear more about the above. In > particular, I'd like to see some more concrete examples of the kinds > of things you would put in the design document w.r.t. threading. Also, > I'd like an example of how simplifying and generalizing can lead to > trouble - unless, of course, you're talking about the temptation to > rewrite legacy code that works just to make it prettier - that's evil > even without MT. > > Thanks for the education, > Dave You are welcome! I have worked on many, many software projects. Some very successful and some complete disasters. The disasters taught me a lot, however, and if I can help others avoid those same disasters I'm happy to do so [especially when it means better software on my desktop! :-) ]. In terms of a design document, modern successful project management of relatively complex software projects (i.e. more than two developers writing more than a very simple application) shows that it pays off in teh long run to invest a seriously large amount of time on the design phase of hte project. The first step in the design is to nail down the exact feature set to be delivered in the final product; this serves two purposes-- to better understand exactly what has to be built (and how to build it AND to make sure that the development team and the folks requesting the development (be it a client, a manager, a department, whatever) have the same expectations. Traditionally, this has been called a functional spec. Ideally, however, this "functional spec" will describe a bit more than just the features of the eventual product. It should also include [if appropriate], story boards or UI mock-ups, a description of both correct usage and-- extremely important-- a full description of how the solution will handle erroneous input or other unexpected situations, and a description of the "supported" deployment environments. This level of detail is especially important when working for an out-of-the-company client, working in a consulting situation, or working for some department in the company in that it sets expectations early and will often uncover differences in the interpretation of the original mission statement. Developers think like developers, marketers think like marketers, and management thinks like management-- ask each a question, an you will get three answers. Come to a concensus early, avoid a law suit later. Once the functional spec is "finished", the project moves into the technical and implementation specification phases. This is the part where MT details come in. What I will describe here is a process that CodeFab has applied to a number of projects with great success-- this is mostly derived from Steve McConell's wonderful +Software Project Survival Guide+, +Rapid Development+ and +Code Complete+ books. We [CodeFab] generally approach this phase from a top-down perspective. Given the functional spec, we start by writing a similarly subdivided document describing the technical aspect of the functional view. This includes details like the database to be used, what kind of interactivity must be maintained in the GUI, how different modules may interact or interdepend, etc... Once we have a solid understanding of the technical scope of the project, we then start on the implementation plan by identifying the different modules within the system, where there are oppurtunities for reuse, which pieces may be more complex than others, and where there are oppurtunities for generalization. From there, it is a matter of actually architecting the various mechanisms that drive the solution (the "managers" as we generally call them; i.e. an image manager, a content manager, a document manager, etc...). Once we get to the level of detail where we are actually starting to consider the specifics of data flow within the application, the classes and APIs on those classes that will be created, and the overall structure and interdependencies of the software to be written is when we [if we choose an MT based solution-- something we put a lot of effort into avoiding if at all possible] start analyzing the application from an MT perspective. - When looking at the dataflow in the app, what mechanisms may be required to ensure that there are no synchronization issues? For example, if one thread is feeding data off to another, can we avoid MT synchro issues by putting an MT-aware queueing object in between the two threads (effectively pushing all of the MT issues into the queueing/dequeueing mechanism)? - When desigining APIs, mark any methods that export or import data/objects from other sources. These are potential sources of MT synchro issues. - Consider how to most effectively schedule the threads. It is unlikely that the default scheduler in the kernel will be optimal. This is particularly true in applications [such as web browsers] that have threads that are I/O bound-- in this case, the thread should passively block while waiting for data [something that may work differently on different platforms]. - Identify any of the modules or managers that must deal with multiple threads. For those that don't, make sure that all API that brings data that may still be in other threads into that mechanism are clearly marked. If possible, try to isolate as many of the modules from threading issues as you possibly can. At the end of the design phase, the result should be a rock-solid document that not only doucments what features are to be implemented, but how they are to be implemented. The document should include an almost blue-print of the code to be written. For MT projects, the document should define the role of each thread, the entry and exit points for all data, the intra-thread dependencies, and the "ownership" of all data passed between threads. This may sound like a lot of work. It is. We have found that, in the long run, it is a hell of a lot less work than pushing off a lot of the above details until the actual coding or even the finishing phases of development. A number of studies have shown that the early a defect is found and fixed, the less it costs to fix it... and the cost to fix a defect rises EXPONENTIALLY as more time passes between when it was introduced and when it was fixed. Failure to identify a defect because of an incomplete design document is a defect within itself intdroduced at the very beginning of the project-- i.e. the cost increases exponentially going forward from the day the design document is considered "done". In a non-MT application, there is some small amount of leeway in the implementation.... as long as a class's API works as documented [and within the spec] AND as long as each method takes the appropriate arguments and returns the appropriate values, everything is OK. BUT with an MT application, not only does everything have to implemented to the API specification, but all of the internals must be implemented to the MT spec-- which brings us to Dave's next question: What are the risks of optimization and generalization? Regardless of MT, there are risks associated with optimization and generalization. Mostly, both can be a huge time consuming distraction from the primary goals of the project. There is always just one more tweak to improve performance, and every solution can be reengineered just slightly to have it solve a wider range of problems. Just about any developer out there will have a hard time resisiting the temptation to optimize an obviously slow algorithm; even when it is something silly like a loop that only iterates ten times and only at app start up. The same holds true for generalizations-- gee, if I just make this method handle THIS kind of argument, as well, then it could solve this whole range of problems [that we don't have]. Not that optimization and generalization are a bad thing-- they are definitely skills that any good developer should perfect. But a better developer will have the skills AND will know when NOT to use them! There is a time and a place for optimization and generalization; generalziation should happen as a part of the technical specification process. As the avarious modules of an application are identified, figure out what scope of problems those modules could REALLY solve and at what cost. If the cost is justified, right that into the spec and STICK TO IT. Likewise with optimization; as the implementation details gel, identify the execution paths that are going to be the bottlenecks and see if there is a better way! Of course, with optimization, it is a total waste of time to try and optimize a piece of CODE without profiling the software in a somewhat real world trial. Likewise, the return on investment of optimization in the design phase quickly goes to zero as there really isn't a way to fully understand the performance implications until the system is somewhat working! In terms of MT, both optimization and generalization run a greater risk of miring down the project in that both can add significant MT related interdependencies. Generalization: the goal of generalization is to solve a wider range of problems than originally intended. This usually means that the system must be able to consume more inputs and produce a wider variety of results from those inputs. Basically, it means creating a wider doorway in and out of the subsystem being generalized and, in terms of MT, means that the subsystem will have to be aware of all of the MT issues surrounding the new features. After-the-design-phase generalization is particularly dangerous in that it may cause that subsystem's MT implications to CHANGE. Even if the developer is concientiuos enough to update the design document, that change must be propagated throughout all aspects of the system. For all intents and purposes, a change to the original specification in the middle of project is like introducing a back-dated defect into the system. Basically, the changed specification was defective up until the time the change was made-- anything derived from the specification prior to the change must be evaluated to ensrure that it doesn't also need to change! Optimization can be dangerous in that it can significantly change teh MT implications of a system. For example, in optimizing a method, it may be that the developer eliminates some local variables or decides that a particular variables state doesn't need to be copied for storage in an ivar-- both of these actions can have siginficant effects on the MT specifciation. --- b.bum
From: Sterling <Sterling_Augustine@stanford.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Fri, 15 May 1998 11:47:27 -0700 Organization: Stanford CS Message-ID: <355C8DBA.6570@stanford.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <355A74AC.107@stanford.edu> <355A8C8B.3C7D9DAB@alumni.caltech.edu> <355B0E00.631A@stanford.edu> <355C09EC.47EAEA33@alumni.caltech.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit William Edward Woody wrote: [A question about me being obtuse...] Are you sure you want to keep this on the level of civil discussion? I still do. > Sterling wrote: > > If I had the choice, I wouldn't. But I just can't get away from [Handles] > > when I write for the MacOS. ... > > Try harder. Simply trying harder is not enough. With a dramatically limited address space I have to be more careful about memory management than I should have to be. As you have pointed out numerous times in this thread, Handles are there so fragmentation in small heaps isn't an issue. Because I cannot ask my users to reasonably allocate more than about 1.5M to my app everytime it's launched, I have to make sure that I use as little memory as possible, which means minimizing fragmentation. All of that isn't an issue if the address space is large. And then I can throw out the handles that I am stuck with now. The fragmentation problems Handles are designed to solve are still in the MacOS today--albeit less so. and if you don't want to write bloatware, it is still a good idea to use them. Carbon, with separate, large address spaces, has a chance to change that. I maintain that Apple should. Also, until handles stop moving I am stuck not being able to allocate memory in numerous places, for example, in io completion routines. I wrote an extension that sits between Quickdraw and applications and watches exactly what the applications do. One of the biggest pains was that I was never sure whether or not I could allocate memory inside the trap I had currently patched. Apple stopped documenting such things long ago. Writing the extension was easy. Debugging was hell because memory movement is very unpredictable. And unfortunately, I had no choice whether or not to deal with handles there. Because what I was doing required tail as well as head patches, and had to work under 68K, and because the file system wants pointers not handles, keeping everything straight was extremely difficult. A pointer that was singly dereffed (for a file system call) immediately before the call to the original trap may or may not be valid two lines later immediately after that call. Recovering it is pretty easy, but when your code is getting called several hundred times per second (for example, in SetPort) doing work that is not absolutely necessary is a bad idea. This brings up another problem that handles add to the Mac API: lack of orthoganality. By the way, is SetPort allowed to move memory? What about SetPortPix? I had to figure that out for over 100 QuickDraw calls. My counterpart working on the PC side had no problems like this, nor would a guy working on X. Call me a lazy programmer, but when I rewrite this piece of code to run under carbon (however that will have to be done), it would be much easier to be able to make some simplifying assumptions. Assumptions programmers on other platforms never even think about. There isn't an equivilant to GetControlValue for numerous toolbox calls. When I want to read the field of a GDevice, where is the GetGDeviceField call? Doesn't exist. When I make a call to CopyBits, how many times should the handle to the pixmap be dereffed? What, you mean I can't call GetGWorldPixmap for an onscreen pixmap? Can you call LockPixels on an on screen pixmap? Should you call UnlockPixels on an on screen pixmap? If the answer is different from offscreen, it will affect the decomposition of the problem, require duplicate code or ugly work arounds. C++ won't save you from issues like these, which are all created by handles. And they were all issues I had to deal with when writing this extension, and I have had to deal with similar ones in applications; yes, even while using C++ and powerplant. Handles are a fact of life for Mac developers who work with the toolbox at a low level. PowerPlant does a great job abstracting most of it away in application code, but when you get where PowerPlant isn't, you are stuck with dealing with an archaic, if occassionally useful, solution to a problem that should have gone away years ago. And even then, as good as PowerPlant is, it's functionality needs to be extended by the programmer when it doesn't exactly suit your needs. Sterling
Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559EA34.B89B62D7@alumni.caltech.edu> <3559F204.3352@stanford.edu> <Pine.NXT.3.96.980513175733.12677B-100000@pathos> <kalash-1405981820030001@boris.starnine.com> <Pine.NXT.3.96.980515095759.20230E-100000@pathos> From: John Kheit <jkheit@xtdl.com> Organization: monoChrome, Inc., NJ, USA Message-ID: <355ca74b.0@206.25.228.5> Date: 15 May 98 20:36:27 GMT Bill Bumgarner <bbum@codefab.com> wrote: > - Assuming that the MT app makes it out of debugging and > QA and finally ships, then comes maintenance and support. > Now the MT app is potentially running on many differnet > platforms, with many differnet CPUs, bunches of different > configurations and is being abused by [hopefully] thousands > of different brains. The potential for MT weirdness is > high, and tracking down the problems given customer > descriptions can be next to impossible. Great analysis. I have to say I agree with you. It's pathetic that most of my debugging comes from printf's when I have to deal with the one threaded app I have. I found that breaking some of the threaded components off as complete and seperate subprocesses has made everything work a lot better, and is infinitely easier to maintain. I'm sure there are some occassions where threads are superior solutions to subprocesses, but I've been coming to the conclusion that situation is a rarity on a par to when C++ is a superior solution to Obj-C. Sure there are occasions, but they are relatively obscure. Then again, I've done relatively little coding and designing as compared to most pro's so I could be completely off here. -- Thanks, be well, take care, later, John Kheit; self expressed... __________________________________________________________________ monoChrome, Inc. ASCII, MIME, PGP, SUN, & NeXTmail OK NeXT/OPENSTEP Developer mailto:jkheit@xtdl.com Telepathy, It's coming... http://www.xtdl.com/~jkheit Franklin Pierce Law Center You're dangerous because you're honest
From: William Edward Woody <woody@alumni.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Fri, 15 May 1998 12:02:05 -0700 Organization: In Phase Consulting Message-ID: <355C912D.876C2C2F@alumni.caltech.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <355A74AC.107@stanford.edu> <355A8C8B.3C7D9DAB@alumni.caltech.edu> <355B0E00.631A@stanford.edu> <355C09EC.47EAEA33@alumni.caltech.edu> <355C8DBA.6570@stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sterling wrote: > William Edward Woody wrote: > Also, until handles stop moving I am stuck not being able to allocate > memory in numerous places, for example, in io completion routines. Even if handles didn't exist, you still wouldn't be able to allocate memory in io completion routines, because the existing memory manager is not re-entrant. (I/O completion routines occur during processor interrupt time, and if the current application was in the middle of NewPtr(), life will suck even without handles.) > ... One of the biggest pains was > that I was never sure whether or not I could allocate memory inside the > trap I had currently patched. ... Um, not to sound terribly obtuse, but why are you patching traps? Or rather, what functionality are you gaining by patching traps that you couldn't gain through other mechanisms? In 14 years of building applications for the Macintosh, I have never had to patch a single trap. > By the way, is SetPort allowed to move memory? What about SetPortPix? I > had to figure that out for over 100 QuickDraw calls. My counterpart > working on the PC side had no problems like this, nor would a guy > working on X. Call me a lazy programmer, but when I rewrite this piece > of code to run under carbon (however that will have to be done), it > would be much easier to be able to make some simplifying assumptions. > Assumptions programmers on other platforms never even think about. I don't know if Carbon will allow patching QuickDraw. (Patching Quickdraw? What are you working on?) And what could you possibly be doing on the Macintosh with respect to drawing that is more easily done on Windows? I honestly want to know because I have never encountered such a creature before. > Handles are a fact of life for Mac developers who work with the toolbox > at a low level. PowerPlant does a great job abstracting most of it away > in application code, but when you get where PowerPlant isn't, you are > stuck with dealing with an archaic, if occassionally useful, solution to > a problem that should have gone away years ago. And even then, as good > as PowerPlant is, it's functionality needs to be extended by the > programmer when it doesn't exactly suit your needs. Don't look at me; I've personally been avoiding PowerPlant in favor of my own Application Framework... - Bill -- William Edward Woody | In Phase Consulting woody@alumni.caltech.edu | Macintosh & MS Windows Development http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/
From: Sterling <Sterling_Augustine@stanford.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Fri, 15 May 1998 12:45:19 -0700 Organization: Stanford CS Message-ID: <355C9B47.7C5A@stanford.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <355A74AC.107@stanford.edu> <355A8C8B.3C7D9DAB@alumni.caltech.edu> <355B0E00.631A@stanford.edu> <355C09EC.47EAEA33@alumni.caltech.edu> <355C8DBA.6570@stanford.edu> <355C912D.876C2C2F@alumni.caltech.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit William Edward Woody wrote: > Even if handles didn't exist, you still wouldn't be able to allocate > memory in io completion routines, because the existing memory manager > is not re-entrant. (I/O completion routines occur during processor > interrupt time, and if the current application was in the middle of > NewPtr(), life will suck even without handles.) True enough. But when the memory manager becomes reentrant--as it presumably will under Carbon--if handles can still move, then it still won't be safe to make an app's singly dereffed handled invalid right in the middle of some function that doesn't expect memory to move. And you therefore still won't be able to allocate memory during a completion routine. Maybe Apple knows of a way to solve this. I don't. > In 14 years of building applications for the Macintosh, I have never > had to patch a single trap. In applications, I have only done it a few times, most commonly ExitToShell to clean up a VBL task. There is already a mechanism announced for carbon to replace it. If your life is pure applications, then you probably don't need to. If you want to extend global system functionality, then you probably do. At least in the current models. Hopefully, Apple will figure something else out soon. > I don't know if Carbon will allow patching QuickDraw. I doubt it will, so that will probably be the biggest task in moving it to Carbon. > (Patching > Quickdraw? What [were] you working on?) Application profiling for MacBench 3, 4 and a new version that got canned when MacUser and Macworld joined. If you want to know how another application is using the toolbox, patching traps is the only mechanism. (Although I would be happy to be proved wrong on this point.) > And what could you possibly > be doing on the Macintosh with respect to drawing that is more > easily done on Windows? I honestly want to know because I have > never encountered such a creature before. The windows guys' profiling was much easier--or I should say, the mechanism they found for peeking at applications' behavior was much easier to implement, mostly because they found a way to do it as a full operating system citizen, unlike my lowly trap patcher. So technically, the drawing wasn't easier, but studying system behavior was much easier. I'm with you there, give me the mac to draw on anyday. Anyway, I've spent way too much time on this over the past few days, so I'll bow out of the discussion now. Sterling
From: William Edward Woody <woody@alumni.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Fri, 15 May 1998 14:18:53 -0700 Organization: In Phase Consulting Message-ID: <355CB13D.82E8A916@alumni.caltech.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <355A74AC.107@stanford.edu> <355A8C8B.3C7D9DAB@alumni.caltech.edu> <355B0E00.631A@stanford.edu> <355C09EC.47EAEA33@alumni.caltech.edu> <355C8DBA.6570@stanford.edu> <355C912D.876C2C2F@alumni.caltech.edu> <355C9B47.7C5A@stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sterling wrote: > > (Patching > > Quickdraw? What [were] you working on?) > > Application profiling for MacBench 3, 4 and a new version that got > canned when MacUser and Macworld joined. If you want to know how another > application is using the toolbox, patching traps is the only mechanism. > (Although I would be happy to be proved wrong on this point.) Ohhhh! That explains a lot! AFAIK, as far as debugging and profiling applications are concerned, the existing Macintosh API sort of sucks rocks. But I don't think this has anything to do with the memory manager--though the lack of a reasonable mechanism, such as provided by Windows or most flavors of UNIX means that the non-reentrancy of the memory manager will get in the way. > > And what could you possibly > > be doing on the Macintosh with respect to drawing that is more > > easily done on Windows? I honestly want to know because I have > > never encountered such a creature before. > > The windows guys' profiling was much easier--or I should say, the > mechanism they found for peeking at applications' behavior was much > easier to implement, mostly because they found a way to do it as a full > operating system citizen, unlike my lowly trap patcher. So technically, > the drawing wasn't easier, but studying system behavior was much easier. > I'm with you there, give me the mac to draw on anyday. And I will agree that the existing mechanism under Windows makes this sort of profiling a hell of a lot easier. (Not that the mechanism is all that great--I've played with it in the past. But that there is an API at all is a lot easier than patching OS traps.) I hope Apple is listening--and provides debugging support services as a fundamental part of the Carbon design, instead of tossing it in as an afterthought. > Anyway, I've spent way too much time on this over the past few days, so > I'll bow out of the discussion now. Thanks and take care, - Bill -- William Edward Woody | In Phase Consulting woody@alumni.caltech.edu | Macintosh & MS Windows Development http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/
From: schuerig@acm.org (Michael Schuerig) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Fri, 15 May 1998 23:38:48 +0200 Organization: Completely Disorganized Message-ID: <1d937nh.hbzh011dr3q8aN@rhrz-isdn3-p13.rhrz.uni-bonn.de> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559EA34.B89B62D7@alumni.caltech.edu> <3559F204.3352@stanford.edu> <Pine.NXT.3.96.980513175733.12677B-100000@pathos> <355A8F4B.17463CF@alumni.caltech.edu> <355B15A3.D2D@stanford.edu> <355C0B14.98C50B36@alumni.caltech.edu> <355C80C2.39A@stanford.edu> Mail-Copies-To: never Sterling <Sterling_Augustine@stanford.edu> wrote: > William Edward Woody wrote: > > > In context, this comment refered to the programmer's interface to > > the Memory Manager, not the underlying source code. > > But the discussion was about what affect carbon would have on handles. > So saying that it is all moot because the API doesn't change misses the > point. > > The fact that the API doesn't change doesn't prevent numerous possible > simplifications Apple could make to its runtime environment. > > My original suggestion was that Apple could say, for example "Under > MacOS X, handles do not move." That would be incredibly simplifying. Not really, if you want to keep backward compatibility you can't take advantage of it. Michael -- Michael Schuerig mailto:schuerig@acm.org http://www.uni-bonn.de/~uzs90z/
From: "Edwin E. Thorne" <edwin.thorne@rauland.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Yellowbox programming, and ragosta Date: 15 May 1998 22:50:03 GMT Organization: WorldWide Access - Midwestern Internet Services - www.wwa.com Message-ID: <01bd8053$3839e0c0$c07498cf@edwint.rauland.com> References: <6j2qnb$r2p$2@newsfep1.sprintmail.com> <6j4eta$rbo$1@crib.bevc.blacksburg.va.us> <35563AB3.B8A9AAC@nstar.net> Michael J. Peck <mjpeck@nstar.net> wrote in article <35563AB3.B8A9AAC@nstar.net>... > Nathan Urban wrote: > > > Others have pointed out your problem. What I don't understand is, > > why are you singling out Joe?? If you've been paying attention at all, > > you'd have noticed that the rest of us _also_ say that it's as simple as > > clicking a box and recompiling, as long as you stay within the Yellow Box. > > I think the point was that while "the rest of [you]" know what you're > talking about, and can answer the question... > > ...draw your own conclusions. It's a matter of "legitimacy of claims". > > MJP Hmmm.......So Joe doesn't know what he's talking about, even though he's saying the same things as the people who you do credit as knowing what they're talking about? Edwin
From: *johnnyc*@or.psychology.dal.ca (John Christie) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Fri, 15 May 1998 22:37:20 -0300 Organization: ISINet, Nova Scotia Message-ID: <*johnnyc*-1505982237210001@jchristie.halifaxcable.dal.ca> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <kalash-1405981828460001@boris.starnine.com> <355C0CDF.6D2314E@alumni.caltech.edu> In article <355C0CDF.6D2314E@alumni.caltech.edu>, William Edward Woody <woody@alumni.caltech.edu> wrote: > Except... > > Memory compaction only occurs in low memory situations, when space is > tight. And other situations which cause memory movement are situations > (like handle resizing) where movement would happen even had we been > using realloc() instead. And major compaction (caused when things are > so fragmented it requires a major effort to fit the handle into the > heap) only occurs when things are so tight that the allocation would > have otherwise failed. Memory compaction occurs in situation like you describe, but that is not the only time. It also occurs when non kosher memory is assigned. Whenever a Pointer is chucked out instead of a Handle then the memory must be moved to make certain the Pointer goes in the lowest (highest?) point. -- You aren't free if you CAN choose - only if you DO choose. All you are is the decisions you make. Remove "*" and ohnny (i.e. jc@) to reply via email
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Designing and Implementing (was: future of mac programming) Date: 16 May 1998 06:27:30 GMT Organization: Idiom Communications Message-ID: <6jjbki$n53$1@news.idiom.com> References: <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559EA34.B89B62D7@alumni.caltech.edu> <3559F204.3352@stanford.edu> <Pine.NXT.3.96.980513175733.12677B-100000@pathos> <kalash-1405981820030001@boris.starnine.com> <Pine.NXT.3.96.980515095759.20230E-100000@pathos> <355c7db9.969266058@news.motu.com> <Pine.NXT.3.96.980515150112.20230T-100000@pathos> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: bbum@codefab.com Bill Bumgarner may or may not have said: [some very good thoughts on the importance of designing before implementing snipped] I would also point out, that there is a tremendous cost involved if the code deviates from the spec without the spec getting revised. I recently worked on an application where the original architect had specified that any time the application changed the current selection, it would post a notification of the change. That is, when it changed "what was selected" as opposed to editing the selected objects. This discipline was not followed, and each of the implementers who needed to write pieces of the app that would bind to the current selection, such as inspector panels and logging routines, had to write incredibly convoluted code that chased through the UI objects and the global window list to try to figure out what the inspectors should be inspecting. I would take a guess that this added something on the order of a hundred hours to each of these components. If the design document had been followed, the app could have saved many pages of duplicated code, and an enormous amount of debugging time. -jcr
From: mpaque@wco.com (Mike Paquette) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer,comp.sys.mac.programmer Subject: NSHosting and MacOS X Followup-To: comp.sys.next.advocacy Date: Sat, 16 May 1998 01:16:11 -0700 Organization: Electronics Service Unit No. 16 Message-ID: <1d93iik.1l170dz8etsaxN@carina47.wco.com> There has been some concern expressed over MacOS X and the loss of the NXHost capability. I'd like to clear a few things up. NX/NS Host Operation NXHost works by redirecting the PostScript binary object stream to a PostScript interpreter running in a Window Server on a different machine than the current application. A PostScript interpreter is required to process this stream. MacOS X will not require a PostScript interpreter. One will not be present in the base system. Therefore, one cannot reasonably process the existing NSHost protocol to a MacOS X system. Remote Display and operation Remote display and operation of Mac systems is an important capability, which in the past has been addressed by third party applications. I realize the importance of this capability to the developer, educational, and enterprise communities. I am very interested in providing this capability. Please note, however, that there have to be priorities. Apple must deliver on the promised core functionality in MacOS X in a timely manner. Additional features such as remote display will be given all due consideration, but cannot be promised for the initial release of the software. Room for expansion in future directions such as this will be considered during the design and implementation process. What You Can Do If you are a MacOS developer, get and run the Carbon Dater from devworld.apple.com. Check out your code and get it Carbon-ready. If you've been patching traps to do remote display, please write up a short blurb on what you've been patching, and what functionality you are looking for. The engineering team will try to provide the needed functionality for you. Send this information to carbon@apple.com. -- Mike Paquette MacOS X Graphics Subsystems
From: Terry Greeniaus <tgree@Phys.UAlberta.CA> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Sat, 16 May 1998 02:15:25 -0600 Organization: University of Alberta, Edmonton, Canada Message-ID: <Pine.ULT.3.91.980516021341.19469C-100000@stoney> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <kalash-1405981828460001@boris.starnine.com> <355C0CDF.6D2314E@alumni.caltech.edu> <kalash-1605980037300001@kalash-home.starnine.com> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII In-Reply-To: <kalash-1605980037300001@kalash-home.starnine.com> On Sat, 16 May 1998, Joe Kalash wrote: > > So the tradeoff is mostly the overhead of the (roughly) 4 bytes per > > handle overhead to maintain the master pointers, and of course the > > brain space needed for programmers to manage their handles correctly. > > Ummm, no. There is also the CPU overhead of maintaining and walking the > information (which is the real issue). Try writting a little program that > allocates and releases a "bunch" of blocks of various sizes. Write it once > with NewPtr/DisposPtr. Then write it with malloc/free. Time both, as you > increase the number of blocks. You will quickly find that malloc/free win. > This is even what Apple will tell you (at least they have told us this > very explictly). Of course, neither NewPtr/DisposPtr or malloc/free have anything to do with handles... Granted however, malloc/free/new/delete is the way to go in modern OS's which actually supprort virtual/mapped memory. However Handles do have their places in the MacOS (otherwise they never would have been invented so many years ago by those Apple programmers...) TG
From: kalash@starnine.com (Joe Kalash) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Sat, 16 May 1998 00:27:52 +0100 Organization: StarNine Technologines, Inc. Message-ID: <kalash-1605980027520001@kalash-home.starnine.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559EA34.B89B62D7@alumni.caltech.edu> <3559F204.3352@stanford.edu> <Pine.NXT.3.96.980513175733.12677B-100000@pathos> <kalash-1405981820030001@boris.starnine.com> <Pine.NXT.3.96.980515095759.20230E-100000@pathos> In article <Pine.NXT.3.96.980515095759.20230E-100000@pathos>, Bill Bumgarner <bbum@codefab.com> wrote: > Problem is that it is extremely difficult to get threading right. And > once it IS done "right", maintaing/debugging/testing a body of code that > relies heavily on threads can be maddening. > > Managing the complexity of an MT application is MUCH MORE difficult than > non-MT. [etc.] Geez, you make it sound equivalent to performing self heart surgery :-). I will freely grant you that programming with threads is more difficult then programming without them. But it can be dealt with. There are lots of heavily threaded systems that can have literally hundreds of programmers busily working on them (database engines like Oracle, Sybase, etc. pop quickly to mind). Kernel work is not threaded in the traditional sense, but it amounts to the same thing, and places like Sun/Apple/Microsoft can have immense oodles of programmers putting there oars into the programming waters (without all that many disasters). It is as much a mindset as anything else. When it comes down to it, threads are just one of the things we are paid to know and use (we is [sic] professionals, after all). -- Joe Kalash StarNine Technologies, Inc. kalash@starnine.com
From: kalash@starnine.com (Joe Kalash) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Sat, 16 May 1998 00:37:30 +0100 Organization: StarNine Technologines, Inc. Message-ID: <kalash-1605980037300001@kalash-home.starnine.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <kalash-1405981828460001@boris.starnine.com> <355C0CDF.6D2314E@alumni.caltech.edu> In article <355C0CDF.6D2314E@alumni.caltech.edu>, William Edward Woody <woody@alumni.caltech.edu> wrote: > So the tradeoff is mostly the overhead of the (roughly) 4 bytes per > handle overhead to maintain the master pointers, and of course the > brain space needed for programmers to manage their handles correctly. Ummm, no. There is also the CPU overhead of maintaining and walking the information (which is the real issue). Try writting a little program that allocates and releases a "bunch" of blocks of various sizes. Write it once with NewPtr/DisposPtr. Then write it with malloc/free. Time both, as you increase the number of blocks. You will quickly find that malloc/free win. This is even what Apple will tell you (at least they have told us this very explictly). > And besides--if you don't think it's a valid tradeoff, don't use 'em. I suspect it will become irrelevant. While the handle operations will still exist in "X", I will bet a really shiny nickle that under the Carbon world they are just simple wrappers around malloc/free. -- Joe Kalash StarNine Technologies, Inc. kalash@starnine.com
From: William Edward Woody <woody@alumni.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Sat, 16 May 1998 04:15:03 -0700 Organization: In Phase Consulting Message-ID: <355D7537.45A26CFA@alumni.caltech.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <kalash-1405981828460001@boris.starnine.com> <355C0CDF.6D2314E@alumni.caltech.edu> <kalash-1605980037300001@kalash-home.starnine.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Joe Kalash wrote: > William Edward Woody <woody@alumni.caltech.edu> wrote: > > So the tradeoff is mostly the overhead of the (roughly) 4 bytes per > > handle overhead to maintain the master pointers, and of course the > > brain space needed for programmers to manage their handles correctly. > > Ummm, no. There is also the CPU overhead of maintaining and walking the > information (which is the real issue). Try writting a little program that > allocates and releases a "bunch" of blocks of various sizes. Write it once > with NewPtr/DisposPtr. Then write it with malloc/free. Time both, as you > increase the number of blocks. You will quickly find that malloc/free win. > This is even what Apple will tell you (at least they have told us this > very explictly). I am aware of this. What I am not aware of is what this has to do with NewHandle()/ DisposeHandle(). Especially as Apple has indicated that NewHandle()/ DisposeHandle() allocate memory by looking for the next free block, without making any attempts to compact memory beforehand. This means that unless memory becomes tight, NewPtr()/DisposePtr() won't move handles around. (That's why the well-groomed Macintosh programmer uses MoveHHi and other memory manager routines.) More importantly, this means that unless memory is tight enough to move handles around, NewPtr()/DisposePtr() don't do anything that malloc()/ free() don't do. (And if memory is that tight, malloc() would fail anyways.) > > And besides--if you don't think it's a valid tradeoff, don't use 'em. > > I suspect it will become irrelevant. While the handle operations will > still exist in "X", I will bet a really shiny nickle that under the Carbon > world they are just simple wrappers around malloc/free. *shrug* See, that's the issue I'm having here. Handles are a powerful tool for preventing memory fragmentation--simply getting rid of 'em because some programmers can't deal with 'em seems rather silly to me. - Bill -- William Edward Woody | In Phase Consulting woody@alumni.caltech.edu | Macintosh & MS Windows Development http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 16 May 1998 12:18:28 GMT Organization: Idiom Communications Message-ID: <6jk06k$n53$13@news.idiom.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <kalash-1405981828460001@boris.starnine.com> <355C0CDF.6D2314E@alumni.caltech.edu> <kalash-1605980037300001@kalash-home.starnine.com> <355D7537.45A26CFA@alumni.caltech.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: woody@alumni.caltech.edu William Edward Woody may or may not have said: --> See, that's the issue I'm having here. Handles are a powerful tool -> for preventing memory fragmentation--simply getting rid of 'em -> because some programmers can't deal with 'em seems rather silly to me. How about getting rid of them because they're not necessary when your OS has a modern, demand-paged virtual memory system? They were a clever hack for a 128K mac with a MC68000 CPU no MMU. MicroSquish followed suit when they were making Windoze, because they didn't want to try to make a full VM system, either. -jcr
From: *johnnyc*@or.psychology.dal.ca (John Christie) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Sat, 16 May 1998 10:44:17 -0300 Organization: ISINet, Nova Scotia Message-ID: <*johnnyc*-1605981044180001@jchristie.halifaxcable.dal.ca> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <kalash-1405981828460001@boris.starnine.com> <355C0CDF.6D2314E@alumni.caltech.edu> <kalash-1605980037300001@kalash-home.starnine.com> <355D7537.45A26CFA@alumni.caltech.edu> In article <355D7537.45A26CFA@alumni.caltech.edu>, William Edward Woody <woody@alumni.caltech.edu> wrote: > What I am not aware of is what this has to do with NewHandle()/ > DisposeHandle(). Especially as Apple has indicated that NewHandle()/ > DisposeHandle() allocate memory by looking for the next free block, > without making any attempts to compact memory beforehand. This means > that unless memory becomes tight, NewPtr()/DisposePtr() won't move > handles around. (That's why the well-groomed Macintosh programmer > uses MoveHHi and other memory manager routines.) More importantly, > this means that unless memory is tight enough to move handles > around, NewPtr()/DisposePtr() don't do anything that malloc()/ > free() don't do. (And if memory is that tight, malloc() would > fail anyways.) That isn't what I get at all. Assign a big Handle into memory. Assign several thousand pointers. Time the assignment of each pointer. You will see a linear increase in the time. Every time a new pointer is assigned memory moves (Disposed is another matter). Furthermore, it has to check all of the current memory to make certain that it is accessing a point that reduces fragmentation. OTOH, Handles get chucked out at the same all the time. Therefore, Handles do provide a cost to those who don't want to use them. The compaction of the heap on each New would not occur if it were not for Handles. simply "choosing" not to use Handles incurs a cost in performance so it isn't as simple as you say. Malloc and free are special circumstances. Most compilers typically reserve a block of memory with a Pointer the first time malloc is called and then assign within it until it is full (and so on). This way the linear decrease in performance that occurs with each call of NewPtr is avoided. BTW, malloc in most compilers calls NewPtr to assign that big block. > See, that's the issue I'm having here. Handles are a powerful tool > for preventing memory fragmentation--simply getting rid of 'em > because some programmers can't deal with 'em seems rather silly to me. Keeping them if most programmers don't want to use them is silly because the few who do want to could write their own implementation if they wanted to. However, Handles system wide interferes with the use of pointers. I guess we should ask around and see what most programmers want to use. -- You aren't free if you CAN choose - only if you DO choose. All you are is the decisions you make. Remove "*" and ohnny (i.e. jc@) to reply via email
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.next.programmer Subject: Re: [Q] What are the implications of MacOS X? Date: 16 May 1998 14:47:52 GMT Organization: P & L Systems Message-ID: <6jk8uo$1cc$52@ironhorse.plsys.co.uk> References: <6jherv$sqg$1@nnrp1.dejanews.com> <johns-1505980828570001@10.0.0.253> <6jhrog$jum$1@news.seicom.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: frank@this.NO_SPAM.net In <6jhrog$jum$1@news.seicom.net> Frank M. Siegert wrote: > Except for the dropped DPS functionality - replaceing DPS by Quickdraw seems > to me like removing the wheels from my car and putting rollerskates in place. > Umm, they haven't replaced it with QD: the new imaging model is PDF. This means that we basically have much the same model (PSmoveto(), PSlineto() etc all work), "just" no pswraps. That may be a big deal to some people, but hopefully they're in a minority, and hopefully things shouldn't be so bad as pswraps are usually used to enhance performance (reducing round trips to the windowserver etc), so if the server is faster... Best wishes, mmalc.
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.next.programmer Subject: Re: [Q] What are the implications of MacOS X? Date: 16 May 1998 14:51:13 GMT Organization: P & L Systems Message-ID: <6jk951$1cc$53@ironhorse.plsys.co.uk> References: <6jherv$sqg$1@nnrp1.dejanews.com> <johns-1505980828570001@10.0.0.253> <6jhrog$jum$1@news.seicom.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: frank@this.NO_SPAM.net In <6jhrog$jum$1@news.seicom.net> Frank M. Siegert wrote: > Except for the dropped DPS functionality - replaceing DPS by Quickdraw seems > to me like removing the wheels from my car and putting rollerskates in place. > To follow up on my previous reply (pointing out the new model is PDF)... Elsewhere Izumi Ohzawa pointed out that another thing that may be lost is preview of ESP files, and the possibility of using EPS on the pasteboard. This may be more of a blow. Frank, you're one of the PS/PDF gurus here -- Izumi asked is there any way of producing "encapsulated PDF"? Best wishes, mmalc.
From: William Edward Woody <woody@alumni.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Sat, 16 May 1998 10:24:50 -0700 Organization: In Phase Consulting Message-ID: <355DCBE2.1EED860F@alumni.caltech.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <kalash-1405981828460001@boris.starnine.com> <355C0CDF.6D2314E@alumni.caltech.edu> <kalash-1605980037300001@kalash-home.starnine.com> <355D7537.45A26CFA@alumni.caltech.edu> <*johnnyc*-1605981044180001@jchristie.halifaxcable.dal.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit John Christie wrote: > I guess we should ask around and see what most programmers want to use. ... And if the answer comes back that they want to use Windows NT? ... Ah, well; this is getting really pretty silly. After all, the point of Carbon is backwards compatability; we just can't ask programmers what they want to use 'cause the Memory Manager API has to preserve source level compatability. - Bill -- William Edward Woody | In Phase Consulting woody@alumni.caltech.edu | Macintosh & MS Windows Development http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/
From: William Edward Woody <woody@alumni.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Sat, 16 May 1998 10:39:15 -0700 Organization: In Phase Consulting Message-ID: <355DCF43.DBF0BD67@alumni.caltech.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <kalash-1405981828460001@boris.starnine.com> <355C0CDF.6D2314E@alumni.caltech.edu> <kalash-1605980037300001@kalash-home.starnine.com> <355D7537.45A26CFA@alumni.caltech.edu> <6jk06k$n53$13@news.idiom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit John C. Randolph wrote: > How about getting rid of them because they're not necessary when your OS > has a modern, demand-paged virtual memory system? > > They were a clever hack for a 128K mac with a MC68000 CPU no MMU. > MicroSquish followed suit when they were making Windoze, because they didn't > want to try to make a full VM system, either. The fundamental argument that has been made by those who don't like Handles are threefold: 1. Handles are no longer required because we are no longer dealing with a "primitive" environment, but will now have a "modern, demand-paged virtual memory system" which is capable of doing just about everything we could imagine behind the programmer's back, including cleaning up after the programmer's mess for him/her. 2. Handles make the system more buggy because they introduce a level of compexity into the programming model which most programmers apparently are incapable of dealing with. (Why Handles are "hard" yet multi-threaded applications are "easy" is a mystery to me...) (And of course those who have made this observations aren't saying that they are personally incapable of dealing with handles, only the hypothetical "other guy" is.) 3. Handles are slow, inefficient, and by their very existance cause the whole system to grind to a screaching halt. I don't buy (1). First, even in a potentially infinite (4 gig) address space, you are still limited by physical RAM and the amount of hard disk space the user is willing to set aside as swap, minus the amount of space run by other applications. You aren't going to win friends and influence enemies if your application's paged set size gets to be a large percentage of RAM + disk space. Further, as your page set grows, so does the chances of your application "page trashing." And if you haven't experienced a program page thrashing on Windows NT or on a VAX, you're missing a vital bit of personal experience. Page thrashing due to a large, inorganized page set is not a pretty site. So we can't simply use the "modern virtual memory system will do everything but wipe our own a**es" argument, as that sort of programmer laziness is not a Good Idea. I don't buy (2), either. Partially because if you don't want to use 'em, or can't deal with 'em, don't use 'em. Many software applications being written for the Macintosh *right* *now* don't use Handles (except for those UI items, like Controls, which force you to). But in some places Handles are Very Useful. Personally, I have a 'XGMemoryFile' and an 'XGDynArray' class in C++ which uses Handles on the Macintosh to keep things together--on Windows and X they're implemented using malloc() and realloc(). Honestly I find the Handle solution a little more eligant. And I certainly don't buy (3), as it's clear to me that if Apple were to support handles on a "modern virtual memory system", they'd know enough to change the algorithms used to manage those handles. I'm also a little disturbed to see the number of posts going by which suggest that the Macintosh memory manager is somehow "primitive" because it was made to work originally on the 128K Macintosh. While I'm not arguing that it is *not* primitive (and trust me: I'm not one of those guys who bleed in 6 colors), I'm not certain that this blanket assertion that "Handles are primitive" is warranted. - Bill -- William Edward Woody | In Phase Consulting woody@alumni.caltech.edu | Macintosh & MS Windows Development http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/
From: "Todd Heberlein" <todd@NetSQ.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: 16 May 1998 18:00:58 GMT Organization: Net Squared, Inc. Message-ID: <01bd80f4$85082140$04387880@test1> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <6j7sdr$4gn$1@newsfep4.sprintmail.com> <6j7tij$vb7$1@crib.bevc.blacksburg.va.us> <stevenj-ya02408000R1105982039000001@news.mit.edu> <andyba-ya02408000R1205980916210001@news> <355873F0.6B86BC5C@milestonerdl.com> <andyba-ya02408000R1405981025210001@news> Andy Bates <andyba@corp.webtv.net> wrote > > Is there a TECHNICAL reason, or just political? > > I would say technical; otherwise, they'd just have ported ... There is also a time-to-market issue. Perhaps Apple just doesn't have enough resources to deliver a high-quality, well tested version to the Intel market in a timely fashion. Also, Carbon was mainly created for porting lots of old code to the new platform. This old code probably has a lot of byte-order dependencies in it, and cost-benefit analysis might prevent most developers from porting the Carbon/PowerPC code to Carbon/Intel. Also, it is questionable how much money Apple could make with an Intel version. Very few, if any, major PC manufacturers offer a non-Microsoft OS on their computer system. I am not even sure you can order one w/o a Microsoft OS. That is, even if you intend to install Linux or Rhapsody on your new Dell, you will still have to pay Microsoft for the OS shipped with your computer! (per processor licensing still lives!) Also, "Intel" usually means the x86 family of chips. Apple may still ship for Merced - which will be out around the time MacOS X is released. Just have to wait and see, I guess. Todd
From: Matthew Cromer <matthew_cromer@iname.com> Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer,comp.sys.mac.programmer Subject: Re: NSHosting and MacOS X Date: 16 May 1998 19:27:42 GMT Organization: NETworthy Inc. Distribution: world Message-ID: <6jkpbe$m2v$1@camel0.mindspring.com> References: <1d93iik.1l170dz8etsaxN@carina47.wco.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <1d93iik.1l170dz8etsaxN@carina47.wco.com> Mike Paquette, mpaque@wco.com writes: >Remote Display and operation > >Remote display and operation of Mac systems is an important capability, >which in the past has been addressed by third party applications. I >realize the importance of this capability to the developer, educational, >and enterprise communities. I am very interested in providing this >capability. The problem with PCAnywhere and Timbuktu is that they allow only one user on a particular computer to use the GUI. NSHosting and X allow multiple GUI logins to a single computer. THis is the screaming advantage of such a protocol. It will be very sad if Apple takes such an important part of the Unix heritage away from MacOS X. Matthew Cromer
From: "Izumi Ohzawa" <izumi@pinoko.berkeley.edu> Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: NSHosting and MacOS X Date: 16 May 1998 20:35:16 GMT Organization: University of California, Berkeley Message-ID: <01bd80c6$be64d100$a4e82080@mizuki.HIP.Berkeley.EDU> References: <1d93iik.1l170dz8etsaxN@carina47.wco.com> Mike Paquette <mpaque@wco.com> wrote in article <1d93iik.1l170dz8etsaxN@carina47.wco.com>... > There has been some concern expressed over MacOS X and the loss of the > NXHost capability. I'd like to clear a few things up. > > NX/NS Host Operation > > NXHost works by redirecting the PostScript binary object stream to a > PostScript interpreter running in a Window Server on a different machine > than the current application. A PostScript interpreter is required to > process this stream. > > MacOS X will not require a PostScript interpreter. One will not be > present in the base system. Therefore, one cannot reasonably process > the existing NSHost protocol to a MacOS X system. Is it really the loss of DPS that makes NSHosting difficult? After all, in principle, there can be Display PDF engine doing similar work as DPS and work on PDF streams from clients. It seems to me, that it's not really the lack of DPS per se, but the new boundary between WindowServer and clients. Previously, it was at the PS stream. Now, the WindowServer hands clients (apps) a shared memory for an area of display, and clients directly writes into it. I can see how this makes it difficult to do remote displays. It would probably have to be based on streaming QuickTime which transports backing store (and sound) across network to the actual remote display (and sound port) on another machine. Still, I haven't heard any explanation as to why they decided to break the client- server boundary at the shared-memory. If multi-threading of DPDF engine is important, well, you can run multiple threads of Display PDF on the server side each listening to PDF streams from clients, either local or remote. Perhaps, it is integration of QuickTime into NSView that drove this decision. I would appreciate more enlightenment, Mike? Thanks.
From: Kristofer Jon Magnusson <kris@xmission.xmission.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: 16 May 1998 14:52:43 -0600 Organization: XMission Internet (801 539 0852) Message-ID: <6jkuar$39a$1@xmission.xmission.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <6j7sdr$4gn$1@newsfep4.sprintmail.com> <6j7tij$vb7$1@crib.bevc.blacksburg.va.us> <stevenj-ya02408000R1105982039000001@news.mit.edu> <andyba-ya02408000R1205980916210001@news> <355873F0.6B86BC5C@milestonerdl.com> <andyba-ya02408000R1405981025210001@news> <01bd80f4$85082140$04387880@test1> In comp.sys.next.programmer Todd Heberlein <todd@NetSQ.com> wrote: : Also, it is questionable how much money Apple could make with an Intel : version. Very few, if any, major PC manufacturers offer a : non-Microsoft OS on their computer system. I am not even sure you can : order one w/o a Microsoft OS. That is, even if you intend to install : Linux or Rhapsody on your new Dell, you will still have to pay : Microsoft for the OS shipped with your computer! (per processor : licensing still lives!) This is true, but not entirely accurate. Lots of people purchase non-MS operating systems for Intel, despite the per-system licensing. : Just have to wait and see, I guess. Go ahead. Without Intel, Apple is just wasting my time. So I'm history. ............kris -- Kristofer Jon Magnusson <kris@xmission.com> "NASA and its partners have identified vacuum as a major attribute of space."
From: see-below@not-my-address.com (Matthew Vaughan) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Sat, 16 May 1998 16:20:33 -0700 Message-ID: <see-below-1605981620340001@209.24.241.16> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <6j7sdr$4gn$1@newsfep4.sprintmail.com> <6j7tij$vb7$1@crib.bevc.blacksburg.va.us> <stevenj-ya02408000R1105982039000001@news.mit.edu> <andyba-ya02408000R1205980916210001@news> <355873F0.6B86BC5C@milestonerdl.com> <andyba-ya02408000R1405981025210001@news> <01bd80f4$85082140$04387880@test1> <6jkuar$39a$1@xmission.xmission.com> In article <6jkuar$39a$1@xmission.xmission.com>, Kristofer Jon Magnusson <kris@xmission.xmission.com> wrote: > In comp.sys.next.programmer Todd Heberlein <todd@NetSQ.com> wrote: > > : Also, it is questionable how much money Apple could make with an Intel > : version. Very few, if any, major PC manufacturers offer a > : non-Microsoft OS on their computer system. I am not even sure you can > : order one w/o a Microsoft OS. That is, even if you intend to install > : Linux or Rhapsody on your new Dell, you will still have to pay > : Microsoft for the OS shipped with your computer! (per processor > : licensing still lives!) > > This is true, but not entirely accurate. Lots of people purchase non-MS > operating systems for Intel, despite the per-system licensing. > > : Just have to wait and see, I guess. > > Go ahead. Without Intel, Apple is just wasting my time. So I'm history. Apple will still ship for Intel. Just not a consumer version of the OS (ie, with backward compatibility for Macintosh applications). .................................................... MATTHEW VAUGHAN matthewv at best dot com (damn spammers...) http://www.best.com/~matthewv/ ....................................................
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.next.programmer Subject: Re: [Q] What are the implications of MacOS X? Date: 16 May 1998 16:17:03 -0700 Organization: Primenet Services for the Internet Message-ID: <B1836C75-7514F@206.165.43.149> References: <6jk951$1cc$53@ironhorse.plsys.co.uk> MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit mmalcolm crawford <malcolm@plsys.co.uk> said: > >Elsewhere Izumi Ohzawa pointed out that another thing that may be lost is >preview of ESP files, and the possibility of using EPS on the pasteboard. >This may be more of a blow. Frank, you're one of the PS/PDF gurus here -- >Izumi asked is there any way of producing "encapsulated PDF"? And you guys wonder why I scream about GX when we're dealing with Apple... Here's another question for you guys: Why can't a GX-like picture shape be used instead of PDF to transfer images over the clipboard? A list of shapes, including other pictures is far better than the pdf format, IMHO, since it allows for easy editing of any component shape/pciture, and is easily converted to PDF with no loss of fidelity. ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: 16 May 1998 16:34:02 -0700 Organization: Primenet Services for the Internet Message-ID: <B1837072-84146@206.165.43.149> References: <see-below-1605981620340001@209.24.241.16> nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Matthew Vaughan <see-below@not-my-address.com> said: >> >> Go ahead. Without Intel, Apple is just wasting my time. So I'm history. > > >Apple will still ship for Intel. Just not a consumer version of the OS >(ie, with backward compatibility for Macintosh applications). I still don't see how MacOS X is going to be the consumer version of MacOS. MacOS X is Rhapsody + Carbon, which is likely larger than Rhapsody. How is that "consumer-oriented?" ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
Date: Sat, 16 May 1998 20:05:15 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Message-ID: <joe.ragosta-1605982005160001@elk32.dol.net> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> In article <B1837072-84146@206.165.43.149>, "Lawson English" <english@primenet.com> wrote: > Matthew Vaughan <see-below@not-my-address.com> said: > > >> > >> Go ahead. Without Intel, Apple is just wasting my time. So I'm history. > > > > > >Apple will still ship for Intel. Just not a consumer version of the OS > >(ie, with backward compatibility for Macintosh applications). > > > > I still don't see how MacOS X is going to be the consumer version of MacOS. > > MacOS X is Rhapsody + Carbon, which is likely larger than Rhapsody. > > How is that "consumer-oriented?" Maybe. But if they've left out the Unix from MacOS X, it could be smaller. I'd also expect MacOS X to have fewer tools. -- Regards, Joe Ragosta See the Complete Macintosh Advocacy Page http://www.dol.net/~Ragosta/complmac.htm
From: johns@sonic.net (John Selhorst) Newsgroups: comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Sat, 16 May 1998 17:11:58 +0100 Organization: John Selhorst Message-ID: <johns-1605981711580001@10.0.0.253> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> In article <B1837072-84146@206.165.43.149>, "Lawson English" <english@primenet.com> wrote: >I still don't see how MacOS X is going to be the consumer version of MacOS. > >MacOS X is Rhapsody + Carbon, which is likely larger than Rhapsody. > >How is that "consumer-oriented?" It still will run the blue box, so consumers can have their cake and eat it too. Johnny
From: macghod@concentric.net Newsgroups: comp.sys.next.programmer Subject: Newby question Date: 17 May 1998 01:04:33 GMT Message-ID: <6jld31$l6u$1@newsfep4.sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I am having problems with the travel advisor tutorial in the openstep developer tutorial. I am at #3, resuse the converter class (p69) it says to select the converter class from the currency converter nib and then copy, then to select NSObject in travelAdvisor.nib, then select paste. I do so and I get Unexpected Error Error: *** -[IBClassDescriptor initWithDictionary:]; selector not recognized What am I doing wrong? -- Running on Openstep 4.2, the best os in the world, and about to get better! Mac/openstep qa work desired, email me for resume NeXTMail and MIME OK!
From: Ryan Kennedy <rck@thegrid.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Sat, 16 May 1998 18:37:43 -0700 Organization: Call America Internet Services +1 (800) 563-3271 Message-ID: <355E3F66.86DD9F0@thegrid.net> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> <joe.ragosta-1605982005160001@elk32.dol.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Joe Ragosta wrote: > But if they've left out the Unix from MacOS X, it could be smaller. Actually, if you're talking about size as in disk space then Unix will be smaller. Unix was made to be small and simple, most kernels can actually be fit on a 1.44 MB disk. Ryan
From: glhansen@copper.ucs.indiana.edu (Gregory Loren Hansen) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: 17 May 1998 03:29:48 GMT Organization: Indiana University, Bloomington Message-ID: <6jlljc$m6$1@flotsam.uits.indiana.edu> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> In article <B1837072-84146@206.165.43.149>, Lawson English <english@primenet.com> wrote: >I still don't see how MacOS X is going to be the consumer version of MacOS. > >MacOS X is Rhapsody + Carbon, which is likely larger than Rhapsody. > >How is that "consumer-oriented?" By the time MacOS X ships, 6 gig hard drives and 64 megs RAM will be the standard minimum configuration. Okay, so I just pulled those numbers out of my butt. But do you remember when 4 megs RAM was the standard in all the systems advertised in the Sunday newspaper? 16 megs RAM was standard a little while ago, but now it's moving toward 32 megs. Disk drives have followed a similar trend, with 2 gigs and 4 gigs being commmon and some bundles already come with 6 gig drives. When I got my Power100 about three years ago, it came with an 850 megger and that was common, give or take a few hundres megs. My estimates may be conservative. -- Stay alert! Trust no one! Keep your laser handy! The Computer is your Friend!
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <1562894772821@digifix.com> Date: 17 May 1998 03:49:39 GMT Organization: Digital Fix Development Message-ID: <10466895377625@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1993. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: veloso@mci2000.com (Manuel Veloso) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <veloso-1605982359000001@nemo.ibex.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559EA34.B89B62D7@alumni.caltech.edu> <3559F204.3352@stanford.edu> <Pine.NXT.3.96.980513175733.12677B-100000@pathos> <kalash-1405981820030001@boris.starnine.com> <Pine.NXT.3.96.980515095759.20230E-100000@pathos> <kalash-1605980027520001@kalash-home.starnine.com> Date: Sun, 17 May 1998 04:56:32 GMT NNTP-Posting-Date: Sun, 17 May 1998 00:56:32 EDT In article <kalash-1605980027520001@kalash-home.starnine.com>, kalash@starnine.com (Joe Kalash) wrote: > In article <Pine.NXT.3.96.980515095759.20230E-100000@pathos>, Bill > Bumgarner <bbum@codefab.com> wrote: > > Problem is that it is extremely difficult to get threading right. And > > once it IS done "right", maintaing/debugging/testing a body of code that > > relies heavily on threads can be maddening. > > > > Managing the complexity of an MT application is MUCH MORE difficult than > > non-MT. > > Geez, you make it sound equivalent to performing self heart surgery :-). > > I will freely grant you that programming with threads is more difficult > then programming without them. But it can be dealt with. Actually, I find that threads make things simpler, since you only have intersection points to worry about. While these points can be a bit hairy, there aren't (or shouldn't) be very many of them. Plus, it's a lot easier to use threads (and understand/organize the code) than to do some kind of weird polling mechanism (bleah) where whole bunches of state are exposed for no particularly good reason. Of course, it also depends a whole lot on how good your tools are. On *nix, the tools pretty much suck, so threading (and debugging) are much harder. Last I heard, gdb (maybe the world's worst debugger) didn't support hardware threads on Solaris, for example. It's kind of hard to debug threaded code w/o a debugger, since there's only so much data you can get from fprintf (which then needs to be wrapped to be thread-safe -and- modified so you can tell which thread's doing the output). ------------------------------------------ Manny Veloso Digital Plumber IT Masters http://www.itmasters.com/ ------------------------------------------ "performance measurement is a drag."
From: William Edward Woody <woody@alumni.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Sun, 17 May 1998 01:14:07 -0700 Organization: In Phase Consulting Message-ID: <355E9C4F.EE69D20@alumni.caltech.edu> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lawson English wrote: > I still don't see how MacOS X is going to be the consumer version of MacOS. > > MacOS X is Rhapsody + Carbon, which is likely larger than Rhapsody. > > How is that "consumer-oriented?" Dunno; strip out the tools and just leave the basic OS + API, and I don't think this thing will have a footprint much larger than Windows 95. (Poking around in Windows 95, I found that the image set size just to boot the thing is 40 megs. Of course the thing uses VM, so it runs okay on a 16 meg machine, which is a fairly standard off-the-shelf configuration nowadays.) By the way, Win98 is than Win95, and yet it's a "consumer only" OS... - Bill -- William Edward Woody | In Phase Consulting woody@alumni.caltech.edu | Macintosh & MS Windows Development http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/
From: M Rassbach <mark@milestonerdl.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Sun, 17 May 1998 08:16:12 -0500 Organization: Inc.Net http://www.inc.net Message-ID: <355EE31C.7834F8EA@milestonerdl.com> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> <joe.ragosta-1605982005160001@elk32.dol.net> <355E3F66.86DD9F0@thegrid.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Ryan Kennedy wrote: > Joe Ragosta wrote: > > > But if they've left out the Unix from MacOS X, it could be smaller. > > Actually, if you're talking about size as in disk space then Unix will be > smaller. Unix was made to be small and simple, most kernels can actually be fit > on a 1.44 MB disk. And on my TAM with OS 8.1, the system used memory is 15 meg. On one of my Unix boxes, it fits in just under 5 meg. So Ragosta, where's this 'larger' FUD of yours coming from?
From: nobody@REPLAY.COM (Anonymous) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 17 May 1998 18:26:14 +0200 Organization: Replay Associates, L.L.P. Message-ID: <6jn336$5kk@basement.replay.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <MPG.fc14db42395aa139896a2@news.supernews.com>, don.brown@cesoft.com (Donald Brown) wrote: >In your apps, avoid strange tricks between applications. Do >NOT< send >an Apple Event from one app to another, pass the address of a buffer, and >have the second app write into the buffer in the first app's heap. Avoid >the system heap. Don, I was just looking at the "Transitioning to MacOS X" document. On page 46 it states "Carbon will fully support the Apple Event Manager" On page 66 it states "Carbon will fully support the Open Scripting Architecture". Where did you hear that you shouldn't support AEs for interapplication communication? Jeffrey jbk-at-world-dot-std-dot-com (remove all dashes and replace the dots and at)
From: jason@jhste1.dyn.ml.org (Jason S.) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: 17 May 1998 19:31:16 GMT Organization: I don't think so Message-ID: <slrn6luevt.5ee.jason@jhste1.dyn.ml.org> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> <joe.ragosta-1605982005160001@elk32.dol.net> <355E3F66.86DD9F0@thegrid.net> <355EE31C.7834F8EA@milestonerdl.com> M Rassbach posted the following to comp.sys.mac.advocacy: >> > But if they've left out the Unix from MacOS X, it could be smaller. >> Actually, if you're talking about size as in disk space then Unix will be >> smaller. Unix was made to be small and simple, most kernels can actually be fit >> on a 1.44 MB disk. >And on my TAM with OS 8.1, the system used memory is 15 meg. On one of my Unix >boxes, it fits in just under 5 meg. >So Ragosta, where's this 'larger' FUD of yours coming from? /usr/bin on my Linux/PowerPC box takes up 78MB. Obviously, leaving this stuff out would save 78MB. Is that so hard to understand? -- If FreeBSD actually did that, I would concede that FreeBSD was any more "correct" than Linux is, but not even the FreeBSD people can justify that kind of performance loss. -- Linus Torvalds on comp.unix.advocacy
Message-ID: <355F3C3C.E888B105@gain-ny.com> Date: Sun, 17 May 1998 15:36:28 -0400 From: "James T. Romano" <jromano@gain-ny.com> Organization: NoSpamAllowed MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Re: C++ and rhapsody References: <6jcvpm$25r$1@newsfep1.sprintmail.com> <6jd7db$555$1@crib.bevc.blacksburg.va.us> <6jdoil$76r$2@news.imaginet.fr> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit UGH!!! The last person that had a very very very long method name in my programming group at work, WAS FIRED!!! Pascal Bourguignon wrote: > nurban@crib.bevc.blacksburg.va.us (Nathan Urban) wrote: > >[...] > >Well, it turned out at some point that an 11-digit number was necessary > >after all in some extreme cases, so I simply went into the code (which I > >hadn't written) and just changed it to obj->Convert(num,buf,11). Except, > >as I found out, that last parameter _wasn't_ the size of the buffer to > >keep it from overflowing. It was the radix of the conversion, and I > >was getting base-11 numbers, leading to an odd bug. If the original > >code had been [obj convertNumber:num toString:buf usingRadix:10], > >that would never have happened. And in C++, no one would write > >obj->convertNumberToStringUsingRadix(num,buf,11), that's just illegible. > > Well, being the fool on the hill that I am, I do, at least since I leaned > Objective-C. But I always liked long descriptive names. So far, my longest > one is 97 characters. > > -- > __Pascal Bourguignon__ | The box said 'Requires Windows 95, > mailto:pjb@imaginet.fr | or better.' So I bought a Macintosh.
From: Sterling <Sterling_Augustine@stanford.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Sun, 17 May 1998 12:46:42 -0700 Organization: Stanford CS Message-ID: <355F3E90.1BD5@stanford.edu> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <6jn336$5kk@basement.replay.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit > Don, > I was just looking at the "Transitioning to MacOS X" document. On page > 46 it states "Carbon will fully support the Apple Event Manager" On page 66 > it states "Carbon will fully support the Open Scripting Architecture". > Where did you hear that you shouldn't support AEs for interapplication > communication? It's not the AppleEvent itself that is the problem, it is sending a pointer inside your address space to another process that is the problem. When you have different address spaces, that pointer will be meaningless to the other application. It will dereference it into it's own address space and find garbage. AppleEvents will live. Interapplication communication via pointers will die. (Actually, I believe there will be some methods, but apple events won't be the way.) Sterling
From: don.brown@cesoft.com (Donald Brown) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Sun, 17 May 1998 15:59:27 -0500 Organization: Prairie Group Message-ID: <MPG.fc90bab12a0eb5d9896ce@news.supernews.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <6jn336$5kk@basement.replay.com> In article <6jn336$5kk@basement.replay.com>, nobody@REPLAY.COM says... > In article <MPG.fc14db42395aa139896a2@news.supernews.com>, > don.brown@cesoft.com (Donald Brown) wrote: > > >In your apps, avoid strange tricks between applications. Do >NOT< send > >an Apple Event from one app to another, pass the address of a buffer, and > >have the second app write into the buffer in the first app's heap. Avoid > >the system heap. > > Don, > I was just looking at the "Transitioning to MacOS X" document. On page > 46 it states "Carbon will fully support the Apple Event Manager" On page 66 > it states "Carbon will fully support the Open Scripting Architecture". > Where did you hear that you shouldn't support AEs for interapplication > communication? There was another post covering this, but here's what I meant: If you have 16k of data, and you wrap it all up into an Apple Event and send it, this will continue to be supported. If instead you send the 4 byte point to the data, so the receiving app will read and write the buffer that's in the sending app's heap, that's the strange trick that won't work. Donald
Date: Sun, 17 May 1998 17:30:43 -0400 From: joe.ragosta@dol.net (Joe Ragosta) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Message-ID: <joe.ragosta-1705981730440001@elk85.dol.net> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> <joe.ragosta-1605982005160001@elk32.dol.net> <355E3F66.86DD9F0@thegrid.net> <355EE31C.7834F8EA@milestonerdl.com> In article <355EE31C.7834F8EA@milestonerdl.com>, mark@milestonerdl.com wrote: > Ryan Kennedy wrote: > > > Joe Ragosta wrote: > > > > > But if they've left out the Unix from MacOS X, it could be smaller. > > > > Actually, if you're talking about size as in disk space then Unix will be > > smaller. Unix was made to be small and simple, most kernels can actually be fit > > on a 1.44 MB disk. > > And on my TAM with OS 8.1, the system used memory is 15 meg. On one of my Unix > boxes, it fits in just under 5 meg. > > So Ragosta, where's this 'larger' FUD of yours coming from? Read what I said. I said that MacOSX _without_ Unix would be smaller than MacOSX _with_ Unix. Unless you've found a way for Unix to contain a negative number of bytes, my statement is clearly correct. -- Regards, Joe Ragosta See the Complete Macintosh Advocacy Page http://www.dol.net/~Ragosta/complmac.htm
From: "Michelle L. Buck" <buck.erik@mcleod.net> Newsgroups: comp.sys.next.programmer Subject: New display model in OS-X Date: Sun, 17 May 1998 22:29:15 -0500 Organization: McleodUSA - http://www.mcleodusa.net Message-ID: <6joa23$ik26@odie.mcleod.net> The people at Apple are brilliant. I am sure that they made the right decisions. Unfortunately, their compromises are not without sacrifices. There are real questions about what the graphics API will look like beyond NSBezierPath, the NSRect functions, NSColor, NSFont, and NSTextView. For instance, how will I get a character path ? Believe it or not, this is a very important issue. In other conversations, after the graphics session, several Apple employees admitted that the decision was recent and no work on the future graphics APIs has been started. The loss of NSHosting is a probably direct result of these decisions. Once the client/server architecture for drawing commands is abandoned, it is very difficult to recover. Applications will be made by default that can not or will not work well over a network even if Apple provides hooks in their APIs for remote display. As my wife's postings to Usenet indicate, the loss of NSHosting is probably necessary in order to support Carbon Quickdraw effectively. There is another even more alarming problem. If a Carbon application takes over the screen and does not give it back, no other application (Yellow or Carbon) will be able to display! A dead or infinitely looping Carbon application will prevent you from using the Process panel to kill it. This question was asked and clarified several times. Carbon applications will be able to take over the entire display and we must rely on their good nature to give it back and let other applications draw. I used to deride Windows for that flaw. PDF is NOT Postscript and true WYSIWYG is probably dead until PDF printers become available. Even then, some capabilities formerly taken for granted may be lost. EPS will now only be supported as bitmap thumbnails. I hope that large collections of EPS art can be converted. No existing PDF distiller has shown much promise. -Erik M. Buck
From: leigh@NOSPAMcs.uwa.edu.au (Leigh Smith) Newsgroups: comp.sys.next.programmer Subject: Re: [Q] What are the implications of MacOS X? Date: 18 May 1998 04:57:19 GMT Organization: The University of Western Australia Message-ID: <6jof3f$qn4$1@enyo.uwa.edu.au> References: <6jk951$1cc$53@ironhorse.plsys.co.uk> <B1836C75-7514F@206.165.43.149> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: english@primenet.com In <B1836C75-7514F@206.165.43.149> "Lawson English" wrote: > And you guys wonder why I scream about GX when we're dealing with Apple... > > Here's another question for you guys: > > Why can't a GX-like picture shape be used instead of PDF to transfer images > over the clipboard? Forgive me if I interpret this wrong, but PDF is a portable cross platform protocol whereas GX-like implies a protocol which would have support only within Apple. Given YellowBox for Windows will continue, it makes sense to support industry standard protocols. > > A list of shapes, including other pictures is far better than the pdf > format, IMHO, since it allows for easy editing of any component > shape/pciture, and is easily converted to PDF with no loss of fidelity. > This sounds like it could be an even higher level API that sits on top of PDF, perhaps as a third party API solution to fill in the gaps left by Apple dropping QuickDraw GX from Carbon. -- Leigh Computer Science, University of Western Australia Smith +61-8-9380-3778 leigh@cs.uwa.edu.au (NeXTMail/MIME) Microsoft - never has so much been made by so few, by pushing so much of so little, on so many with so little resistance.
From: Stephen Peters <portnoy@ai.mit.edu> Newsgroups: comp.sys.next.programmer Subject: Re: [Q] What are the implications of MacOS X? Date: 18 May 1998 01:18:36 -0400 Organization: Massachvsetts Institvte of Technology Sender: portnoy@crispy-critters Message-ID: <us5ogwww40j.fsf@ai.mit.edu> References: <6jherv$sqg$1@nnrp1.dejanews.com> <johns-1505980828570001@10.0.0.253> <6jhrog$jum$1@news.seicom.net> <6jk951$1cc$53@ironhorse.plsys.co.uk> mmalcolm crawford <malcolm@plsys.co.uk> writes: > Elsewhere Izumi Ohzawa pointed out that another thing that may be lost is > preview of ESP files, and the possibility of using EPS on the pasteboard. > This may be more of a blow. Frank, you're one of the PS/PDF gurus here -- > Izumi asked is there any way of producing "encapsulated PDF"? Yeah, I was just coming to that realization myself today. I'm also guessing that it removes the ease with which every OpenStep program adds in support for reading and writing EPS files. Unfortunately, this realization came at the same moment I realized that I'd been using EPS as a document interchange format for the past four years, and had *tons* of EPS files lying around. I've been using them constantly because they're easy to hack at (for me, at least) if there's something wrong, they can be used natively by most every word processing program on the planet (including LaTeX, my choice), and they're incredibly easy to manipulate within those programs. I'm really getting worried about this move. Sure, an EPDF might be possible, but how long will it be before I can easily use it on platforms *outside* MacOS-X? Maybe I should just buy Rhapsody and then never upgrade to MacOS X. Hell, I've been running NeXTSTEP 3.2 for the past five years... :-) -- Stephen L. Peters portnoy@ai.mit.edu PGP fingerprint: BFA4 D0CF 8925 08AE 0CA5 CCDD 343D 6AC6 "Poodle: The other white meat." -- Sherman, Sherman's Lagoon
From: "Ken Schuller" <schullersite@NOworldnet.att.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Mon, 18 May 1998 01:32:48 -0500 Organization: AT&T WorldNet Services Message-ID: <6jokj8$gq8@bgtnsc02.worldnet.att.net> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> <paul-1705982318100001@206.169.118.19> Paul Hamilton wrote in message ... > >Well Windows 98 is basically NT right? so no difference there. > >It's all relative anyway. > >Paul. >-- >Paul Hamilton >Software Engineer >mBED Software >paul@mbed.com >http://www.mbed.com Ohhhh boy... Not unless M$ changed the kernel and added NTFS support... Ken Schuller Network Systems Specialist NovaNET Learning, Inc. ======================== "In computing because it beats working for a living." I speak for me. Remove the obvious spam foil to reply via e-mail. ==========================================
From: drifterusa@sprintmail.com (John Bauer) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Sun, 17 May 1998 23:34:43 -0500 Organization: KAOS Message-ID: <199805172334432399994@sdn-ts-001txhousp01.dialsprint.net> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <6j7sdr$4gn$1@newsfep4.sprintmail.com> <6j7tij$vb7$1@crib.bevc.blacksburg.va.us> <stevenj-ya02408000R1105982039000001@news.mit.edu> <andyba-ya02408000R1205980916210001@news> <355873F0.6B86BC5C@milestonerdl.com> <andyba-ya02408000R1405981025210001@news> <01bd80f4$85082140$04387880@test1> <6jkuar$39a$1@xmission.xmission.com> <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> <joe.ragosta-1605982005160001@elk32.dol.net> <355E3F66.86DD9F0@thegrid.net> <355EE31C.7834F8EA@milestonerdl.com> M Rassbach wrote: > Ryan Kennedy wrote: > > > Joe Ragosta wrote: > > > > > But if they've left out the Unix from MacOS X, it could be smaller. > > > > Actually, if you're talking about size as in disk space then Unix will > > be smaller. Unix was made to be small and simple, most kernels can > > actually be fit on a 1.44 MB disk. > > And on my TAM with OS 8.1, the system used memory is 15 meg. On one of my > Unix boxes, it fits in just under 5 meg. > > So Ragosta, where's this 'larger' FUD of yours coming from? Lawson says: Mac OS X = Rhapsody + Carbon, ergo, X is probably larger than Rhapsody. Joe says: Mac OS X = Rhapsody + Carbon - BSD app environment/tools - dev tools, ergo, X may be smaller than Rhapsody. You offer anecdotal evidence that does nothing to show that what's needed to run Carbon is larger than the Unix and dev tools Rhapsody comes with. And you accuse Joe of spreading FUD? -- John Bauer <remove NOT from email address above, just like the spammers do>
From: paul@mbed.com (Paul Hamilton) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Sun, 17 May 1998 23:18:10 -0800 Organization: mBED Software Message-ID: <paul-1705982318100001@206.169.118.19> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> In article <B1837072-84146@206.165.43.149>, "Lawson English" <english@primenet.com> wrote: > > I still don't see how MacOS X is going to be the consumer version of MacOS. > > MacOS X is Rhapsody + Carbon, which is likely larger than Rhapsody. > > How is that "consumer-oriented?" Well Windows 98 is basically NT right? so no difference there. It's all relative anyway. Paul. -- Paul Hamilton Software Engineer mBED Software paul@mbed.com http://www.mbed.com
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: Q:Catching Uncaught Exceptions Date: Mon, 18 May 1998 11:07:24 +0200 Organization: Square B.V. Message-ID: <355FFA4C.BB677321@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I'm trying to catch exceptions that I should have catched somewhere else but didn't. The documentation says one should register an handler with NSSetUncaughtExceptionHandler. I do this, but when I raise an exception on purpose (after setting the handler) it doesn't get caught by the handler. The code included below is a part of my application. The applicationWillFinishLaunching registers the handler, the applicationDidFinishLaunching raises an exception. It doesn't get caught. Why-oh-why? Maurice le Rutte. volatile void SQUncaughtExceptionHandler(NSException *exception) { NSLog(@"Name:%@ Reason:%@ UserInfo:%@",[exception name], [exception reason],[exception userInfo]); exit(-1); } @implementation AppController -(void)applicationWillFinishLaunching:(NSNotification *)notification { NSSetUncaughtExceptionHandler(SQUncaughtExceptionHandler); } - (void)applicationDidFinishLaunching:(NSNotification *)notification { [ [NSException exceptionWithName:@"Name" reason:@"Reason" userInfo:[NSDictionary dictionary]] raise]; } @end -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto:mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+ Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" and "to trick into believing or accepting as genuine something false and often preposterous"
From: William Edward Woody <woody@alumni.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Mon, 18 May 1998 02:19:29 -0700 Organization: In Phase Consulting Message-ID: <355FFD21.B50FF5C1@alumni.caltech.edu> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> <paul-1705982318100001@206.169.118.19> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Paul Hamilton wrote: > Well Windows 98 is basically NT right? so no difference there. Ummmm, no it's not. Microsoft has indicated that the next generation after Win98 will be a stripped down NT (or perhaps not). But the current Win98 kernel is basically Win95 + IE4 with a new front end and a bunch of new drivers and services. Of course the NT kernel itself is not all that big--the size of WinNT or Win95/98 is the rest of the junk. So whether or not Win98 contains the NT kernel or not is somewhat moot to the discussion at hand. - Bill (Who notes that in my dual-boot Windows environment, Win95 and WinNT take about the same amount of space.) -- William Edward Woody | In Phase Consulting woody@alumni.caltech.edu | Macintosh & MS Windows Development http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.next.programmer Subject: Re: [Q] What are the implications of MacOS X? Date: 18 May 1998 02:41:00 -0700 Organization: Primenet Services for the Internet Message-ID: <B185503A-4EB22@206.165.43.155> References: <6jof3f$qn4$1@enyo.uwa.edu.au> MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Leigh Smith <leigh@NOSPAMcs.uwa.edu.au> said: >In <B1836C75-7514F@206.165.43.149> "Lawson English" wrote: >> And you guys wonder why I scream about GX when we're dealing with >Apple... >> >> Here's another question for you guys: >> >> Why can't a GX-like picture shape be used instead of PDF to transfer >images >> over the clipboard? > >Forgive me if I interpret this wrong, but PDF is a portable cross platform >protocol whereas GX-like implies a protocol which would have support only >within Apple. Given YellowBox for Windows will continue, it makes sense >to >support industry standard protocols. >> >> A list of shapes, including other pictures is far better than the pdf >> format, IMHO, since it allows for easy editing of any component >> shape/pciture, and is easily converted to PDF with no loss of fidelity. >> > >This sounds like it could be an even higher level API that sits on top of >PDF, perhaps as a third party API solution to fill in the gaps left by Apple >dropping QuickDraw GX from Carbon. But why not use it for transfering images over the clipboard? If the image is merely to be cut/copied/pasted, there's no reason to use an "industry-standard protocol" between applications when a simpler and more easily edited (programatically, not textually) solution could be used instead. ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: "Gabor Greif" <gabor.greif@no.drsolomon.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 18 May 98 12:44:23 +0200 Organization: noris network GmbH Message-ID: <B185DDB5-9C3BC@192.168.1.8> References: <355C9B47.7C5A@stanford.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit nntp://news.noris.de/comp.sys.mac.programmer.codewarrior, nntp://news.noris.de/comp.sys.next.programmer On Fre, 15. Mai 1998 21:45 Uhr, Sterling <mailto:Sterling_Augustine@stanford.edu> wrote: >True enough. But when the memory manager becomes reentrant--as it >presumably will under Carbon--if handles can still move, then it still >won't be safe to make an app's singly dereffed handled invalid right in >the middle of some function that doesn't expect memory to move. And you >therefore still won't be able to allocate memory during a completion >routine. Maybe Apple knows of a way to solve this. I don't. > Without being an expert I would guess that there will be no asynchronous completion routines in Carbon. The async routines are needed only in the absence of preemptive threads. If you have threads you simply issue a threaded request (that looks like a synchronous one) and your thread will be scheduled again when the request has completed. Maybe the completion routines will be supported for the masochistic ones among us :-) Gabor
From: wilkie@cg.tuwien.ac.at (Alexander Wilkie) Newsgroups: comp.sys.next.programmer Subject: Re: New display model in OS-X Date: 18 May 1998 10:46:22 GMT Organization: Vienna University of Technology, Austria Message-ID: <6jp3hu$5mr$1@news.tuwien.ac.at> References: <6joa23$ik26@odie.mcleod.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: buck.erik@mcleod.net In <6joa23$ik26@odie.mcleod.net> "Michelle L. Buck" wrote: > The people at Apple are brilliant. Given what you say later on, this sounds rather hollow to me. Especially: > There is another even more alarming problem. If a Carbon application takes > over the screen and does not give it back, no other application (Yellow or > Carbon) will be able to display! A dead or infinitely looping Carbon > application will prevent you from using the Process panel to kill it. If this is true, then either the engineers at Apple all got a major lobotomy as part of their "health plan" and/or they have lost all professional decency. > I used to deride Windows for that flaw. It's not a flaw. If the abovestated is true, then the McX display system is _broken_. Period. > PDF is NOT Postscript and true WYSIWYG is probably dead until PDF printers > become available. Which will probably happen sometime after hell has frozen over solidly. > EPS will now only be supported as bitmap thumbnails. I'm speechless. Is Lawson English the new CTO at Apple? I mean, they've got a good (read: not perfect, but quite decent) working display model right now. And they go out and discard it in favour of _this_? Are they crazy? Someone has said that the display model switch was politically motivated. I guess this means "politics" as in "North Korean trade policy" or such... Somebody please pinch me. I want to wake up NOW... $0.2E-32 Alexander Wilkie -- e-mail: wilkie@cg.tuwien.ac.at (NeXTMail preferred, MIME o.k.) www : http://www.cg.tuwien.ac.at/staff/AlexanderWilkie.html pgp : finger wilkie@www.cg.tuwien.ac.at
From: see-below@not-my-address.com (Matthew Vaughan) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 18 May 1998 04:40:35 -0700 Message-ID: <see-below-1805980440350001@209.24.240.12> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <6jn336$5kk@basement.replay.com> In article <6jn336$5kk@basement.replay.com>, nobody@REPLAY.COM (Anonymous) wrote: > In article <MPG.fc14db42395aa139896a2@news.supernews.com>, > don.brown@cesoft.com (Donald Brown) wrote: > > >In your apps, avoid strange tricks between applications. Do >NOT< send > >an Apple Event from one app to another, pass the address of a buffer, and > >have the second app write into the buffer in the first app's heap. Avoid > >the system heap. > > Don, > I was just looking at the "Transitioning to MacOS X" document. On page > 46 it states "Carbon will fully support the Apple Event Manager" On page 66 > it states "Carbon will fully support the Open Scripting Architecture". > Where did you hear that you shouldn't support AEs for interapplication > communication? You're only reading the first half of his second sentence. Read the second half of that sentence again. .................................................... MATTHEW VAUGHAN matthewv at best dot com (damn spammers...) http://www.best.com/~matthewv/ ....................................................
From: see-below@not-my-address.com (Matthew Vaughan) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 18 May 1998 04:41:56 -0700 Message-ID: <see-below-1805980441570001@209.24.240.12> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <6jn336$5kk@basement.replay.com> <355F3E90.1BD5@stanford.edu> In article <355F3E90.1BD5@stanford.edu>, Sterling <Sterling_Augustine@stanford.edu> wrote: > > Don, > > I was just looking at the "Transitioning to MacOS X" document. On page > > 46 it states "Carbon will fully support the Apple Event Manager" On page 66 > > it states "Carbon will fully support the Open Scripting Architecture". > > Where did you hear that you shouldn't support AEs for interapplication > > communication? > > It's not the AppleEvent itself that is the problem, it is sending a > pointer inside your address space to another process that is the > problem. When you have different address spaces, that pointer will be > meaningless to the other application. It will dereference it into it's > own address space and find garbage. > > AppleEvents will live. Interapplication communication via pointers will > die. (Actually, I believe there will be some methods, but apple events > won't be the way.) More simply put, the two programs can no longer directly access the same address space, right? .................................................... MATTHEW VAUGHAN matthewv at best dot com (damn spammers...) http://www.best.com/~matthewv/ ....................................................
From: don.brown@cesoft.com (Donald Brown) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 18 May 1998 09:14:02 -0500 Organization: Prairie Group Message-ID: <MPG.fc9fe2378ddbc2d9896d8@news.supernews.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <6jn336$5kk@basement.replay.com> <355F3E90.1BD5@stanford.edu> <see-below-1805980441570001@209.24.240.12> In article <see-below-1805980441570001@209.24.240.12>, see-below@not-my- address.com says... > In article <355F3E90.1BD5@stanford.edu>, Sterling > <Sterling_Augustine@stanford.edu> wrote: > > > > Don, > > > I was just looking at the "Transitioning to MacOS X" document. On page > > > 46 it states "Carbon will fully support the Apple Event Manager" On page 66 > > > it states "Carbon will fully support the Open Scripting Architecture". > > > Where did you hear that you shouldn't support AEs for interapplication > > > communication? > > > > It's not the AppleEvent itself that is the problem, it is sending a > > pointer inside your address space to another process that is the > > problem. When you have different address spaces, that pointer will be > > meaningless to the other application. It will dereference it into it's > > own address space and find garbage. > > > > AppleEvents will live. Interapplication communication via pointers will > > die. (Actually, I believe there will be some methods, but apple events > > won't be the way.) > > > More simply put, the two programs can no longer directly access the same > address space, right? Right...but to some extent that's a buzzword. I thought an example might be easier to understand just what effect it has. Donald
From: seanl@cs.umd.edu (Sean Luke) Newsgroups: comp.sys.next.programmer Subject: Re: Graphics question Date: 18 May 1998 15:35:27 GMT Organization: U of Maryland, Dept. of Computer Science, Coll. Pk., MD 20742 Message-ID: <6jpkfv$nl$2@walter.cs.umd.edu> References: <355b5852.0@news.depaul.edu> <6jfmln$138$2@news.xmission.com> NNTP-Posting-User: seanl Don Yacktman (don@misckit.com) wrote: : Jonathan W Hendry <jhendry@shrike.depaul.edu> wrote: : > : > I want to take a bitmap and 'sepia-tone' it. Convert it : > into shades of one hue (any hue, not necessarily sepia brown). : So, you have to do this: : (R,G,B) -> (H,S,B) : H = c1 : S = c2 : B *= c3 : (H,S,B) -> (R,G,B) There's an easier way. What Jon wants to do, really, is to project colors onto the color vector for sepia. Let's say you have a color vector R=<r,g,b> that you want to convert into some brightness value variation on sepia, which you've defined as the color vector S=<x,y,z>. Then you project R onto S, which I *think* is S times the scalar value (S . R)|R|/|S| (is that right?). Your hitch is going to be when the length of the resultant vector is greater than 1; you can either normalize your whole picture by the maximum vector value of any vector in the picture, but the easiest thing to do, I think, is to just divide S by 2, since its largest length can't be bigger than that (actually, it technically can't be bigger than Sqrt(3), but dividing by 2 is more efficient if you're dealing with bytes). Sean Luke seanl@cs.umd.edu
From: kalash@starnine.com (Joe Kalash) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 18 May 1998 18:14:20 -0800 Organization: StarNine Technologines, Inc. Message-ID: <kalash-1805981814200001@boris.starnine.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <kalash-1405981828460001@boris.starnine.com> <355C0CDF.6D2314E@alumni.caltech.edu> <kalash-1605980037300001@kalash-home.starnine.com> <355D7537.45A26CFA@alumni.caltech.edu> <6jk06k$n53$13@news.idiom.com> <355DCF43.DBF0BD67@alumni.caltech.edu> In article <355DCF43.DBF0BD67@alumni.caltech.edu>, William Edward Woody <woody@alumni.caltech.edu> wrote: > 3. Handles are slow, inefficient, and by their very existance > cause the whole system to grind to a screaching halt. > > And I certainly don't buy (3), as it's clear to me that if Apple > were to support handles on a "modern virtual memory system", they'd > know enough to change the algorithms used to manage those handles. Oh, jeez. Among all the things Apple has to do, rewritting the Memory Manager into UNIX to support moving your handles about is sure to be at the bottom of their list. The number of programmers who will need such a tool is tiny (if you find yourself thrashing, working on an algorithim level is normally a lot easier, and more productive). Even my little 386 BSDI machine has a per application data limit of 33 meg. Please, if you are sure Handles are the answer to a problem you have, by all means implement them. The brk() and sbrk() functions (what underlies malloc) are really easy to use. But don't make me pay for all this. -- Joe Kalash StarNine Technologies, Inc. kalash@starnine.com
From: pete@ohm.york.ac.uk (-bat.) Newsgroups: comp.sys.next.programmer Subject: Re: Graphics question Date: 18 May 1998 17:15:49 GMT Organization: The University of York, UK Sender: pcf1@york.ac.uk Message-ID: <6jpqc5$jpn$1@pump1.york.ac.uk> References: <6jpkfv$nl$2@walter.cs.umd.edu> seanl@cs.umd.edu (Sean Luke) writes: > There's an easier way. What Jon wants to do, really, is to project colors > onto the color vector for sepia. Let's say you have a color vector R=<r,g,b> ..i.e. extract the brightness values for each pixel (unless I missed something). Which you can do with: grey = (0.30 red) + (0.59 green) + (0.11 blue) which correspond approximately to the weightings your eye gives the different colours. thats what I do and it works fine for me. -bat.
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.next.programmer Subject: Re: Graphics question Date: 18 May 1998 17:24:05 GMT Organization: MiscKit Development Message-ID: <6jpqrl$mhk$3@news.xmission.com> References: <355b5852.0@news.depaul.edu> <6jfmln$138$2@news.xmission.com> <6jpkfv$nl$2@walter.cs.umd.edu> seanl@cs.umd.edu (Sean Luke) wrote: > Don Yacktman (don@misckit.com) wrote: > : Jonathan W Hendry <jhendry@shrike.depaul.edu> wrote: > : > > : > I want to take a bitmap and 'sepia-tone' it. Convert it > : > into shades of one hue (any hue, not necessarily sepia brown). > > : So, you have to do this: > > : (R,G,B) -> (H,S,B) > : H = c1 > : S = c2 > : B *= c3 > : (H,S,B) -> (R,G,B) > > There's an easier way. What Jon wants to do, really, is to > project colors onto the color vector for sepia. [...] I would argue that my approach is easier for most people to understand than the raw vector approach, and hence easier for many to implement...which is why I suggested it. If you look at my approach closely, I actually _am_ doing what amounts to a vector projection, but I'm doing it in HSB space instead of RGB space. This is analogous to moving a vector expressed in cartesian coordinates into polar coordinates because it simplifies the calculations in some instances. In this case, the projection is of a 3-D vector onto a 1-D line, since c1 and c2 are constants. And c3 is the normalizing value that you suggested... Now, anyone who would like to move the whole thing back into RGB space is certainly welcome to do so. A little extra math up front would make for a significantly faster algorithm, assuming you get the math right. :-) -- Later, -Don Yacktman don@misckit.com <a href="http://www.misckit.com/don.html">My home page</a>
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.next.programmer Subject: Re: Graphics question Date: 18 May 1998 17:31:21 GMT Organization: MiscKit Development Message-ID: <6jpr99$mhk$4@news.xmission.com> References: <6jpkfv$nl$2@walter.cs.umd.edu> <6jpqc5$jpn$1@pump1.york.ac.uk> pete@ohm.york.ac.uk (-bat.) wrote: > seanl@cs.umd.edu (Sean Luke) writes: > > There's an easier way. What Jon wants to do, really, is to > > project colors onto the color vector for sepia. Let's say > > you have a color vector R=<r,g,b> > > ..i.e. extract the brightness values for each pixel (unless I missed > something). Which you can do with: > > grey = (0.30 red) + (0.59 green) + (0.11 blue) > > which correspond approximately to the weightings your eye gives the > different colours. thats what I do and it works fine for me. Yeah, and then to tint it, you take your sepia color and multiply each r,g,b by the gray value for the pixel: [r|g|b]1 are inputs [r|g|b]2 are outputs gr = 0.30 * r1 + 0.59 * g1 + 0.11 * b1; // get brightness r2 = gr * 0.44; // add tint g2 = gr * 0.27; b2 = gr * 0.15; Note that the last three constants, if changed, would cause a different tinting. -- Later, -Don Yacktman don@misckit.com <a href="http://www.misckit.com/don.html">My home page</a>
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Followup-To: comp.sys.mac.advocacy,comp.sys.next.advocacy Date: 18 May 1998 18:20:16 GMT Organization: P & L Systems Message-ID: <6jpu50$1cc$102@ironhorse.plsys.co.uk> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: english@primenet.com NOTE: FOLLOWUPS TO COMP.SYS.MAC.ADVOCACY,COMP.SYS.NEXT.ADVOCACY In <B1837072-84146@206.165.43.149> "Lawson English" wrote: > I still don't see how MacOS X is going to be the consumer version of MacOS. > Why am I not surprised. > MacOS X is Rhapsody + Carbon, which is likely larger than Rhapsody. > > How is that "consumer-oriented?" > Because people like Adobe are porting their consumer products to Carbon, so there will be Brand Name consumer-level apps which Mac users will be happy with, and because X > 9 -- it's Apple's future OS. Or are you going to tell me that Apple's going to drop the consumer market? mmalc.
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: [Q] What are the implications of MacOS X? Date: 19 May 1998 18:31:32 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6jsj64$rs9$3@ns3.vrx.net> References: <6jherv$sqg$1@nnrp1.dejanews.com> <johns-1505980828570001@10.0.0.253> <6jhrog$jum$1@news.seicom.net> <6jrr6r$4rf$10@ns3.vrx.net> <6jssqn$ara$1@news.digifix.com> <6jt28k$60m$1@tribune.usask.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: eric@skatter.USask.Ca In <6jt28k$60m$1@tribune.usask.ca> eric@skatter.USask.Ca claimed: > I'd like to read about these `language features'. I've got several > applications we use for real-time display of data acquired from various > monitors connected to our linear accelerator. These apps make pretty heavy > use of compositing and user paths. Are these constructs going to make the > jump to the new graphics system? Yup. Both were mentioned specifically. Apparently the compositing system will also allow transparency universally too, not just in the NSImage and such. Maury
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: [Q] What are the implications of MacOS X? Date: 19 May 1998 18:32:34 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6jsj82$rs9$4@ns3.vrx.net> References: <6jherv$sqg$1@nnrp1.dejanews.com> <johns-1505980828570001@10.0.0.253> <6jhrog$jum$1@news.seicom.net> <6jrr6r$4rf$10@ns3.vrx.net> <6jsuk2$nd6$1@news.seicom.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: frank@this.NO_SPAM.net In <6jsuk2$nd6$1@news.seicom.net> Frank M. Siegert claimed: > Maury, get a life! Geez, what I do now? > It all depends on the type of app you are working on. Fortunately I did went > away from DPS quite a while ago. So, no problem? Maury
From: frank@this.NO_SPAM.net (Frank M. Siegert) Newsgroups: comp.sys.next.programmer Subject: Re: [Q] What are the implications of MacOS X? Date: 19 May 1998 21:46:42 GMT Organization: Frank's Area 51 Message-ID: <6jsuk2$nd6$1@news.seicom.net> References: <6jherv$sqg$1@nnrp1.dejanews.com> <johns-1505980828570001@10.0.0.253> <6jhrog$jum$1@news.seicom.net> <6jrr6r$4rf$10@ns3.vrx.net> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Cc: maury@remove_this.istar.ca In <6jrr6r$4rf$10@ns3.vrx.net> Maury Markowitz wrote: > Why? If you wrote good code you shouldn't see much of a difference except > that it will get faster - apparently a lot. > > I write a graphics app. I'm looking at about 50 lines that have to change. > Boo hoo. You comment above seems to indicate that you don't understand the > issues involved. Maury, get a life! I did write an interactive PostScript tracer/debugger beside other stuff using the DPS layer in a more direct way, think about rewriting such applications. If you use the appkit and the DPS layer only in a passive way, be happy. If you don't - well, bad karma. And don't tell me my programs are 'bad' or 'good' because you are standing on the side of luck. It all depends on the type of app you are working on. Fortunately I did went away from DPS quite a while ago. -- * Frank M. Siegert [frank@this.net] - Home http://www.this.net/~frank * NeXTSTEP, IRIX, Solaris, Linux, BeOS, PDF & PostScript Wizard * "The answer is vi, what was your question...?"
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: [Q] What are the implications of MacOS X? Date: 18 May 1998 20:30:21 GMT Organization: Digital Fix Development Message-ID: <6jq5ot$bcb$1@news.digifix.com> References: <6jof3f$qn4$1@enyo.uwa.edu.au> <B185503A-4EB22@206.165.43.155> In-Reply-To: <B185503A-4EB22@206.165.43.155> On 05/18/98, "Lawson English" wrote: >But why not use it for transfering images over the clipboard? If the image >is merely to be cut/copied/pasted, there's no reason to use an >"industry-standard protocol" between applications when a simpler and more >easily edited (programatically, not textually) solution could be used >instead. > It increases the size of the codebase, which increases the cost. It complicates the handling. Fact is, there is no reason that you couldn't put your own data on the pasteboard in addition to the PDF and make it available to the application. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: M Rassbach <mark@milestonerdl.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Mon, 18 May 1998 06:01:48 -0500 Organization: Inc.Net http://www.inc.net Message-ID: <3560151C.879C676@milestonerdl.com> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> <paul-1705982318100001@206.169.118.19> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Paul Hamilton wrote: > Well Windows 98 is basically NT right? so no difference there. Nope > It's all relative anyway. Win 98 is the last living relative of DOS. Microsoft said so in a press release. To 'add functionality' (avoid concente decree over DOS/Windows/Win 95) Micro$oft is making a 'break' from DOS....and is going to package NT as the desktop....next release.
From: eric@skatter.USask.Ca Newsgroups: comp.sys.next.programmer Subject: Re: [Q] What are the implications of MacOS X? Date: 19 May 1998 22:48:52 GMT Organization: University of Saskatchewan Message-ID: <6jt28k$60m$1@tribune.usask.ca> References: <6jherv$sqg$1@nnrp1.dejanews.com> <johns-1505980828570001@10.0.0.253> <6jhrog$jum$1@news.seicom.net> <6jrr6r$4rf$10@ns3.vrx.net> <6jssqn$ara$1@news.digifix.com> sanguish@digifix.com (Scott Anguish) wrote: > The problem is that there is a mis-understanding. > > DPS isn't being replaced by Quickdraw. DPS is being replaced >by a new engine that has all the drawing features of DPS without the >language features. > > Enhanced Quickdraw and the Yellow Box graphics classes both >sit on top of the new Imaging Model. > I'd like to read about these `language features'. I've got several applications we use for real-time display of data acquired from various monitors connected to our linear accelerator. These apps make pretty heavy use of compositing and user paths. Are these constructs going to make the jump to the new graphics system? -- Eric Norum eric@skatter.usask.ca Saskatchewan Accelerator Laboratory Phone: (306) 966-6308 University of Saskatchewan FAX: (306) 966-6058 Saskatoon, Canada.
From: cahisnay <cahisnay@erols.com> Newsgroups: comp.sys.next.programmer Subject: OpenWrite API Date: Tue, 19 May 1998 21:07:12 -0400 Organization: PDH, Inc. Message-ID: <35622CC0.B484DFB4@erols.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi folks, Hope you can help. I have access to the Lighthouse product OpenWrite PRO API version 2.1. I would like to exercize the API through a NEXTSTEP distributed object interface. What .h files do I need and where are they located. Any information is hugely appreciated. Thanks in advance. Carole H.
From: spam_panda@mis.net (Jason Stephenson) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <spam_panda-1905982118030001@lexcas36.mis.net> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <6j84dp$vln$1@crib.bevc.blacksburg.va.us> Date: Wed, 20 May 1998 01:13:50 GMT NNTP-Posting-Date: Tue, 19 May 1998 18:13:50 PDT In article <6j84dp$vln$1@crib.bevc.blacksburg.va.us>, nurban@vt.edu wrote: > In article <MPG.fc14db42395aa139896a2@news.supernews.com>, don.brown@cesoft.com (Donald Brown) wrote: > > > Here is the path they have described today: > > > You learn and program with the Mac APIs. There is a small percentage of > > those that are going away and being replaced, but you will mostly be > > working with the APIs we've been using for decades. Go ahead and learn > > the Mac API. > > Better advice: if you're writing a new app, forget the Mac API. > OpenStep is infinitely better. Might as well develop good habits from > the beginning, and not waste your time. > Best advice: Ditch the MacOS entirely. Get MkLinux, 68K Linux, or NetBSD. Help develop an operating system. Join the Free Software Revolution! Besides, programming for these other OSes is *much* easier than programming the MacOS or Windows NT. Besides, if you believe the press, Apple is dying and Microsoft is about to be split up into a kajillion pieces. How can you count on an OS when you don't have the source code?
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: [Q] What are the implications of MacOS X? Date: 19 May 1998 21:16:07 GMT Organization: Digital Fix Development Message-ID: <6jssqn$ara$1@news.digifix.com> References: <6jherv$sqg$1@nnrp1.dejanews.com> <johns-1505980828570001@10.0.0.253> <6jhrog$jum$1@news.seicom.net> <6jrr6r$4rf$10@ns3.vrx.net> In-Reply-To: <6jrr6r$4rf$10@ns3.vrx.net> On 05/19/98, Maury Markowitz wrote: >In <6jhrog$jum$1@news.seicom.net> Frank M. Siegert claimed: >> Except for the dropped DPS functionality - replaceing DPS by Quickdraw >seems >> to me like removing the wheels from my car and putting rollerskates in >place. > > Why? If you wrote good code you shouldn't see much of a difference except >that it will get faster - apparently a lot. > The problem is that there is a mis-understanding. DPS isn't being replaced by Quickdraw. DPS is being replaced by a new engine that has all the drawing features of DPS without the language features. Enhanced Quickdraw and the Yellow Box graphics classes both sit on top of the new Imaging Model. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: don.brown@cesoft.com (Donald Brown) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Tue, 19 May 1998 21:47:20 -0500 Organization: Prairie Group Message-ID: <MPG.fcc0036a7f17c1f9896f1@news.supernews.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <6j84dp$vln$1@crib.bevc.blacksburg.va.us> <spam_panda-1905982118030001@lexcas36.mis.net> In article <spam_panda-1905982118030001@lexcas36.mis.net>, spam_panda@mis.net says... > In article <6j84dp$vln$1@crib.bevc.blacksburg.va.us>, nurban@vt.edu wrote: > > > In article <MPG.fc14db42395aa139896a2@news.supernews.com>, > don.brown@cesoft.com (Donald Brown) wrote: > > > > > Here is the path they have described today: > > > > > You learn and program with the Mac APIs. There is a small percentage of > > > those that are going away and being replaced, but you will mostly be > > > working with the APIs we've been using for decades. Go ahead and learn > > > the Mac API. > > > > Better advice: if you're writing a new app, forget the Mac API. > > OpenStep is infinitely better. Might as well develop good habits from > > the beginning, and not waste your time. > > > > Best advice: Ditch the MacOS entirely. Get MkLinux, 68K Linux, or NetBSD. > Help develop an operating system. Join the Free Software Revolution! > Besides, programming for these other OSes is *much* easier than > programming the MacOS or Windows NT. > > Besides, if you believe the press, Apple is dying and Microsoft is about > to be split up into a kajillion pieces. How can you count on an OS when > you don't have the source code? > The problem is with both the other OSes and OpenStep, when you create something, how many people will be running it? Donald
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer,comp.sys.next.advocacy Subject: Re: future of mac programming Followup-To: comp.sys.mac.advocacy,comp.sys.next.advocacy Date: 19 May 1998 23:18:25 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6jti21$6a$1@crib.bevc.blacksburg.va.us> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <6j84dp$vln$1@crib.bevc.blacksburg.va.us> <spam_panda-1905982118030001@lexcas36.mis.net> In article <spam_panda-1905982118030001@lexcas36.mis.net>, spam_panda@mis.net (Jason Stephenson) wrote: > In article <6j84dp$vln$1@crib.bevc.blacksburg.va.us>, nurban@vt.edu wrote: > > In article <MPG.fc14db42395aa139896a2@news.supernews.com>, > don.brown@cesoft.com (Donald Brown) wrote: > > > Here is the path they have described today: > > > You learn and program with the Mac APIs. There is a small percentage of > > > those that are going away and being replaced, but you will mostly be > > > working with the APIs we've been using for decades. Go ahead and learn > > > the Mac API. > > Better advice: if you're writing a new app, forget the Mac API. > > OpenStep is infinitely better. Might as well develop good habits from > > the beginning, and not waste your time. > Best advice: Ditch the MacOS entirely. Get MkLinux, 68K Linux, or NetBSD. > Help develop an operating system. Join the Free Software Revolution! Gee, I just thought I wanted to use the system that lets me get my work done the easiest. When did I have to join a revolution to do that? I'm not interested in developing an operating system anyway. I'm interested in writing tools for the operating system I like, which is not Linux. (Well, Linux isn't bad. Just not the best. Good core OS. But what's on top of it could be better.) FYI: at various times, I've had NEXTSTEP, OPENSTEP for Mach, Rhapsody, Windows 95, Windows NT, OS/2, Linux, and the Hurd installed on my computer. (I'm also going to put OpenBSD on my old DECstation.) As of right now, I've got OPENSTEP (replacing NEXTSTEP), Rhapsody, NT, 95 (which I never use anymore and am going to delete), and have already deleted OS/2, Linux, and the Hurd (I might put the Hurd back sometime to fool around with 'cause it's cool, especially if they ever decide to port it to Fluke). I use OPENSTEP about 95% of the time and NT and Rhapsody the other 5%, though the Rhapsody percentage is due to increase once I get RDR2. OPENSTEP is simply a better user and developer environment than Linux, for me at least. My next computer will not be a PC, it will be a Mac. It may or may not have MkLinux. Of course, with ever-increasing hard disk sizes, I guess I might as well throw it on even if I never really use it. > Besides, programming for these other OSes is *much* easier than > programming the MacOS or Windows NT. <snicker> Spoken like a true Linux bigot. Ignorance is bliss, eh? You obviously have never programmed for OpenStep. I wonder if you even know what it is? > Besides, if you believe the press, I don't. Do _you_? How naive. > How can you count on an OS when you don't have the source code? I've got most of what I need, and MacOS X will have more than Linux offers for my interests. [Followups to comp.sys.next.advocacy, comp.sys.mac.advocacy.]
From: *johnnyc*@or.psychology.dal.ca (John Christie) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 20 May 1998 10:28:37 -0300 Organization: ISINet, Nova Scotia Message-ID: <*johnnyc*-2005981028370001@jchristie.halifaxcable.dal.ca> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <6jn336$5kk@basement.replay.com> <355F3E90.1BD5@stanford.edu> <see-below-1805980441570001@209.24.240.12> In article <see-below-1805980441570001@209.24.240.12>, see-below@not-my-address.com (Matthew Vaughan) wrote: > More simply put, the two programs can no longer directly access the same > address space, right? Yes, and YAY! Anyone who really wants this terribly unsafe way of moving information around inside OS X raise their hand now. -- You aren't free if you CAN choose - only if you DO choose. All you are is the decisions you make. Remove "*" and ohnny (i.e. jc@) to reply via email
From: John Hornkvist <sorry@no.more.spams> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 20 May 1998 14:44:14 GMT Organization: Chalmers Tekniska Högskola Sender: john@haddock.cd.chalmers.se (John Hprnkvist) Message-ID: <Et9FLq.J6F@haddock.cd.chalmers.se> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <6jn336$5kk@basement.replay.com> <355F3E90.1BD5@stanford.edu> <see-below-1805980441570001@209.24.240.12> <*johnnyc*-2005981028370001@jchristie.halifaxcable.dal.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: *johnnyc*@or.psychology.dal.ca In <*johnnyc*-2005981028370001@jchristie.halifaxcable.dal.ca> John Christie wrote: >> More simply put, the two programs can no longer directly access the same >> address space, right? > Yes, and YAY! Anyone who really wants this terribly unsafe way of > moving information around inside OS X raise their hand now. This is actually a good way of passing information around. You just need to be a bit careful about it... :) For example the Unix call "fork" gains a lot of performance from being implemented with shared memory -- copy on write shared memory, mind you. Since all new processes are started with a "fork" call, which used to take a copy of the calling process' entire memory, which is normally followed by an "exec" where that memory area is written over with a new program, copy on write memory saves a lot of unnecessary copying. Mach (2 at least) uses copy on write shared memory for message passing. This saves quite a bit of unnecessary copying when a large lump of data is sent. Migrating threads is an even cooler way of doing this; instead of passing data around you pass behaviour around. That can make message passing quite a bit faster -- impressively so, even. I guess it is or will be used in the OSF Mach 3 version that Mac OS X builds on. (Migrating threads can cut the number of (expensive) full context switches down a whole lot. Instead of switching to another task (thread + memory protection), you just switch the memory protection. A bit risky if you don't implement it right, but incredibly powerful.) Regards, John Hornkvist Name: nhoj Address: cd.chalmers.se
From: *johnnyc*@or.psychology.dal.ca (John Christie) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 20 May 1998 14:59:02 -0300 Organization: ISINet, Nova Scotia Message-ID: <*johnnyc*-2005981459020001@jchristie.halifaxcable.dal.ca> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <6jn336$5kk@basement.replay.com> <355F3E90.1BD5@stanford.edu> <see-below-1805980441570001@209.24.240.12> <*johnnyc*-2005981028370001@jchristie.halifaxcable.dal.ca> <Et9FLq.J6F@haddock.cd.chalmers.se> In article <Et9FLq.J6F@haddock.cd.chalmers.se>, sorry@no.more.spams wrote: > In <*johnnyc*-2005981028370001@jchristie.halifaxcable.dal.ca> John Christie > wrote: > >> More simply put, the two programs can no longer directly access the same > >> address space, right? > > Yes, and YAY! Anyone who really wants this terribly unsafe way of > > moving information around inside OS X raise their hand now. > > This is actually a good way of passing information around. You just need to > be a bit careful about it... :) > > For example the Unix call "fork" gains a lot of performance from being > implemented with shared memory -- copy on write shared memory, mind you. Apple's claim that you won't be able to directly pass pointers between programs and jumping to the conclusion that they won't implement available Unix services doesn't necessarily follow. If Apple doesn't, a third party will implement shared memory functionaliy through messaging. However, it won't work the current way and Apple is telling everyone to avoid the current, and dangerous method. They want everyone to write for Carbon and that isn't here yet so the best they can do is tell everyone what not to do with what they have been doing. If you look carefully at the Carbon Paper there is a tremendous amount of functionality already described at WWDC that isn't even mentioned as being available. Mostly all that paper says is how to use what you have in a way that will work with Carbon when it is here, what will work, and what won't. It does not relate much about the new functionality (like GX text and new QD) and only talks about a few of the new APIs briefly (like Nav Services). -- You aren't free if you CAN choose - only if you DO choose. All you are is the decisions you make. Remove "*" and ohnny (i.e. jc@) to reply via email
From: dbk@mcs.com (Dan "Bud" Keith) Newsgroups: comp.sys.next.programmer Subject: Why no g++ 2.8/libg++ under Rhapsody? What will it take? Date: Wed, 20 May 1998 13:27:45 -0500 Organization: Emergent Systems Message-ID: <1d9bwd7.938m751b8p48aN@dbklap.bb.opentext.com> I need to port a significant amount of C++ code to Rhapsody DR2 (please don't tell me that I should use Objective C, thank you) and the Apple-supplied version of gcc is 2.7.2.1, which has inadequate template support (I can't even compile deque.h!!!). I have heard that there is a g++ version 2.8 that supports templates and a libg++ that has the SGI STL. I was wondering what is entailed in building this for Rhapsody. Is it a futile task? I heard that Apple has proprietary modifications to support the Objective-C runtime? Are these modification available to the public? Doesn't the GPL require such availability? Metrowerks appears to have dropped the ball on delivering C++ in the short term, and I'd like to have my software porting underway before CR1 of Rhapsody. Does anyone else share these problems with me? thanks for any info, bud ----- Dan Keith (dbk@mcs.com) -----
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: My _first_ reboot! Date: 20 May 1998 13:58:40 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6junig$is5$2@ns3.vrx.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Today I had to reboot my OpenStep box. This is newsworthy indeed, because I haven't had to do this since starting here four months ago. And it wasn't even a crash, I ran out of swap space and couldn't be bothered to wait for it to come back to life. OpenStep rocks. Maury
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: 20 May 1998 15:34:54 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6jvb8u$2do$1@crib.bevc.blacksburg.va.us> References: <6junig$is5$2@ns3.vrx.net> In article <6junig$is5$2@ns3.vrx.net>, maury@remove_this.istar.ca (Maury Markowitz) wrote: > Today I had to reboot my OpenStep box. This is newsworthy indeed, because > I haven't had to do this since starting here four months ago. And it wasn't > even a crash, I ran out of swap space and couldn't be bothered to wait for it > to come back to life. And, of course, the eternally-growing swapfile problem will be removed with Mach 3.0. For some reason I seem to have really bad luck with OpenStep, though. :( The whole system seems to wedge every month or two, can't even telnet in. I've never been able to figure out what causes it, it's usually something completely innocuous like clicking on an app's window. Other people don't seem to have this problem.
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: 20 May 1998 14:45:23 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6juqa3$kv2$1@ns3.vrx.net> References: <6junig$is5$2@ns3.vrx.net> <6jvb8u$2do$1@crib.bevc.blacksburg.va.us> <6jvbhn$2f3$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: nurban@crib.bevc.blacksburg.va.us In <6jvbhn$2f3$1@crib.bevc.blacksburg.va.us> Nathan Urban claimed: > Of course, I should also point out that some people would still kill > for an operating system that only crashes once every month or two. :) You kidding? I wish my last NT box would have crashed less than twice a WEEK. Maury
From: Bill Bumgarner <bbum@codefab.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Wed, 20 May 1998 16:16:05 -0400 Organization: Spacelab.net Internet Access Message-ID: <Pine.NXT.3.96.980520161114.26221R-100000@pathos> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> <joe.ragosta-1605982005160001@elk32.dol.net> <355E3F66.86DD9F0@thegrid.net> <355EE31C.7834F8EA@milestonerdl.com> <joe.ragosta-1705981730440001@elk85.dol.net> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII To: Joe Ragosta <joe.ragosta@dol.net> In-Reply-To: <joe.ragosta-1705981730440001@elk85.dol.net> On Sun, 17 May 1998, Joe Ragosta wrote: > In article <355EE31C.7834F8EA@milestonerdl.com>, mark@milestonerdl.com wrote: > > > Ryan Kennedy wrote: > > > > > Joe Ragosta wrote: > > > > > > > But if they've left out the Unix from MacOS X, it could be smaller. > > > > > So Ragosta, where's this 'larger' FUD of yours coming from? > > Read what I said. > > I said that MacOSX _without_ Unix would be smaller than MacOSX _with_ > Unix. Unless you've found a way for Unix to contain a negative number of > bytes, my statement is clearly correct. But MacOSX IS UNIX. Period. There is NOT an option to have MacOSX WITHOUT UNIX. Now, certainly, Apple could have a minimal MacOSX that doesn't have a lot of the Unix command-line stuff... or could ship a MacOSX without a lot of the Unix APIs... but IT WILL BE UNIX. Actually, that isn't totally true. The Kernel IS NOT unix. The kernel is Mach. On top of the kernel, you will find a set of POSIX compliant 4.4 BSD APIs. So, if every MacOSX configuration is to be posix compliant [a good thing, btw], then that will basically determine the minimum size of the OS image. Now, on TOP of the 4.4 BSD APIs, you have a series of command line tools; ls, rm, ln, etc.etc.etc.etc.... these are the optional bits that are not guaranteed to be present in the system. So, the question is, how much of the POSIX compliant APIs will be guaranteed to be present? Answer that, and you are a long way towards answering how small MacOSX could possibly be. b.bum
From: Bill Bumgarner <bbum@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: Wed, 20 May 1998 16:26:38 -0400 Organization: Spacelab.net Internet Access Message-ID: <Pine.NXT.3.96.980520161846.26221T-100000@pathos> References: <6junig$is5$2@ns3.vrx.net> <6jvb8u$2do$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII To: nurban@vt.edu In-Reply-To: <6jvb8u$2do$1@crib.bevc.blacksburg.va.us> There are a couple of factors that can contribute: If it is an Intel box, it may just be hardware weirdness.... I have seen Intel boxes crash when a Taxi passes our office and is broadcasting on an overpowered CB. I have other Intel boxes that ARE NOT stable with hte case open. Others that crash when you look at 'em and yet others that are stable. Definitely drop premium memory in the box-- it is a few $$$ more expensive, but well worth it in terms of stability. If it is a NeXT box, it could be memory that is on the verge of flaking out.... Symptoms for sporadic crashes like this are usually very random-- basically, something has thrashed memory in an unpredictable way. By clicking on a Window is not really that innocuous-- it often involves paging in a bunch of memory... if you HAPPEN to have a SIMM that HAPPENs to very occasionally blow a bit, then that paging may WRITE a bad bit out to disk only to page in later or may READ something into memory that is corrupted. Interesting to note that on Intel, Operating Systems that use more effecient paging algorithms seem to be LESS stable in terms of triggering funky memory problems because they typically cause MORE paging activity in that: - the user can usually get away with running way beyond physical memory because the swapping subsystem is fast - because I/O is optimized, data moves from drive <-> memory in a very fast, optimized manner - because the paging is optimized it is typically not the case that you are moving many consecutivie pages, but may be moving lots of discontiguous pages very quickly... a much noisier sequence of events. If memory errors out due to cross talk [which seems to frequently be the case], then there will be more chance of failure due to the relatively chaotic nature of a good pager vs. the relatively linear behaviour of a pager that tends to swap out whole apps or sets of contiguous pages [like NT-- which exhibits the yucky behviouar of swapping in/out WHOLE processes worth of memory]. At least, that is what I have observed from dealing with many, many machines across many, many operating systems. b.bum On 20 May 1998, Nathan Urban wrote: > In article <6junig$is5$2@ns3.vrx.net>, maury@remove_this.istar.ca (Maury Markowitz) wrote: > > > Today I had to reboot my OpenStep box. This is newsworthy indeed, because > > I haven't had to do this since starting here four months ago. And it wasn't > > even a crash, I ran out of swap space and couldn't be bothered to wait for it > > to come back to life. > > And, of course, the eternally-growing swapfile problem will be removed > with Mach 3.0. > > For some reason I seem to have really bad luck with OpenStep, though. :( > The whole system seems to wedge every month or two, can't even telnet in. > I've never been able to figure out what causes it, it's usually something > completely innocuous like clicking on an app's window. Other people > don't seem to have this problem. > >
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: 20 May 1998 16:43:52 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6jvfa8$2pj$1@crib.bevc.blacksburg.va.us> References: <6junig$is5$2@ns3.vrx.net> <6jvb8u$2do$1@crib.bevc.blacksburg.va.us> <Pine.NXT.3.96.980520161846.26221T-100000@pathos> In article <Pine.NXT.3.96.980520161846.26221T-100000@pathos>, Bill Bumgarner <bbum@codefab.com> wrote: > On 20 May 1998, Nathan Urban wrote: > > For some reason I seem to have really bad luck with OpenStep, though. :( > > The whole system seems to wedge every month or two, can't even telnet in. > There are a couple of factors that can contribute: > If it is an Intel box, it may just be hardware weirdness.... ... and probably is. That's the best guess I can come up with, at least, since whatever triggers it seems to be completely unpredictable. > Definitely drop premium memory in the box-- it is a few $$$ more > expensive, but well worth it in terms of stability. Already tried that.. got a memory upgrade, found out that it wasn't "compatible enough" with the existing RAM and was causing all kinds of crashes, so replaced the whole lot.. though that still doesn't rule out memory problems. > Interesting to note that on Intel, Operating Systems that use more > effecient paging algorithms seem to be LESS stable in terms of triggering > funky memory problems because they typically cause MORE paging activity Does this tend to afflict non-Intel systems too? If not, why not? > [a bunch of interesting reasons] > If memory errors out due to cross talk [which seems to frequently be the > case], then there will be more chance of failure due to the relatively > chaotic nature of a good pager vs. the relatively linear behaviour of a > pager that tends to swap out whole apps or sets of contiguous pages [like > NT-- which exhibits the yucky behviouar of swapping in/out WHOLE processes > worth of memory]. You're joking, right? Don't tell me that NT actually implements whole-process _swapping_ over paging! I knew that NT VM performance was bad, but not _that_ bad! This would explain much.. Thanks for a fascinating discussion of the oddities of VM-induced instabilities, by the way..
From: lars.farm@ite.mh.se (Lars Farm) Newsgroups: comp.sys.next.programmer Subject: Re: Why no g++ 2.8/libg++ under Rhapsody? What will it take? Date: Wed, 20 May 1998 22:58:50 +0200 Organization: pv Message-ID: <1d9cmmp.110myoj1wkowieN@dialup168-2-56.swipnet.se> References: <1d9bwd7.938m751b8p48aN@dbklap.bb.opentext.com> Cache-Post-Path: mn8!s-49817@dialup168-2-56.swipnet.se Dan "Bud" Keith <dbk@mcs.com> wrote: > Metrowerks appears to have dropped the ball on delivering C++ in the > short term, and I'd like to have my software porting underway before CR1 > of Rhapsody. Does anyone else share these problems with me? Yes. -- Lars Farm; lars.farm@ite.mh.se - Limt/Channelmatic: lars.farm@limt.se
From: John Hornkvist <sorry@no.more.spams> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 20 May 1998 18:17:57 GMT Organization: Chalmers Tekniska Högskola Sender: john@haddock.cd.chalmers.se (John Hprnkvist) Message-ID: <Et9pHx.JI8@haddock.cd.chalmers.se> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <6jn336$5kk@basement.replay.com> <355F3E90.1BD5@stanford.edu> <see-below-1805980441570001@209.24.240.12> <*johnnyc*-2005981028370001@jchristie.halifaxcable.dal.ca> <Et9FLq.J6F@haddock.cd.chalmers.se> <*johnnyc*-2005981459020001@jchristie.halifaxcable.dal.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: *johnnyc*@or.psychology.dal.ca In <*johnnyc*-2005981459020001@jchristie.halifaxcable.dal.ca> John Christie wrote: > In article <Et9FLq.J6F@haddock.cd.chalmers.se>, sorry@no.more.spams wrote: > > This is actually a good way of passing information around. You just need to > > be a bit careful about it... :) > > > > For example the Unix call "fork" gains a lot of performance from being > > implemented with shared memory -- copy on write shared memory, mind you. > > Apple's claim that you won't be able to directly pass pointers between > programs and jumping to the conclusion that they won't implement available > Unix services doesn't necessarily follow. If Apple doesn't, a third party > will implement shared memory functionaliy through messaging. However, it > won't work the current way and Apple is telling everyone to avoid the > current, and dangerous method. They want everyone to write for Carbon and > that isn't here yet so the best they can do is tell everyone what not to > do with what they have been doing. Oh, I'm not the least concerned with message passing or shared memory in Mac OS X. I've never programmed the toolbox or any other MacOS API, and I don't think I ever will, either. I have been programming NEXTSTEP and OPENSTEP for the last four years, and expect to keep on doing that for years. At least as a hobby. :) My point was that sharing memory between applications is good, it is just a matter of sharing the right stuff. MacOS currently shares everything, which is not a very good idea -- at least with current compilers, operating systems and programming languages*. With a reasonable operating system, like Mach, you can share memory responsibly. Luckily, with the NeXT purchase Apple got one of the best kernel minds in the business, and he is in a position to call the shots. * Some may argue that memory protection is compiler and language issue, and does not belong in operating systems and processors. That would certainly be cool -- if they can get it to work, which I wouldn't bet the house on... Regards, John Hornkvist Name: nhoj Address: cd.chalmers.se
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: 20 May 1998 16:42:11 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6jv153$pib$1@ns3.vrx.net> References: <6junig$is5$2@ns3.vrx.net> <6jveu0$klt$1@news.idiom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: jcr.remove@this.phrase.idiom.com In <6jveu0$klt$1@news.idiom.com> John C. Randolph claimed: > Log out, and type "exit" in the login window (no password.) The login window > will go away for a few seconds, and then reappear. This should allow Mach to > reclaim the vm that the window server was using. Ahhhh, I had assumed a logout would do that. I'll do this from now on. Maury
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: 20 May 1998 15:39:35 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6jvbhn$2f3$1@crib.bevc.blacksburg.va.us> References: <6junig$is5$2@ns3.vrx.net> <6jvb8u$2do$1@crib.bevc.blacksburg.va.us> In article <6jvb8u$2do$1@crib.bevc.blacksburg.va.us>, nurban@vt.edu wrote: > For some reason I seem to have really bad luck with OpenStep, though. :( > The whole system seems to wedge every month or two, can't even telnet in. Of course, I should also point out that some people would still kill for an operating system that only crashes once every month or two. :) (Of 24/7 operation, let me point out, though it's punctuated by periodic reboots every couple of days to a week or so whenever OmniWeb or a few other apps gobble up my swap space.)
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Objective C++ (was Re: Infoworld on the Yellow Box's future) Date: 20 May 1998 15:53:40 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6jvcc4$2i5$1@crib.bevc.blacksburg.va.us> References: <nospampischke-1905981902030001@pm2-18-2.tor.idirect.com> <nospampischke-1905982350420001@ts7-32t-16.idirect.com> <35626D26.B7616464@alum.mit.edu> <35633314.0@206.25.228.5> In article <35633314.0@206.25.228.5>, jkheit@xtdl.com wrote: > Anyone know for sure? I read the Obj-C++ thing on one of Scott's > WWDC note pages at www.stepwise.com Yes, does anyone know in detail what changes they are making in the Obj-C/Obj-C++ language? I've been wondering about that. And are they all things to improve C++ integration, or are there new Obj-C features too?
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: 20 May 1998 20:37:20 GMT Organization: Idiom Communications Message-ID: <6jveu0$klt$1@news.idiom.com> References: <6junig$is5$2@ns3.vrx.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: maury@remove_this.istar.ca Maury Markowitz may or may not have said: -> Today I had to reboot my OpenStep box. This is newsworthy indeed, because -> I haven't had to do this since starting here four months ago. And it wasn't -> even a crash, I ran out of swap space and couldn't be bothered to wait for it -> to come back to life. -> -> OpenStep rocks. -> -> Maury Suggestion: Restart your window server once in a while. Log out, and type "exit" in the login window (no password.) The login window will go away for a few seconds, and then reappear. This should allow Mach to reclaim the vm that the window server was using. Many user apps are able to cause the window server to leak memory, usually by allocating NSImages and not removing them properly before exiting. -jcr
From: spammers@ruin.the.internet.channelu.com Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: 20 May 1998 23:17:39 GMT Organization: Michigan State University Message-ID: <6jvoaj$72f$3@msunews.cl.msu.edu> References: <6junig$is5$2@ns3.vrx.net> <6jvb8u$2do$1@crib.bevc.blacksburg.va.us> <Pine.NXT.3.96.980520161846.26221T-100000@pathos> <6jvfa8$2pj$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: nurban@crib.bevc.blacksburg.va.us In <6jvfa8$2pj$1@crib.bevc.blacksburg.va.us> Nathan Urban wrote: > In article <Pine.NXT.3.96.980520161846.26221T-100000@pathos>, Bill Bumgarner <bbum@codefab.com> wrote: > > > On 20 May 1998, Nathan Urban wrote: > > > > For some reason I seem to have really bad luck with OpenStep, though. :( > > > The whole system seems to wedge every month or two, can't even telnet in. > > > There are a couple of factors that can contribute: > > > If it is an Intel box, it may just be hardware weirdness.... > > ... and probably is. That's the best guess I can come up with, at least, > since whatever triggers it seems to be completely unpredictable. Yeah could be anything. On my primary box I had done some hardware & software swapping and started getting wedges nearly every time I ran a new version of RadicalNews. Took me a day to pin it down to RAM. I examined the pads and they were heavily scratched. Thing is I'm sure they wern't when I got them. I examined carefully and it appeared someone went over the tin leads with a slight amount of tin solder (my suspicion). First time they went it they were fine, after that who can say.. Heat can cause problems too. Vibration also especially if you have bad connectors. Check all your connectors make sure you don't have corrosion, and that they seat tightly. Beyond that its a matter of having replacements to swap in/out for everything to try to isolate. > > Definitely drop premium memory in the box-- it is a few $$$ more > > expensive, but well worth it in terms of stability. > > Already tried that.. got a memory upgrade, found out that it wasn't > "compatible enough" with the existing RAM and was causing all kinds of > crashes, so replaced the whole lot.. though that still doesn't rule > out memory problems. You should also be able to turn down the RAM speed/timings this might help. And yes it doesn't rule out memory problems. BTW: Once I found it my Openstep box (a P-Pro 200 OC'd to 225=75x3) is rock solid. Though I don't leave it on all the time. My Indigo2 & Turbo NeXT Cube (3.3 BTW). Now those are ROCK solid over months and now 2 years. I should try it just to see if 4.2 on my Pro is just as solid. Randy rencsok at channelu dot com argus dot cem dot msu dot edu spammers works also :) Randy Rencsok General UNIX, NeXTStep, IRIX Admining, Turbo Software Consulting, Programming, etc.)
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: 20 May 1998 23:30:53 GMT Organization: Idiom Communications Message-ID: <6jvp3d$62e$2@news.idiom.com> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> <joe.ragosta-1605982005160001@elk32.dol.net> <355E3F66.86DD9F0@thegrid.net> <355EE31C.7834F8EA@milestonerdl.com> <joe.ragosta-1705981730440001@elk85.dol.net> <Pine.NXT.3.96.980520161114.26221R-100000@pathos> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: bbum@codefab.com Bill Bumgarner may or may not have said: -> -> On Sun, 17 May 1998, Joe Ragosta wrote: -> -> > In article <355EE31C.7834F8EA@milestonerdl.com>, mark@milestonerdl.com wrote: -> > -> > > Ryan Kennedy wrote: -> > > -> > > > Joe Ragosta wrote: -> > > > -> > > > > But if they've left out the Unix from MacOS X, it could be smaller. -> > > > -> > > So Ragosta, where's this 'larger' FUD of yours coming from? -> > -> > Read what I said. -> > -> > I said that MacOSX _without_ Unix would be smaller than MacOSX _with_ -> > Unix. Unless you've found a way for Unix to contain a negative number of -> > bytes, my statement is clearly correct. -> -> But MacOSX IS UNIX. Period. There is NOT an option to have MacOSX -> WITHOUT UNIX. -> -> Now, certainly, Apple could have a minimal MacOSX that doesn't have a lot -> of the Unix command-line stuff... or could ship a MacOSX without a lot of -> the Unix APIs... but IT WILL BE UNIX. -> -> Actually, that isn't totally true. -> -> The Kernel IS NOT unix. The kernel is Mach. On top of the kernel, you -> will find a set of POSIX compliant 4.4 BSD APIs. So, if every MacOSX -> configuration is to be posix compliant [a good thing, btw], then that will -> basically determine the minimum size of the OS image. -> -> Now, on TOP of the 4.4 BSD APIs, you have a series of command line tools; -> ls, rm, ln, etc.etc.etc.etc.... these are the optional bits that are not -> guaranteed to be present in the system. Of course, if you take a bunch of those simple, common UNIX programs like sed and the like out of the product, then you'll have a needlessly crippled system. /bin on my OpenStep 4.2 system is 4.2 megs. Thats about four cents worth of storage. The UNIX /bin directory on Rhapsody is similar. It ain't broke, but if Apple tries to "fix" it by taking out a bunch of these tools, it will be a hell of a lot of work. -jcr
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: 21 May 1998 00:30:44 GMT Organization: MiscKit Development Message-ID: <6jvsjk$5qh$1@news.xmission.com> References: <6junig$is5$2@ns3.vrx.net> <6jvb8u$2do$1@crib.bevc.blacksburg.va.us> nurban@crib.bevc.blacksburg.va.us (Nathan Urban) wrote: > For some reason I seem to have really bad luck with OpenStep, though. :( > The whole system seems to wedge every month or two, can't even telnet in. > I've never been able to figure out what causes it, it's usually something > completely innocuous like clicking on an app's window. Other people > don't seem to have this problem. Could be a minor hardware problem. Intel HW is full of those. :-/ I had a similar problem, where everything would freeze, on one of my Intel boxes. It got annoying, because it would tend to happen only a short while after the computer was brought up. So I opened up the box and discovered that the problem was that the CPU fan's wires had wiggled their way loose and so the fan wasn't running. That explained everything--the machine was locking up because the CPU was going into thermal shutdown. Certainly not OPENSTEP's fault! Once I fixed it, the machine became solid as a rock. Damn Intel processors that draw more power than a Christmas tree...and run hot enough to cook dinner on... -- Later, -Don Yacktman don@misckit.com <a href="http://www.misckit.com/don.html">My home page</a>
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Objective C++ (was Re: Infoworld on the Yellow Box's future) Date: 21 May 1998 03:40:52 GMT Organization: Digital Fix Development Message-ID: <6k07o4$gc0$1@news.digifix.com> References: <nospampischke-1905981902030001@pm2-18-2.tor.idirect.com> <nospampischke-1905982350420001@ts7-32t-16.idirect.com> <35626D26.B7616464@alum.mit.edu> <35633314.0@206.25.228.5> <6jvcc4$2i5$1@crib.bevc.blacksburg.va.us> In-Reply-To: <6jvcc4$2i5$1@crib.bevc.blacksburg.va.us> On 05/20/98, Nathan Urban wrote: >In article <35633314.0@206.25.228.5>, jkheit@xtdl.com wrote: > >> Anyone know for sure? I read the Obj-C++ thing on one of Scott's >> WWDC note pages at www.stepwise.com > >Yes, does anyone know in detail what changes they are making in the >Obj-C/Obj-C++ language? I've been wondering about that. And are they all >things to improve C++ integration, or are there new Obj-C features too? > I can answer that. The changes are NOT the syntax, rather they are adding some of the "other stuff". Features like declaration as statement and exceptions. They syntax issue is resolved. The traditional Objective-C syntax is staying. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Objective C++ (was Re: Infoworld on the Yellow Box's future) Date: 21 May 1998 00:06:58 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6k0992$3mk$1@crib.bevc.blacksburg.va.us> References: <nospampischke-1905981902030001@pm2-18-2.tor.idirect.com> <35633314.0@206.25.228.5> <6jvcc4$2i5$1@crib.bevc.blacksburg.va.us> <6k07o4$gc0$1@news.digifix.com> In article <6k07o4$gc0$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: > On 05/20/98, Nathan Urban wrote: > The changes are NOT the syntax, rather they are adding some of > the "other stuff". I figured they'd leave the syntax, after their previous aborted attempt, but wanted to be sure. > Features like declaration as statement and exceptions. What is "declaration as statement"? (I'm sure I know what it is, just not under that name.) Any improvements to the Obj-C object model?
From: arti@lava.DOTnet (Art Isbell - remove "DOT") Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: 21 May 1998 05:57:43 GMT Organization: LavaNet, Inc. Distribution: world Message-ID: <6k0fon$5pr@mochi.lava.net> References: <6junig$is5$2@ns3.vrx.net> <6jvb8u$2do$1@crib.bevc.blacksburg.va.us> <Pine.NXT.3.96.980520161846.26221T-100000@pathos> <6jvfa8$2pj$1@crib.bevc.blacksburg.va.us> nurban@crib.bevc.blacksburg.va.us (Nathan Urban) wrote: > You're joking, right? Don't tell me that NT actually implements > whole-process _swapping_ over paging! I knew that NT VM performance > was bad, but not _that_ bad! This would explain much.. Regardless of how NT implements VM, the disk seek noise and sluggishness when changing app focus seems incredible, especially considering that my PC has 96 MB of RAM. Then I installed Diskkeeper Lite, a disk defragmenter. It reported 54% fragmentation! fsck typically reports 0.3% on my Mach systems. Unfortunately, defragmentation risks disk corruption and data loss unless Service Pack 2 or greater has been installed which I understand breaks the samba that comes with OS/Mach 4.2. So backing up to tape, reformatting, and reinstalling is supposed to be a time-consuming alternative. Did that - still 32% fragmentation despite almost 600 MB of free space! Does NTFS implement any kind of anti-fragmentation algorithms?? Doesn't seem like it to me. And remember, folks, NT is Microsoft's high-performance operating system :-) -- Art Isbell NeXT/MIME Mail: arti@lavaDOTnet IDX Systems Corporation Voice/Fax: +1 808 394 0511 (for whom I don't speak) Voice Mail: +1 808 394 0495 Healthcare Info Technology US Mail: Honolulu, HI 96825-2638
From: mpaque@wco.com (Mike Paquette) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Wed, 20 May 1998 23:10:32 -0700 Organization: Electronics Service Unit No. 16 Message-ID: <1d9bz7g.119xbtvbb77k5N@cetus160.wco.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <6j7sdr$4gn$1@newsfep4.sprintmail.com> <6j7tij$vb7$1@crib.bevc.blacksburg.va.us> <stevenj-ya02408000R1105982039000001@news.mit.edu> <andyba-ya02408000R1205980916210001@news> <355873F0.6B86BC5C@milestonerdl.com> <andyba-ya02408000R1405981025210001@news> <01bd80f4$85082140$04387880@test1> <6jkuar$39a$1@xmission.xmission.com> <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> <joe.ragosta-1605982005160001@elk32.dol.net> <355E3F66.86DD9F0@thegrid.net> <355EE31C.7834F8EA@milestonerdl.com> <joe.ragosta-1705981730440001@elk85.dol.net> <Pine.NXT.3.96.980520161114.26221R-100000@pathos> Bill Bumgarner <bbum@codefab.com> wrote: > The Kernel IS NOT unix. The kernel is Mach. On top of the kernel, you > will find a set of POSIX compliant 4.4 BSD APIs. So, if every MacOSX > configuration is to be posix compliant [a good thing, btw], then that will > basically determine the minimum size of the OS image. > > So, the question is, how much of the POSIX compliant APIs will be > guaranteed to be present? > > Answer that, and you are a long way towards answering how small MacOSX > could possibly be. Well, one could pull out the POSIX subsystem (BSD et al) and save about 70 Kbytes of code, plus memory used in support of the implementation (I/O buffers, assorted caches, etc). This would slightly reduce the disk footprint and shave a bit off the minimum memory requirement. Of course, how one would do things like file I/O, networking, user input and GUI display becomes problematic, and would be an exercise left to the user. (A naked Mach microkernel does very little that would be interesting to a typical desktop computer user.)
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: 21 May 1998 02:24:19 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6k0haj$41v$1@crib.bevc.blacksburg.va.us> References: <6junig$is5$2@ns3.vrx.net> <Pine.NXT.3.96.980520161846.26221T-100000@pathos> <6jvfa8$2pj$1@crib.bevc.blacksburg.va.us> <6k0fon$5pr@mochi.lava.net> In article <6k0fon$5pr@mochi.lava.net>, arti@lava.DOTnet (Art Isbell - remove "DOT") wrote: > nurban@crib.bevc.blacksburg.va.us (Nathan Urban) wrote: > > You're joking, right? Don't tell me that NT actually implements > > whole-process _swapping_ over paging! I knew that NT VM performance > > was bad, but not _that_ bad! This would explain much.. > Regardless of how NT implements VM, the disk seek noise and sluggishness > when changing app focus seems incredible, especially considering that my PC > has 96 MB of RAM. Oh, this is without doubt. I've got a 64 MB system at work and the disk is _constantly_ grinding during various tasks.. and when I'm in certain phases of compiling, it's sometimes all but impossible to even switch apps! (Points to not just VM problems, but scheduler problems too.) I don't know _how_ they could have implemented such a braindead system, but my OPENSTEP for Mach box at home (on a slower system) blows NT out of the water in terms of interactive performance. (I'm sure that the window server saving context switches and the universal double buffering have a lot to do with that too, though. But the NT VM and scheduler seem to be really bad.)
From: paul@mbed.com (Paul Hamilton) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Thu, 21 May 1998 01:15:29 -0800 Organization: mBED Software Message-ID: <paul-2105980115290001@206.169.118.19> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <6jn336$5kk@basement.replay.com> <355F3E90.1BD5@stanford.edu> <see-below-1805980441570001@209.24.240.12> <*johnnyc*-2005981028370001@jchristie.halifaxcable.dal.ca> In article <*johnnyc*-2005981028370001@jchristie.halifaxcable.dal.ca>, *johnnyc*@or.psychology.dal.ca (John Christie) wrote: > In article <see-below-1805980441570001@209.24.240.12>, > see-below@not-my-address.com (Matthew Vaughan) wrote: > > > More simply put, the two programs can no longer directly access the same > > address space, right? > > Yes, and YAY! Anyone who really wants this terribly unsafe way of > moving information around inside OS X raise their hand now. > v | \O | /\ Paul. -- Paul Hamilton Software Engineer mBED Software paul@mbed.com http://www.mbed.com
From: paul@mbed.com (Paul Hamilton) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Thu, 21 May 1998 02:23:09 -0800 Organization: mBED Software Message-ID: <paul-2105980223100001@206.169.118.19> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> <joe.ragosta-1605982005160001@elk32.dol.net> <355E3F66.86DD9F0@thegrid.net> <355EE31C.7834F8EA@milestonerdl.com> <joe.ragosta-1705981730440001@elk85.dol.net> <Pine.NXT.3.96.980520161114.26221R-100000@pathos> In article <Pine.NXT.3.96.980520161114.26221R-100000@pathos>, Bill Bumgarner <bbum@codefab.com> wrote: > On Sun, 17 May 1998, Joe Ragosta wrote: > > > In article <355EE31C.7834F8EA@milestonerdl.com>, mark@milestonerdl.com wrote: > > > > > Ryan Kennedy wrote: > > > > > > > Joe Ragosta wrote: > > > > > > > > > But if they've left out the Unix from MacOS X, it could be smaller. > > > > > > > So Ragosta, where's this 'larger' FUD of yours coming from? > > > > Read what I said. > > > > I said that MacOSX _without_ Unix would be smaller than MacOSX _with_ > > Unix. Unless you've found a way for Unix to contain a negative number of > > bytes, my statement is clearly correct. > > But MacOSX IS UNIX. Period. There is NOT an option to have MacOSX > WITHOUT UNIX. > > Now, certainly, Apple could have a minimal MacOSX that doesn't have a lot > of the Unix command-line stuff... or could ship a MacOSX without a lot of > the Unix APIs... but IT WILL BE UNIX. > Right. Does it matter if it's BIG? Does "consumer level" mean small? Not on my powerbook. The system folder is about 60MB of stuff. Add to this all the applications, and you get up to about 200MB of CORE stuff. I know I'm not a consumer, but remember that on Windows consumer means Office 98, and that sucks about 100MB of your hard drive! So "consumer level" doesn't mean "low disk footprint". Does it mean "low memory footprint"? No, most machines from Apple come with 32MB of RAM (even more). This is heaps for a real OS with decent virtual memory. We will save space and RAM by moving towards a more unified and powerful OS Kernel (like Mach). We won't gain disk footprint or RAM footprint. "Consumer level" is all about how cheap you can make the boxes I think. Paul. -- Paul Hamilton Software Engineer mBED Software paul@mbed.com http://www.mbed.com
From: *johnnyc*@or.psychology.dal.ca (John Christie) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Thu, 21 May 1998 10:31:07 -0300 Organization: ISINet, Nova Scotia Message-ID: <*johnnyc*-2105981031080001@jchristie.halifaxcable.dal.ca> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <B17CCA7D-359B7@206.165.43.153> <6j9ve8$g6d$1@news01.deltanet.com> <35596510.F5627440@alumni.caltech.edu> <mem-1305981347230001@surfer.pha.jhu.edu> In article <mem-1305981347230001@surfer.pha.jhu.edu>, mem@jhu.edu (Mel Martinez) wrote: > In article <35596510.F5627440@alumni.caltech.edu>, William Edward Woody > <woody@alumni.caltech.edu> wrote: > > If you write for Carbon, it's a recompile for MacOS 8. So for > > systems that cannot run Carbon apps, just follow the guidelines > > Hmm... it may be even simpler than that. Unless I'm missing something, > shouldn't the Carbon ABI (note the 'B' as in Application _Binary_ > Interface) be a (well behaved) subset of the mac OS 8 PPC ABI? In other > words, it's not clear to me that one should have to recompile a properly > written Carbon API application to run on both Mac OS X and Mac OS 8.x > (with Carbon plug-in) on PPC. Yes, if you compiled for Carbon then you run in both automatically. However, if you compiled with the OS8 versions of the same APIs then you would need to -- You aren't free if you CAN choose - only if you DO choose. All you are is the decisions you make. Remove "*" and ohnny (i.e. jc@) to reply via email
From: John Hornkvist <sorry@no.more.spams> Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: Thu, 21 May 1998 11:02:53 GMT Organization: Chalmers Tekniska Högskola Sender: john@haddock.cd.chalmers.se (John Hprnkvist) Message-ID: <EtB00t.52L@haddock.cd.chalmers.se> References: <6junig$is5$2@ns3.vrx.net> <Pine.NXT.3.96.980520161846.26221T-100000@pathos> <6jvfa8$2pj$1@crib.bevc.blacksburg.va.us> <6k0fon$5pr@mochi.lava.net> <6k0haj$41v$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: nurban@crib.bevc.blacksburg.va.us In <6k0haj$41v$1@crib.bevc.blacksburg.va.us> Nathan Urban wrote: > Oh, this is without doubt. I've got a 64 MB system at work and the > disk is _constantly_ grinding during various tasks.. and when I'm in > certain phases of compiling, it's sometimes all but impossible to even > switch apps! (Points to not just VM problems, but scheduler problems > too.) I don't know _how_ they could have implemented such a braindead > system, but my OPENSTEP for Mach box at home (on a slower system) blows > NT out of the water in terms of interactive performance. > > (I'm sure that the window server saving context switches and the universal > double buffering have a lot to do with that too, though. But the NT VM > and scheduler seem to be really bad.) IIRC, NT uses a VMS-like virtual memory system "adopted" to workstations. Primary paging is not system wide, but "private" to each process. A process has a FIFO of pages, and a second chance queue. When a process starts to generate a lot of page faults, more pages are allocated to it, i.e. the FIFO for that process gets longer. As I understand it, the VM system inherits "features" from minis where the system administrator allocates pages to users. In NT the allocation is among processes instead of among users, and the kernel is responsible for the queue sizes. Obvious problems with this approach is that the VM system reacts slowly when the work load changes, and that it is always more inefficient to have many queues than to have one. I like random replacement, because there is not worst case -- possibly coupled with a clock hand algorithm to copy pages to disk during moments when the workload is light. Regards, John Hornkvist Name: nhoj Address: cd.chalmers.se
Subject: Re: My _first_ reboot! From: "Robert A. Decker" <comrade@umich.edu> Newsgroups: comp.sys.next.programmer References: <6jvb8u$2do$1@crib.bevc.blacksburg.va.us> MIME-Version: 1.0 Message-ID: <B189C84A-3C2CE@141.214.128.36> Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Date: Thu, 21 May 1998 15:58:19 GMT NNTP-Posting-Date: Thu, 21 May 1998 11:58:19 EDT On Wed, May 20, 1998 2:34 PM, Nathan Urban <mailto:nurban@crib.bevc.blacksburg.va.us> wrote: > For some reason I seem to have really bad luck with OpenStep, though. :( > Mine goes down about twice a month. It's always OmniWeb that does it. The entire machine locks up and I can't even telnet into it... rob -- <mailto: "Robert A. Decker" comrade@umich.edu> <http://hmrl.cancer.med.umich.edu/Rob/index.ssi> Programmer Analyst - Health Media Research Lab University of Michigan Comprehensive Cancer Center "Get A Life" quote #2: "Has anyone seen Chris? I have some last minute instructions for the scene where he wrestles the evil monkey." -Get a Life
From: Andre-John Mas <ama@fabre.act.qc.ca> Newsgroups: comp.sys.next.programmer Subject: [Q] So will Rhapsody be released? Date: Wed, 20 May 1998 08:11:00 -0500 Organization: Communications Accessibles Montreal, Quebec Canada Message-ID: <Pine.LNX.3.95.980520080725.10572B-100000@fabre.act.qc.ca> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Now that Apple has announced MacOS X (a fusion of Rhapsody and MacOS) will Rhapsody as we are using actually be released to the public, or will we have to wait for MacOS X before we see any elements of Rhapsody? OR, will Rhapsody as we know it be used as a development platform for releasing X-platform programs? Thanks AJ ---------------------------------------------------------- Andre-John Mas mailto:ama@act.qc.ca ----------------------------------------------------------
From: scott@leorg.ucdavis.edu (Ryan Scott) Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: 21 May 1998 17:22:51 GMT Organization: UC Davis - Dept. Applied Science Message-ID: <6k1ntb$dm5$1@mark.ucdavis.edu> References: <6junig$is5$2@ns3.vrx.net> <6jvb8u$2do$1@crib.bevc.blacksburg.va.us> <6jvsjk$5qh$1@news.xmission.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: don@misckit.com In <6jvsjk$5qh$1@news.xmission.com> Don Yacktman wrote: > I had a similar problem, where everything would freeze, on one of my Intel > boxes. It got annoying, because it would tend to happen only a short while > after the computer was brought up. So I opened up the box and discovered > that the problem was that the CPU fan's wires had wiggled their way loose and > so the fan wasn't running. That explained everything--the machine was > locking up because the CPU was going into thermal shutdown. Certainly not > OPENSTEP's fault! > > Once I fixed it, the machine became solid as a rock. > > Damn Intel processors that draw more power than a Christmas tree...and run > hot enough to cook dinner on... > > Someone sent this to me a couple of day ago. A paragraph later in the announcement explains why the Intel processors run so hot ;-) MICROSOFT TESTS NUCLEAR DEVICE AT SECRET HANFORD FACILITY REDMOND (BNN)--World leaders reacted with stunned silence as Microsoft Corp. (MSFT) conducted an underground nuclear test at a secret facility in eastern Washington state. The device, exploded at 9:22 am PDT (1622 GMT/12:22 pm EDT) today, was timed to coincide with talks between Microsoft and the US Department of Justice over possible antitrust action. "Microsoft is going to defend its right to market its products by any and all necessary means," said Microsoft CEO Bill Gates. "Not that I'm anti-government" he continued, "but there would be few tears shed in the computer industry if Washington were engulfed in a bath of nuclear fire." Scientists pegged the explosion at around 100 kilotons. "I nearly dropped my latte when I saw the seismometer" explained University of Washington geophysicist Dr. Whoops Blammover, "At first I thought it was Mt. Rainier, and I was thinking, damn, there goes the mountain bike vacation." In Washington, President Clinton announced the US Government would boycott all Microsoft products indefinitely. Minutes later, the President reversed his decision. "We've tried sanctions since lunchtime, and they don't work," said the President. Instead, the administration will initiate a policy of "constructive engagement" with Microsoft. Microsoft's Chief Technology Officer Nathan Myrhvold said the test justified Microsoft's recent acquisition of the Hanford Nuclear Reservation from the US Government. Not only did Microsoft acquire "kilograms of weapons grade plutonium" in the deal, said Myrhvold, "but we've finally found a place to dump those millions of unsold copies of Microsoft Bob." Myrhvold warned users not to replace Microsoft NT products with rival operating systems. "I can neither confirm nor deny the existence of a radioisotope thermoelectric generator inside of every Pentium II microprocessor," said Myrhvold, "but anyone who installs an OS written by a bunch of long-hairs on the Internet is going to get what they deserve." The existence of an RTG in each Pentium II microprocessor would explain why the microprocessors, made by the Intel Corporation, run so hot. The Intel chips "put out more heat than they draw in electrical power" said Prof. E. E. Thymes of MIT. "This should finally dispell those stories about cold fusion." Rumors suggest a second weapons development project is underway in California, headed by Microsoft rival Sun Microsystems. "They're doing all of the development work in Java," said one source close to the project. The development of a delivery system is said to be holding up progress. "Write once, bomb anywhere is still a dream at the moment." Meanwhile, in Cupertino, California, Apple interim-CEO Steve Jobs was rumored to be in discussion with Oracle CEO Larry Ellison about deploying Apple's Newton technology against Microsoft. "Newton was the biggest bomb the Valley has developed in years," said one hardware engineer. "I'd hate to be around when they drop that product a second time." posted on 14 May 1998 Copyright 1998 by the Bogus News Network. -- ________________________________________________ Ryan P. Scott Laser and Electro-Optics Research Group UC Davis - Department of Applied Science Tel: (530)754-4358 Fax: (530)752-1652 Email: scott@leorg.ucdavis.edu ________________________________________________
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: [Q] So will Rhapsody be released? Date: 21 May 1998 13:30:36 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6k1obs$677$1@crib.bevc.blacksburg.va.us> References: <Pine.LNX.3.95.980520080725.10572B-100000@fabre.act.qc.ca> In article <Pine.LNX.3.95.980520080725.10572B-100000@fabre.act.qc.ca>, Andre-John Mas <ama@fabre.act.qc.ca> wrote: > Now that Apple has announced MacOS X (a fusion of Rhapsody > and MacOS) will Rhapsody as we are using actually be > released to the public, or will we have to wait for MacOS X > before we see any elements of Rhapsody? Rhapsody 1.0 will ship this fall. But that will be the last release.
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: 21 May 1998 13:29:29 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6k1o9p$66j$1@crib.bevc.blacksburg.va.us> References: <6jvb8u$2do$1@crib.bevc.blacksburg.va.us> <B189C84A-3C2CE@141.214.128.36> In article <B189C84A-3C2CE@141.214.128.36>, "Robert A. Decker" <comrade@umich.edu> wrote: > On Wed, May 20, 1998 2:34 PM, Nathan Urban <mailto:nurban@crib.bevc.blacksburg.va.us> wrote: > > For some reason I seem to have really bad luck with OpenStep, though. :( > Mine goes down about twice a month. It's always OmniWeb that does it. The > entire machine locks up and I can't even telnet into it... While the things that produce my lockups seem to be kind of unpredictable, it _does_ seem that more than the usual share of them occur with OmniWeb (even considering the fact that I use it more). I wonder if there's something to that..
From: rmcassid@uci.edu (Robert Cassidy) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Objective C++ (was Re: Infoworld on the Yellow Box's future) Date: Thu, 21 May 1998 10:10:44 -0700 Organization: UC Irvine Message-ID: <rmcassid-2105981010440001@dante.eng.uci.edu> References: <nospampischke-1905981902030001@pm2-18-2.tor.idirect.com> <nospampischke-1905982350420001@ts7-32t-16.idirect.com> <35626D26.B7616464@alum.mit.edu> <35633314.0@206.25.228.5> <6jvcc4$2i5$1@crib.bevc.blacksburg.va.us> <6k07o4$gc0$1@news.digifix.com> In article <6k07o4$gc0$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: > They syntax issue is resolved. The traditional Objective-C >syntax is staying. But are they officially changing the name to Objective-C++? I smell marketing... -Bob Cassidy
Subject: Re: My _first_ reboot! From: "Robert A. Decker" <comrade@umich.edu> Newsgroups: comp.sys.next.programmer References: <6k1o9p$66j$1@crib.bevc.blacksburg.va.us> MIME-Version: 1.0 Message-ID: <B189E14D-9A3AF@141.214.128.36> Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Date: Thu, 21 May 1998 17:45:01 GMT NNTP-Posting-Date: Thu, 21 May 1998 13:45:01 EDT On Thu, May 21, 1998 12:29 PM, Nathan Urban <mailto:nurban@crib.bevc.blacksburg.va.us> wrote: > While the things that produce my lockups seem to be kind of unpredictable, > it _does_ seem that more than the usual share of them occur with OmniWeb > (even considering the fact that I use it more). I wonder if there's > something to that.. > I actually pretty much stopped using OmniWeb at work (I just use Netscape on my Mac sitting next to my OpenStep box). However, I continue to use it on my home machine which seems much more stable. I love OmniWeb and think it's the best browser out there. The difference between my home machine and work machine is that SCSI is installed at work, and the work machine has a PII 233 and the home machine has a PPro 200. I'm still not completely sure if OmniWeb is causing the problem. I've had the work machine go down twice on its own in about a year. However, I think those crashes are unrelated to the OmniWeb crashes because they occurred when the building cooling system was off and my office got up to 90 degrees F! rob -- <mailto: "Robert A. Decker" comrade@umich.edu> <http://hmrl.cancer.med.umich.edu/Rob/index.ssi> Programmer Analyst - Health Media Research Lab University of Michigan Comprehensive Cancer Center "Get A Life" quote #4: "I feel like Cinderella in more ways than I care to go into." -Chris Elliott
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: NSHosting and MacOS X Date: 18 May 98 21:52:10 Organization: Is a sign of weakness Message-ID: <SCOTT.98May18215210@slave.doubleu.com> References: <1d93iik.1l170dz8etsaxN@carina47.wco.com> <01bd80c6$be64d100$a4e82080@mizuki.HIP.Berkeley.EDU> In-reply-to: "Izumi Ohzawa"'s message of 16 May 1998 20:35:16 GMT In article <01bd80c6$be64d100$a4e82080@mizuki.HIP.Berkeley.EDU>, "Izumi Ohzawa" <izumi@pinoko.berkeley.edu> writes: Still, I haven't heard any explanation as to why they decided to break the client- server boundary at the shared-memory. If multi-threading of DPDF engine is important, well, you can run multiple threads of Display PDF on the server side each listening to PDF streams from clients, either local or remote. Perhaps, it is integration of QuickTime into NSView that drove this decision. I would appreciate more enlightenment, Mike? Specifically, why couldn't local display be done as-is for Carbon and Bluebox apps, and client/server display (not necessarily DPS) for YellowBox apps? It's not like there's any compatibility problems there. And _PLEASE_ do not tell me that in order to draw with my advanced object-oriented environment, I have to revert to a palpimset function-based API, with 4 APIs to do the exact same thing, depending on what timeframe of API you want to use. It sounds suspiciously like building a fairly advanced new kernel using all the avante garde OS principles, only to firmly anchor it with Win32... -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: 19 May 98 12:31:27 Organization: Is a sign of weakness Message-ID: <SCOTT.98May19123127@slave.doubleu.com> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> In-reply-to: "Lawson English"'s message of 16 May 1998 16:34:02 -0700 In article <B1837072-84146@206.165.43.149>, "Lawson English" <english@primenet.com> writes: Matthew Vaughan <see-below@not-my-address.com> said: >> Go ahead. Without Intel, Apple is just wasting my time. So I'm >> history. > >Apple will still ship for Intel. Just not a consumer version of >the OS (ie, with backward compatibility for Macintosh >applications). I still don't see how MacOS X is going to be the consumer version of MacOS. MacOS X is Rhapsody + Carbon, which is likely larger than Rhapsody. How is that "consumer-oriented?" When consumers are buying 32M machines with 6Gig hard drives, how is an OS targetted for 4M machines with 100M drives "consumer-oriented"? Machines that have been sold over the last year are most likely going to be the _minimum_ for MacOS X, which won't ship for over a year. By the time MacOS X ships, consumers are going to be getting machines with >64M RAM and >10Gig drives. I mean, efficiency is to be admired, but not at the expense of utility, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: [Q] What are the implications of MacOS X? Date: 19 May 98 12:34:17 Organization: Is a sign of weakness Message-ID: <SCOTT.98May19123417@slave.doubleu.com> References: <6jof3f$qn4$1@enyo.uwa.edu.au> <B185503A-4EB22@206.165.43.155> <6jq5ot$bcb$1@news.digifix.com> In-reply-to: sanguish@digifix.com's message of 18 May 1998 20:30:21 GMT In article <6jq5ot$bcb$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) writes: On 05/18/98, "Lawson English" wrote: >But why not use it for transfering images over the clipboard? If >the image is merely to be cut/copied/pasted, there's no reason to >use an "industry-standard protocol" between applications when a >simpler and more easily edited (programatically, not textually) >solution could be used instead. It increases the size of the codebase, which increases the cost. It complicates the handling. How so? If indeed MacOS X uses some variant of QuickDraw for drawing, then passing that variant's basic shapes over the clipboard shouldn't be any harder than passing EPS is today in OpenStep. Fact is, there is no reason that you couldn't put your own data on the pasteboard in addition to the PDF and make it available to the application. I'd expect that you could either archive NSBezier and the like and pass them, or ask them to perform a "print to pasteboard" type operation, with the results being EPS, PDF, or some QuickDraw shapes. With very little programmer intervention required, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: NSHosting and MacOS X Date: 21 May 1998 15:16:28 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6k1uic$6dp$1@crib.bevc.blacksburg.va.us> References: <1d93iik.1l170dz8etsaxN@carina47.wco.com> <01bd80c6$be64d100$a4e82080@mizuki.HIP.Berkeley.EDU> <SCOTT.98May18215210@slave.doubleu.com> In article <SCOTT.98May18215210@slave.doubleu.com>, scott@doubleu.com (Scott Hess) wrote: > And _PLEASE_ do not tell me that in order to draw with my advanced > object-oriented environment, I have to revert to a palpimset > function-based API, I've run through all of the words, acronyms, and potential misspellings I can think of, to no avail. So tell me.. what the heck is "palpimset"??
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Objective C++ (was Re: Infoworld on the Yellow Box's future) Date: 21 May 1998 19:33:00 GMT Organization: Idiom Communications Message-ID: <6k1vhc$34b$3@news.idiom.com> References: <nospampischke-1905981902030001@pm2-18-2.tor.idirect.com> <nospampischke-1905982350420001@ts7-32t-16.idirect.com> <35626D26.B7616464@alum.mit.edu> <35633314.0@206.25.228.5> <6jvcc4$2i5$1@crib.bevc.blacksburg.va.us> <6k07o4$gc0$1@news.digifix.com> <rmcassid-2105981010440001@dante.eng.uci.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: rmcassid@uci.edu Robert Cassidy may or may not have said: -> In article <6k07o4$gc0$1@news.digifix.com>, sanguish@digifix.com (Scott -> Anguish) wrote: -> -> > They syntax issue is resolved. The traditional Objective-C -> >syntax is staying. -> -> But are they officially changing the name to Objective-C++? I smell marketing... The name of the compiler is "GCC". Now, for quite some time, the NeXT flavor of that compiler has been able to take both Obj-C, and C++ code in the same source file, with just a bit of trickiness in declaring symbols. So, the language that NeXT (Apple)'s compiler can parse is Objective-C++. Clear? Don't worry. Just quit using C++, it sucks. -jcr
From: Greg_Anderson@afs.com (Gregory H. Anderson) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: NSHosting and MacOS X Date: 21 May 1998 19:38:00 GMT Organization: Anderson Financial Systems Inc. Message-ID: <6k1vqo$itb@shelob.afs.com> References: <6k1uic$6dp$1@crib.bevc.blacksburg.va.us> Nathan Urban writes > In article <SCOTT.98May18215210@slave.doubleu.com>, scott@doubleu.com (Scott Hess) wrote: > > > And _PLEASE_ do not tell me that in order to draw with my advanced > > object-oriented environment, I have to revert to a palpimset > > function-based API, > > I've run through all of the words, acronyms, and potential misspellings > I can think of, to no avail. So tell me.. what the heck is > "palpimset"?? I believe Scott meant to say "palimpsest", which according to Webster is "writing material (as a parchment or tablet) used one or more times after earlier writing has been erased". Sometimes Monsieur Hess's erudition exceeds his orthography. 8^) -- Gregory H. Anderson | "We're in the land of the blind, Visionary Ophthalmologist | selling working eyeballs, and they Anderson Financial Systems | balk at the choice of color." -- Tony greg@afs.com (NeXTmail OK) | Lovell, on Mac user reactions to NeXT
From: Stephan Trebels <stephan@ncube.de> Newsgroups: comp.sys.next.programmer Subject: Graphic cards with 16bit/565 bit layout Date: Fri, 22 May 1998 05:43:16 +0200 Organization: Nacamar Data Communications Message-ID: <3564F454.C1E1B5F@ncube.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi, is it possible to run a 16bit 5-6-5 (i.e. RRRRR-GGGGGBBBBB) display in OPENSTEP for Mach 4.2? I have a Graphics card and a working driver for it, but it operates in 5-6-5 mode, which I could not get running, yet. Any success anyone? Thanks, Stephan
From: William Edward Woody <woody@alumni.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Thu, 21 May 1998 12:44:46 -0700 Organization: In Phase Consulting Message-ID: <3564842E.8D9A4D99@alumni.caltech.edu> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> <SCOTT.98May19123127@slave.doubleu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Scott Hess wrote: > When consumers are buying 32M machines with 6Gig hard drives, how is > an OS targetted for 4M machines with 100M drives "consumer-oriented"? > Machines that have been sold over the last year are most likely going > to be the _minimum_ for MacOS X, which won't ship for over a year. > > By the time MacOS X ships, consumers are going to be getting machines > with >64M RAM and >10Gig drives. I mean, efficiency is to be admired, > but not at the expense of utility, Just keep in mind that the minimum install footprint for Windows 98 (Golden Master 0) is 178 megabytes, and that's without installing any optional utilities whatsoever. - Bill -- William Edward Woody | In Phase Consulting woody@alumni.caltech.edu | Macintosh & MS Windows Development http://www.alumni.caltech.edu/~woody | http://www.pandawave.com/
From: cms@macisp.net (cms) Newsgroups: comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin Subject: Next Turbo Color: Radius Help needed. **** Date: Thu, 21 May 1998 15:18:30 -0500 Organization: CMS Message-ID: <cms-2105981518300001@209.26.71.138> Hi.. We are thinking to purchase the Next System.. but want to use it.. as a Radius Server for Internet Access. Has anyone.. done this.. or is these a pipe dream. I know livingston has the sorce code free for BSDI. Please Email.. cms@macisp.net thank's to all. Rick
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: [Q] What are the implications of MacOS X? Date: 21 May 1998 20:27:42 GMT Organization: Digital Fix Development Message-ID: <6k22nu$6k7$1@news.digifix.com> References: <6jof3f$qn4$1@enyo.uwa.edu.au> <B185503A-4EB22@206.165.43.155> <6jq5ot$bcb$1@news.digifix.com> <SCOTT.98May19123417@slave.doubleu.com> In-Reply-To: <SCOTT.98May19123417@slave.doubleu.com> On 05/19/98, Scott Hess wrote: >In article <6jq5ot$bcb$1@news.digifix.com>, > sanguish@digifix.com (Scott Anguish) writes: > On 05/18/98, "Lawson English" wrote: > >But why not use it for transfering images over the clipboard? If > >the image is merely to be cut/copied/pasted, there's no reason to > >use an "industry-standard protocol" between applications when a > >simpler and more easily edited (programatically, not textually) > >solution could be used instead. > > It increases the size of the codebase, which increases the cost. It > complicates the handling. > >How so? If indeed MacOS X uses some variant of QuickDraw for drawing, >then passing that variant's basic shapes over the clipboard shouldn't >be any harder than passing EPS is today in OpenStep. > Well, the only way to encapsulate graphics in Quickdraw (that is the classic Quickdraw, not the GX beast, which is officially gone in Mac OS X) is using PICT. That really doesn't give you anything more than what you'd get using PDF. Lawson is talking about supporting some sort of QD-GX shape stream capability. I'm saying that putting the extra work into adding a new metadata format for that when there are already YB options that will work quite well (As you say below, archiving NSBezier's as an example) > Fact is, there is no reason that you couldn't put your own data on > the pasteboard in addition to the PDF and make it available to the > application. > >I'd expect that you could either archive NSBezier and the like and >pass them, or ask them to perform a "print to pasteboard" type >operation, with the results being EPS, PDF, or some QuickDraw shapes. >With very little programmer intervention required, Yes.. this is workable way to do it. It doesn't encompass all the options though (like you can't embed text in an NSBezierPath).. but this type of approach is much more appropriate than adding a huge amount of code to support something like has been suggested. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: rmcassid@uci.edu (Robert Cassidy) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Objective C++ (was Re: Infoworld on the Yellow Box's future) Date: Thu, 21 May 1998 13:48:04 -0700 Organization: UC Irvine Message-ID: <rmcassid-2105981348040001@dante.eng.uci.edu> References: <nospampischke-1905981902030001@pm2-18-2.tor.idirect.com> <nospampischke-1905982350420001@ts7-32t-16.idirect.com> <35626D26.B7616464@alum.mit.edu> <35633314.0@206.25.228.5> <6jvcc4$2i5$1@crib.bevc.blacksburg.va.us> <6k07o4$gc0$1@news.digifix.com> <rmcassid-2105981010440001@dante.eng.uci.edu> <6k1vhc$34b$3@news.idiom.com> In article <6k1vhc$34b$3@news.idiom.com>, jcr.remove@this.phrase.idiom.com wrote: > Robert Cassidy may or may not have said: >-> But are they officially changing the name to Objective-C++? I smell >marketing... > >The name of the compiler is "GCC". Now, for quite some time, the NeXT flavor >of that compiler has been able to take both Obj-C, and C++ code in the same >source file, with just a bit of trickiness in declaring symbols. So, the >language that NeXT (Apple)'s compiler can parse is Objective-C++. > >Clear? Yep. Marketing. C is old regardless of applied words, C++ is modern regardless of technical limitations. Give 'em what you want so long as it has the name they want. Good decision. >Don't worry. Just quit using C++, it sucks. Gave it up some years ago. Made my head hurt. -Bob Cassidy
From: xray@cs.brandeis.edu (Nathan G. Raymond) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: NSHosting and MacOS X Date: 21 May 1998 21:41:09 GMT Organization: Brandeis University - Computer Science Dept. Message-ID: <6k271l$9ul$1@new-news.cc.brandeis.edu> References: <01bd80c6$be64d100$a4e82080@mizuki.HIP.Berkeley.EDU> <SCOTT.98May18215210@slave.doubleu.com> <6k1uic$6dp$1@crib.bevc.blacksburg.va.us> In article <6k1uic$6dp$1@crib.bevc.blacksburg.va.us> nurban@vt.edu writes: >In article <SCOTT.98May18215210@slave.doubleu.com>, scott@doubleu.com (Scott Hess) wrote: > >> And _PLEASE_ do not tell me that in order to draw with my advanced >> object-oriented environment, I have to revert to a palpimset >> function-based API, > >I've run through all of the words, acronyms, and potential misspellings >I can think of, to no avail. So tell me.. what the heck is "palpimset"?? I think he meant "palimpsest", which according to the American Heritage Dictionary means: palimpsest n. 1. A manuscript, typically of papyrus or parchment, that has been written on more than once, with the earlier writing incompletely erased and often legible. 2. An object, a place, or an area that reflects its history: "Spaniards in the sixteenth century . . . saw an ocean moving south . . . through a palimpsest of bayous and distributary streams in forested paludal basins" (John McPhee). -- Nathan Raymond http://www.everythingmac.com
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <3564a935.76048282@News.canadawired.com> Control: cancel <3564a935.76048282@News.canadawired.com> Date: 21 May 1998 22:33:19 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.3564a935.76048282@News.canadawired.com> Sender: mlebel@netpointer.com (#1Mario) Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: mdadgar@apple.com (Mark Dadgar) Newsgroups: comp.sys.next.programmer Subject: Disk Fragmentation (was Re: My _first_ reboot!) Date: 22 May 1998 00:10:02 GMT Organization: Apple Computer, Inc. Distribution: world Message-ID: <6k2foq$j08$1@news.apple.com> References: <6k0fon$5pr@mochi.lava.net> In article <6k0fon$5pr@mochi.lava.net> arti@lava.DOTnet (Art Isbell - remove "DOT") writes: > nurban@crib.bevc.blacksburg.va.us (Nathan Urban) wrote: > > > You're joking, right? Don't tell me that NT actually implements > > whole-process _swapping_ over paging! I knew that NT VM performance > > was bad, but not _that_ bad! This would explain much.. > > Regardless of how NT implements VM, the disk seek noise and > sluggishness when changing app focus seems incredible, especially Isn't it awful? It FEELS like swapping. > considering that my PC has 96 MB of RAM. Then I installed > Diskkeeper Lite, a disk defragmenter. It reported 54% > fragmentation! fsck typically reports 0.3% on my Mach systems. This incredible disparity is partly due to the fact that these are really two separate things. On the PC, with the FAT/NTFS filesystems, files become physically fragmented across the disk. That is, logically contiguous blocks of the file are not allocated to physically contiguous blocks on the disk. So you get a lot of head-thrash when you do sequential I/O on the file. This is what Diskkeeper is reporting. On Mach systems, the filesystem is basically the BSD Fast File System. BSD FFS is pretty good about keeping data contiguous on disk. It'll actually move files around on-the-fly to keep them from fragmenting. That's why the filesystem reserves a percentage of free space for maintenance (10% by default, but that was designed in when large disks were 100MB) and slows down dramatically when the disk fills up. So there tends to be a very small amount of file fragmentation. Over time (read: years), you do tend to accumulate a little bit of fragmentation, and the only way to fix it is to dump/newfs/restore the filesystem. However, the % of fragmentation that fsck reports is NOT the kind of file fragmentation that we're thinking of. What it really is is the reported usage of a disk-space optimization algorithm built into the filesystem to prevent wasted space by storing lots of small files. It's a long explanation (and it's described well in _The Design and Implementation of the BSD 4.3 Operating System_ by Leffler, McKusick, et al) so I won't go into it here. Suffice it to say that your Mach system probably has next to no real file fragmentation in the PC sense. - Mark -- Mark Dadgar Product Marketing Apple Computer mdadgar@apple.com
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: 22 May 1998 00:38:02 GMT Organization: Idiom Communications Message-ID: <6k2hda$2rn$1@news.idiom.com> References: <6jvb8u$2do$1@crib.bevc.blacksburg.va.us> <B189C84A-3C2CE@141.214.128.36> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: comrade@umich.edu "Robert A. Decker" may or may not have said: -> On Wed, May 20, 1998 2:34 PM, Nathan Urban -> <mailto:nurban@crib.bevc.blacksburg.va.us> wrote: -> > For some reason I seem to have really bad luck with OpenStep, though. :( -> > -> -> Mine goes down about twice a month. It's always OmniWeb that does it. The -> entire machine locks up and I can't even telnet into it... Omniweb has a few memory leaks, and I would guess that some of the leaked objects are NSImages or are otherwise causing memory in the windowserver to grow. My workaround is just to log out and restart my window server once in a while. -jcr
From: Henry McGilton <henry@trilithon.com> Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: NSHosting and MacOS X Date: Thu, 21 May 1998 17:47:58 -0700 Organization: Trilithon Research and Trading Message-ID: <3564CB3E.AA974741@trilithon.com> References: <1d93iik.1l170dz8etsaxN@carina47.wco.com> <01bd80c6$be64d100$a4e82080@mizuki.HIP.Berkeley.EDU> <SCOTT.98May18215210@slave.doubleu.com> <6k1uic$6dp$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Nathan Urban wrote: * I've run through all of the words, acronyms, and * potential misspellings I can think of, to no avail. * So tell me.. what the heck is "palpimset"?? I assume it's just a mis-spelling of Palimpsest. About as obsolete as a Zumbooruk. ........ Henry ============================================================= Henry McGilton | Trilithon Software, and, Boulevardier, Java Composer | Pacific Research and Trading -----------------------------+------------------------------- mailto:henry@trilithon.com | http://www.trilithon.com =============================================================
From: "Chris Van Buskirk" <cvbuskirk@home.com> Newsgroups: comp.sys.next.programmer Subject: parsing html files.... Message-ID: <W9591.206$ON2.1836919@news.rdc1.nj.home.com> Date: Fri, 22 May 1998 02:12:06 GMT NNTP-Posting-Date: Thu, 21 May 1998 19:12:06 PDT Organization: @Home Network What classes in nextstep are used for parsing. Something like stringtokenizer or streamtokenizer is java. I am way new to yellowbox java apis, and need some help....maybe a book:) -chris
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: parsing html files.... Date: 21 May 1998 22:59:22 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6k2pma$7ef$1@crib.bevc.blacksburg.va.us> References: <W9591.206$ON2.1836919@news.rdc1.nj.home.com> In article <W9591.206$ON2.1836919@news.rdc1.nj.home.com>, "Chris Van Buskirk" <cvbuskirk@home.com> wrote: > What classes in nextstep are used for parsing. Mostly NSScanner.
From: "Michelle L. Buck" <buck.erik@mcleod.net> Newsgroups: comp.sys.next.programmer Subject: Re: parsing html files.... Date: Fri, 22 May 1998 01:58:36 -0500 Organization: McleodUSA - http://www.mcleodusa.net Message-ID: <6k37qi$a8u2@odie.mcleod.net> References: <W9591.206$ON2.1836919@news.rdc1.nj.home.com> Chris Van Buskirk wrote in message ... >What classes in nextstep are used for parsing. Something like >stringtokenizer or streamtokenizer is java. I am way new to >yellowbox java apis, and need some help....maybe a book:) > >-chris > > I am sorry, I just noticed you specified NeXTstep. 10 years ago as a college student, I wrote a very elegant recursive/descent parser in Objective-C. You should still probably use one of the free lex/yacc grammars available. , but I don't know its status. Property lists are automatically parsed in case that matters to you. The lex/yacc or flex/bison grammar for HTML is available on the net and I can only assume DR2 still has those tools. DR2 has Java 1.1.2 and therefore probably has the standard parsing capabilities. At WWDC it was claimed that Interface Builder now works with Java Beans. That may do what you want.
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: NSBrowser doesn't want to reload; Why? (+sample code) Date: 22 May 1998 15:20:51 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6mb5un.36d.rog@talisker.ohm.york.ac.uk> i'm having a problem with NSBrowser using an active delegate. even though i'm loading the cells correctly, (i think!) they aren't displayed, unless some extra rows are added in which case only the extra cells are displayed. at the end of this article, i've included a little bit of code which demonstrates the problem. it should compile (under OpenStep) with just cc tstbrowser.m -framework AppKit (no nib files required) running it and clicking on the "Go!" item in the menu will bring up a window containing a browser and a textfield. typing in the text field and pressing Enter *should* change the first entry of the browser to match. it doesn't work, even though the createRowsForColumn delegate method is being invoked as it ought to be. if anyone can point out what i'm doing wrong i'd be most appreciative! thanks in advance, rog. add a -DEXTRAROW option to show what happens when an extra row is added each time the browser is reloaded. PPS. does anyone know under what circumstances an NSBrowser will sort its component cells into alphabetical order? this behaviour appears to be undocumented. #### snip here: beginning of tstbrowser.m ###### #import <AppKit/AppKit.h> // proposed contents of NSBrowser NSString *browser_contents[] = { @"one", @"two", @"three", @"four" }; #define NUMENTRIES (sizeof(browser_contents) / sizeof(browser_contents[0])) @implementation BrowserHandler: NSObject { NSBrowser *browser; BOOL changed; // whether the browser has been changed since last reloaded } - (void)fieldchanged:sender // the text field has changed: instruct the browser to reload itself. { changed = YES; browser_contents[0] = [[sender stringValue] retain]; // ignore memory leaks! [browser validateVisibleColumns]; } - (void)browser:(NSBrowser *)sender createRowsForColumn:(int)col inMatrix:(NSMatrix *)matrix // load all cells in the first column in the browser from the strings held // in the array browser_contents. { int i; // load each row in turn from browser_contents. for (i = 0; i < NUMENTRIES; i++) { NSBrowserCell *cell = [[NSBrowserCell alloc] initTextCell:browser_contents[i]]; [cell setLeaf:YES]; [matrix addRowWithCells:[NSArray arrayWithObject:cell]]; } #ifdef EXTRAROW { static int count; // add some new cells each time, to demonstrate the problem.... for (i = 0; i < count; i++) { NSBrowserCell *cell = [[NSBrowserCell alloc] initTextCell:[NSString stringWithFormat:@"extra %d", i]]; [cell setLeaf:YES]; [matrix addRowWithCells:[NSArray arrayWithObject:cell]]; } count++; } #endif changed = NO; } - (BOOL)browser:(NSBrowser *)sender isColumnValid:(int)column { return !changed; } #define STYLE (NSTitledWindowMask \ | NSClosableWindowMask \ | NSMiniaturizableWindowMask \ | NSResizableWindowMask) - (void)createwin:sender // create a window containing an NSBrowser and an NSTextField. // typing something into the NSTextField should change the first element // of browser_contents, and then reload the browser to reflect this. // in fact, this doesn't actually happen, for reasons unknown. { NSButton *field; NSRect contentr, fieldr; NSWindow *win = [[NSWindow alloc] initWithContentRect:(NSRect){{200, 200}, {300, 300}} styleMask:STYLE backing:NSBackingStoreBuffered defer:NO]; // place field in bottom left of window. field = [[NSTextField alloc] initWithFrame:(NSRect){{0, 0}, {10, 10}}]; [field setTarget:self]; [field setAction:@selector(fieldchanged:)]; [field setStringValue:@"XXXXXXXXXXXXXXXXXXXXX"]; [field sizeToFit]; [field setStringValue:@""]; fieldr = [field frame]; fieldr.origin.x = fieldr.origin.y = 0; [field setFrame:fieldr]; [[win contentView] addSubview:field]; contentr = [[win contentView] bounds]; contentr.origin.y += fieldr.size.height; contentr.size.height -= fieldr.size.height; // place browser in whatever space is left in the window browser = [[NSBrowser alloc] initWithFrame:contentr]; [browser setAllowsBranchSelection:NO]; [browser setAllowsEmptySelection:YES]; [browser setAllowsMultipleSelection:YES]; [browser setHasHorizontalScroller:NO]; [browser setMaxVisibleColumns:1]; [browser setDelegate:self]; [browser setTitled:YES]; [browser setTitle:@"column zero" ofColumn:0]; [[win contentView] addSubview:browser]; changed = YES; [win makeKeyAndOrderFront:self]; } @end int main(int argc, char **argv) { NSAutoreleasePool *pool; id menu, item, obj; pool = [[NSAutoreleasePool alloc] init]; [NSApplication sharedApplication]; obj = [[BrowserHandler alloc] init]; // create rudimentary main menu containing only "Quit" and "Go!" menu = [[NSMenu alloc] initWithTitle:@"Application"]; item = [menu addItemWithTitle:@"Go!" action:@selector(createwin:) keyEquivalent:@"g"]; [item setTarget:obj]; item = [menu addItemWithTitle:@"Quit" action:@selector(stop:) keyEquivalent:@"q"]; [item setTarget:NSApp]; [NSApp setMainMenu:menu]; [NSApp run]; [pool release]; return 0; }
From: Tony Berke <TonyBerke@Perkins-Hill.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Fri, 22 May 1998 12:22:16 -0400 Organization: Perkins Hill Engineering Message-ID: <3565A638.3983@Perkins-Hill.com> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> <joe.ragosta-1605982005160001@elk32.dol.net> <355E3F66.86DD9F0@thegrid.net> <355EE31C.7834F8EA@milestonerdl.com> <joe.ragosta-1705981730440001@elk85.dol.net> <Pine.NXT.3.96.980520161114.26221R-100000@pathos> <paul-2105980223100001@206.169.118.19> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Paul Hamilton <paul@mbed.com> Paul Hamilton wrote: > > > Does it matter if it's BIG? Does "consumer level" mean small? > < big snip > > "Consumer level" is all about how cheap you can make the boxes I think. > > Paul. You folks have spilled a lot of ink as to how many meg or RAM or disk space an OS can use before it is/isn't a "Consumer OS". I think Paul is right as far as he goes, but I've got to point something out; what consumers really want is easy managability, and above all, apps!! RAM's cheap, and so is disk space. So is processor power, for that matter. Most programmers aren't targeting the MacOS anymore, and for very good reason... except for certain market segments, there aren't enough Macs out there to warrant diverting attention from the 93% of the people running WinTel stuff. This reduces the desirability of the platform for many consumers. Right now, the last significant Mac-centric development efforts seem to be for things like high-end graphics, video, and audio systems. These are hardly "consumer-level" systems, and arguably, neither are the educational users. IMO, Apple's only chance for making MacOS X remain (or return to being, depending on your viewpoint) a "consumer OS" is to make it easy (and I mean *really* easy) for developers to move their apps from Win32 to the Mac. If that takes memory and/or disk footprint, that's utterly trivial. Having a limited selection of applications, on the other hand, is very nontrivial. I didn't make it to WWDC, so I don't know if nowadays Apple sees things this way or not. I hope so. -- Tony Berke Perkins Hill Engineering
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: NSHosting and MacOS X Date: 22 May 1998 17:39:27 GMT Organization: Technical University of Berlin, Germany Message-ID: <6k4d8f$ge0$1@news.cs.tu-berlin.de> References: <1d93iik.1l170dz8etsaxN@carina47.wco.com> <01bd80c6$be64d100$a4e82080@mizuki.HIP.Berkeley.EDU> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit "Izumi Ohzawa" <izumi@pinoko.berkeley.edu> writes: >Mike Paquette <mpaque@wco.com> wrote in article ><1d93iik.1l170dz8etsaxN@carina47.wco.com>... >> There has been some concern expressed over MacOS X and the loss of the >> NXHost capability. I'd like to clear a few things up. >> >> NX/NS Host Operation >> >> NXHost works by redirecting the PostScript binary object stream to a >> PostScript interpreter running in a Window Server on a different machine >> than the current application. A PostScript interpreter is required to >> process this stream. >> >> MacOS X will not require a PostScript interpreter. One will not be >> present in the base system. Therefore, one cannot reasonably process >> the existing NSHost protocol to a MacOS X system. >Is it really the loss of DPS that makes NSHosting difficult? >After all, in principle, there can be Display PDF engine doing similar work >as DPS >and work on PDF streams from clients. Re-read Mike's post. He was very clear in separating (a) NSHost capability, the current protocol which requires DPS (b) remote display capability, which does not Supporting option (a) is not possible because there is no Postscript interpreter included with MacOS X. There seems to be no problem in supporting (b) except for available time and priorities. Additionally, there may be support for third parties to provide this capability. As usual, it all makes sense. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: Andre-John Mas <ama@fabre.act.qc.ca> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: Fri, 22 May 1998 13:11:20 -0500 Organization: Communications Accessibles Montreal, Quebec Canada Message-ID: <Pine.LNX.3.95.980522130546.18546B-100000@fabre.act.qc.ca> References: <see-below-1605981620340001@209.24.241.16> <B1837072-84146@206.165.43.149> <joe.ragosta-1605982005160001@elk32.dol.net> <355E3F66.86DD9F0@thegrid.net> <355EE31C.7834F8EA@milestonerdl.com> <joe.ragosta-1705981730440001@elk85.dol.net> <Pine.NXT.3.96.980520161114.26221R-100000@pathos> <paul-2105980223100001@206.169.118.19> <3565A638.3983@Perkins-Hill.com> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII In-Reply-To: <3565A638.3983@Perkins-Hill.com> The question of lack of software is one very good reason why Apple must get the Yellow box right and have implementations of it available for other platforms. It is also important that Apple builds an excellent run time environment for Java. When I say excellent I mean stable, fast, and easy to configure. Although Java is still in its infency Apple should work with Sun and possibly Symantec to build the ideas JRE. If Apple plays it cards right, they might be able to nullify some of the current arguments against buying into the MacOS platform. Though nullifying myths and prejudice is a totally different issue. AJ ---------------------------------------------------------- Andre-John Mas mailto:ama@act.qc.ca ----------------------------------------------------------
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: NSHosting and MacOS X Date: 22 May 1998 14:27:41 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6k4g2t$8u6$1@crib.bevc.blacksburg.va.us> References: <1d93iik.1l170dz8etsaxN@carina47.wco.com> <01bd80c6$be64d100$a4e82080@mizuki.HIP.Berkeley.EDU> <6k4d8f$ge0$1@news.cs.tu-berlin.de> In article <6k4d8f$ge0$1@news.cs.tu-berlin.de>, marcel@cs.tu-berlin.de (Marcel Weiher) wrote: > Re-read Mike's post. > > He was very clear in separating > > (a) NSHost capability, the current protocol which requires DPS > (b) remote display capability, which does not Of course, (b) is practically useless, IMHO, if it requires blasting bitmaps over a network. I guess it would be nice to have for emergencies, when you really need it you really need it, but probably not something you want to use day-in and day-out (especially if a bunch of people want to use it, all on the same subnet). I'm wondering what Apple could do to leave hooks for NSHosting.. I'll assume for the moment that you want to just NSHost Yellow apps that stick to the PS (and PDF?) functions and the AppKit APIs that use them. Perhaps they could just have an if() statement at the top of the implementation of each of those functions that, if some global flag is true, would jump into some vector of replacement functions that could insert themselves into a network command stream. (You could even do some tricky compiler stuff too so that the replacement function would use the same stack as the function that called it so you don't have to push everything on again immediately, but it's dangerous.)
From: NOSPAMmbkennelNOSPAM@yahoo.com (Matt Kennel) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: NSHosting and MacOS X Date: 22 May 1998 18:43:35 GMT Organization: University of California at San Diego Message-ID: <slrn6mbhqm.894.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu> References: <1d93iik.1l170dz8etsaxN@carina47.wco.com> <01bd80c6$be64d100$a4e82080@mizuki.HIP.Berkeley.EDU> <6k4d8f$ge0$1@news.cs.tu-berlin.de> <6k4g2t$8u6$1@crib.bevc.blacksburg.va.us> : :> Re-read Mike's post. :> :> He was very clear in separating :> :> (a) NSHost capability, the current protocol which requires DPS :> (b) remote display capability, which does not : :Of course, (b) is practically useless, IMHO, if it requires blasting :bitmaps over a network. I guess it would be nice to have for emergencies, :when you really need it you really need it, but probably not something :you want to use day-in and day-out (especially if a bunch of people want :to use it, all on the same subnet). :I'm wondering what Apple could do to leave hooks for NSHosting.. :I'll assume for the moment that you want to just NSHost Yellow apps that :stick to the PS (and PDF?) functions and the AppKit APIs that use them. Probably so. Unless they are able to redirect Quickdraw calls in the carbon layer too. I'm guessing that the QD protocol and concepts are not particularly suitable for good remote display. :Perhaps they could just have an if() statement at the top of the :implementation of each of those functions that, if some global flag is :true, would jump into some vector of replacement functions that could :insert themselves into a network command stream. (You could even do :some tricky compiler stuff too so that the replacement function would :use the same stack as the function that called it so you don't have to :push everything on again immediately, but it's dangerous.) 'if'?? Surely they would dispatch via standard object mechanisms instead. Remember that drawing is only half the problem---one must suck input as well as display configuration from the pipe, as well as dealing with window manager events. -- * Matthew B. Kennel/Institute for Nonlinear Science, UCSD - * "People who send spam to Emperor Cartagia... vanish! _They say_ that * there's a room where he has their heads, lined up in a row on a desk... * _They say_ that late at night, he goes there, and talks to them... _they *- say_ he asks them, 'Now tell me again, how _do_ you make money fast?'"
From: "David Waffen" <dmwaff@erols.com> Newsgroups: comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin Subject: Re: Next Turbo Color: Radius Help needed. **** Date: Fri, 22 May 1998 15:17:59 -0400 Organization: Erol's Internet Services Message-ID: <6k4j7e$lll$1@winter.news.erols.com> References: <cms-2105981518300001@209.26.71.138> cms wrote in message ... > >Hi.. > >We are thinking to purchase the Next System.. > ================= You may want to think about what version of NeXT. v.3.3 is not 2k compliant. Apple has mention Patches. OpenStep, Rap and later versions are, but support is limited. I have not used the later ones, but Next v3.3 as an OS needs a static ip address. NeXTStep does not support DCHP nor ISDN modems, that I have heard. If there is a configuration let me know. I have to use an acend pipeline 50 router at home to connect to my ISDN line and work network. HONESTLY, I love my NextStep but applications are limited, Lighthouse products blow, and think it will become a dinosaur. Hopefully OpenStep and Rhapsody will stay. David
Subject: Re: Next Turbo Color: Radius Help needed. **** Newsgroups: comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin References: <cms-2105981518300001@209.26.71.138> <6k4j7e$lll$1@winter.news.erols.com> In-Reply-To: <6k4j7e$lll$1@winter.news.erols.com> From: nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) Followup-To: comp.sys.next.sysadmin Message-ID: <J%l91.246$ON2.2266091@news.rdc1.nj.home.com> Date: Fri, 22 May 1998 21:21:45 GMT NNTP-Posting-Date: Fri, 22 May 1998 14:21:45 PDT Organization: @Home Network On 05/22/98, "David Waffen" wrote: > You may want to think about what version of NeXT. > v.3.3 is not 2k compliant. There is only one part of 3.3 that was not compliant, setting the Date with Preferences.app. There's already a fix for this. > Apple has mention Patches. Haven't heard this myself, but I don't know what else needs patching > Next v3.3 as an OS needs a static ip address. False. Dynamic PPP works just fine. > NeXTStep does not support DCHP nor ISDN > modems, that I have heard. If there is a > configuration let me know. I believe that there are folks out there who have DHCP working under 3.3.... personally I haven't had to use it. I've used my 3.3 NeXTStation with a static IP and a cable modem, and gotten speeds just as fast as with my P-133 > HONESTLY, I love my NextStep but applications are > limited, Lighthouse products blow, and think it > will become a dinosaur. Hopefully OpenStep and > Rhapsody will stay. OpenStep is dead and Rhapsody won't be around for Intel (which will probably drive me to the world of Windows unless PPC hardware suddenly becomes an affordable alternative). OpenStep 4.2 will slowly become obsolete.... there will more than likely never be a Javascript-enabled browser for OS, nor Real Audio, nor a decent QuickTime/AVI/MPG anything else.... I expect to find myself booting into Win95 more and more as I want to use the Internet. It's sad, but after 7 years in the NeXTStep world, it feels like the end is near for me.... I'll probably keep using Unix in another form, probably FreeBSD, but I think the no-Intel future for Apple will be the final mistake.... NeXT, Inc found out that the world didn't want better-but-more-expensive hardware, they want sucky PC hardware because it's cheaper. I'd be happy to be wrong, but after that announcement my hope for an enjoyable OS in the future died.... TjL
From: Makoto Sadahiro <sadahiro@cs.utexas.edu> Newsgroups: comp.sys.next.programmer Subject: how can I study? Date: Fri, 22 May 1998 16:52:33 -0500 Organization: U of Texas Austin Message-ID: <3565F3A1.794B@cs.utexas.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi. I just got NeXT Cube with NS3.3dev. What would be a first recommended reading? I have bit experience programming C++ on unix, but have no idea how to develop software on NeXT. I kind of started to blowse around online docs in /lib. If you can directly me for some good reading, will be really apprecited. Thanks. -- ~Makoto Sadahiro A Designer Of Old Code. sadahiro@cs.utexas.edu www.cs.utexas.edu/users/sadahiro
From: PsxModifications <generation5@hotmail.com> Newsgroups: comp.sys.next.programmer Subject: @@@ Sony Playstation Modification Chips @@@@ Date: 22 May 1998 23:24:09 GMT Organization: PsxModChips Message-ID: <6k51ep$n31$10564@m5.stny.lrun.com> The hottest Mod chip on the web is back! GENERATION 5 We are settled in our new location and ready to bring quality back to the market. GENERATION 5 mod chips allow you to: Play Backups (Gold, Silvers) Play Imports Play Originals You will be able to install our 4 pin model (PIC 12c508) within minutes. Detailed installation instructions are available as well as many satisfied customers. And or course, you get the BEST PRICE. Check out our easy ordering at our webpage. www.undergroundxpress.com/planetplaystation Mod chip installations are only 30$. Take advantage of them. Check out our webpage for more!!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6k51ep$n31$10564@m5.stny.lrun.com> Control: cancel <6k51ep$n31$10564@m5.stny.lrun.com> Date: 22 May 1998 23:34:53 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6k51ep$n31$10564@m5.stny.lrun.com> Sender: PsxModifications <generation5@hotmail.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: izumi@pinoko.berkeley.edu (Izumi Ohzawa) Newsgroups: comp.sys.next.programmer Subject: Re: NSHosting and MacOS X Date: 23 May 1998 03:14:58 GMT Organization: University of California, Berkeley Distribution: world Message-ID: <6k5evi$l6j$1@agate.berkeley.edu> References: <6k4d8f$ge0$1@news.cs.tu-berlin.de> In article <6k4d8f$ge0$1@news.cs.tu-berlin.de> marcel@cs.tu-berlin.de (Marcel Weiher) writes: >Re-read Mike's post. > >"Izumi Ohzawa" <izumi@pinoko.berkeley.edu> writes: >>Is it really the loss of DPS that makes NSHosting difficult? > >>Mike Paquette <mpaque@wco.com> wrote in article >>> >>> NX/NS Host Operation >>> >>> NXHost works by redirecting the PostScript binary object stream to a >>> PostScript interpreter running in a Window Server on a different machine >>> than the current application. A PostScript interpreter is required to >>> process this stream. I re-read Mike's post, and it still does not make sense. May be it's just me. What I thought possible is described below, rephrasing Mike's paragraph above. I wish this were what they said and did, and I still don't understand why this isn't possible: Diplay PostScript is gone, but DPS is now replaced by Display PDF. NXHost now works by redirecting the new PDF binary object stream to a PDF processing engine running in a WindowServer on a different machine than the current application. Though the representation of drawing primitives sent from client to server has changed from PostScript to PDF, this change is transparent to the programmer unless she has used complex PSwraps that send procedure defs, etc. not available in PDF specification. To achieve multi-threaded execution of PDF streams from different applications, WindowServer now runs a thread of proxy PDF interpreter for each remote client application (local clients still draws directly via PDF calls that draw into rects via the Interceptor). Again, this change is transparent to the user and programmer, as long as drawing has been programmed using basic common primitives supported by PDF and PS. So, everything just works (TM). -------------------------------- Sadly, this is not what we got. Why not, why not, why not?
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.pen,comp.sys.powerpc Subject: cmsg cancel <6k0ect$4q2$48@news.metrocom.ru> Control: cancel <6k0ect$4q2$48@news.metrocom.ru> Date: 23 May 1998 05:01:18 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6k0ect$4q2$48@news.metrocom.ru> Sender: alexnik@tomcat.ru Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <23059803.3101@pussy.here> Control: cancel <23059803.3101@pussy.here> Date: 23 May 1998 07:04:38 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.23059803.3101@pussy.here> Sender: MovieStar@pussy.here Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <1998052307115501.DAA16026@ladder03.news.aol.com> Date: 23 May 1998 07:11:55 GMT Organization: AOL http://www.aol.com References: <6jc50h$eh@fridge.shore.net> I don't see why you couldn't set the port on a per-thread basis. Assuming Carbon follows the MacOS API model, spawning multiple drawing threads might just as well require each thread to have it's own copy of QD Globals so that Draw operations in threads don't cross each other. Also, isn't this just a classic race condition that can be arbitrated via standard mutex means? The MacOS currently doesn't support multi-threaded drawing. In fact PowerPC threads are pretty restrictive and cooperative right now. If Carbon were to support multi-threading al'a Mach, it would require the programmer to write all new threading code anyway. Besides, this is all arbitrated by the core OS services which live between the individual boxes and the kernel anyway. It could be invisibly resolved to a restricted API like carbon. <<<< I don't see how the Mac API could be made thread safe. Suppose you have two threads, thread A, and thread B. Thread A calls SetPort(), and then starts drawing. At about the same time, thread B calls SetPort to a different port, and then starts drawing. The problem is a global current port, so both threads will end up drawing into the same port. >>>>> -- Dr. Nuketopia Read the Blue Glow in Tubes FAQ at http://www.persci.com/~larrysb This address gets lots-o-spam. Please note that your letter is *not* spam in the subject line.
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Message-ID: <1998052307482001.DAA17251@ladder03.news.aol.com> Date: 23 May 1998 07:48:20 GMT Organization: AOL http://www.aol.com References: <6jvp3d$62e$2@news.idiom.com> <<<< Of course, if you take a bunch of those simple, common UNIX programs like sed and the like out of the product, then you'll have a needlessly crippled system. >>> What the hell is my mom gonna do with "sed"? I mean seriously, if you want to compact OS-X, you can leave out the /usr/bin , /sbin /libexec and all the compilers, libraries, utilities, network daemons, vi, emacs, all the shared libraries and so forth. Only us nerds use or understand(?) that stuff anyway. Leave it on the distribution CD, or even offer it as a seperate option purchase. Good grief. -- Dr. Nuketopia Read the Blue Glow in Tubes FAQ at http://www.persci.com/~larrysb This address gets lots-o-spam. Please note that your letter is *not* spam in the subject line.
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <1998052308052000.EAA21989@ladder01.news.aol.com> Date: 23 May 1998 08:05:20 GMT Organization: AOL http://www.aol.com References: <6j84dp$vln$1@crib.bevc.blacksburg.va.us> <<< Better advice: if you're writing a new app, forget the Mac API. OpenStep is infinitely better. Might as well develop good habits from the beginning, and not waste your time. >>> Oh bullshit. There are plenty of cases where OpenStep just plain sux. I had to break it to you, as good as it is it suffers from the same problem *all* frameworks have: They are great as long as what you want to do fits with the frameworks designer's goals. Start coloring outside the lines a little and you wind up having to backtrack over what the framework has left out, or worse, has done wrong. To make matters worse, you can't see the source code, like you can with C++ frameworks. If you want to write a Nextstep app today, and your market is already running OpenStep or Windows + OpenStep, well good. Do it. Also, OpenStep can be quite s l o w, even on fast hardware. I've run it on Pentium Pro 200's, PowerMac 604e/200's and G3/266's. It's slow. Not the Unix part, but the yellowbox part is slow. Even the current bluebox is more responsive than the Yellowbox. I had reserved all judgement until I got DR2, but that's my opinion. It positively crawls. What if you already have significant investment in application code, or even libraries of domain specific code built up over years of development? -- Dr. Nuketopia Read the Blue Glow in Tubes FAQ at http://www.persci.com/~larrysb This address gets lots-o-spam. Please note that your letter is *not* spam in the subject line.
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer,comp.sys.next.advocacy Subject: Re: future of mac programming Date: 23 May 1998 04:29:49 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> References: <6j84dp$vln$1@crib.bevc.blacksburg.va.us> <1998052308052000.EAA21989@ladder01.news.aol.com> In article <1998052308052000.EAA21989@ladder01.news.aol.com>, larrysb@aol.com (LarrySB) wrote: > [I wrote:] > <<< > Better advice: if you're writing a new app, forget the Mac API. > OpenStep is infinitely better. Might as well develop good habits from > the beginning, and not waste your time. > >>> > Oh bullshit. > There are plenty of cases where OpenStep just plain sux. Okay, give me some examples where the Mac API is better. Go ahead and produce some reasons why I'd rather use it over OpenStep. All you produced was a bunch of whining about how you can't do anything outside the lines, which is of course patently ludicrous, not to mention completely unsubstantiated. Sure, the frameworks don't _help_ you very much to do things they weren't designed to do (though OpenStep is still unusually flexible in that regard), but they don't go out of their way to make those things _harder_ -- i.e., harder than it would be _without_ a framework. If worse comes to worst, you simply end up not using the framework for some portion of your code, and you don't end up any worse than before. > Also, OpenStep can be quite s l o w, even on fast hardware. I've run it on > Pentium Pro 200's, PowerMac 604e/200's and G3/266's. It's slow. Not the Unix > part, but the yellowbox part is slow. Even the current bluebox is more > responsive than the Yellowbox. I had reserved all judgement until I got DR2, > but that's my opinion. It positively crawls. I don't know what the heck you're running, but OpenStep is quite _fast_. I haven't compared it to MacOS, not having a Mac, but it's certainly faster than Windows 95 or NT, and I find it quite responsive even on my Pentium/120. (At least OPENSTEP 4.2; I don't have DR2 yet and DR1 is no comparison.) Are you telling me that you've tried DR2 (not DR1) and still think it "positively crawls"?? In what respect do you think Yellow Box is slow, and what did you run to form that opinion? Are you running on a RAM-starved machine or have a really high bit depth? > What if you already have significant investment in application code, or even > libraries of domain specific code built up over years of development? That's why I said NEW apps, which is what the original poster was interested in -- he was interested in Rhapsody/MacOS X programming and, of the two, wanted to know which he should learn. I can't think of a compelling reason for him to spend time learning the Carbon APIs if he's interested in MacOS X programming and not MacOS 8.x.
From: frank@this.NO_SPAM.net (Frank M. Siegert) Newsgroups: comp.sys.next.programmer Subject: Re: NSHosting and MacOS X Date: 23 May 1998 12:03:42 GMT Organization: Frank's Area 51 Distribution: world Message-ID: <6k6duu$qou$1@news.seicom.net> References: <6k4d8f$ge0$1@news.cs.tu-berlin.de> <6k5evi$l6j$1@agate.berkeley.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Cc: izumi@pinoko.berkeley.edu In <6k5evi$l6j$1@agate.berkeley.edu> Izumi Ohzawa wrote: > ... > Sadly, this is not what we got. Why not, why not, why not? Mh, just speculating... - The original model is not based on DPDF alone but for performance (or whatever) sake Quickdraw and QT can draw to physical regions on the screen thus remote hosting wouldn't work for (some?) Carbon based applications. - Because it is not considered to be of any importance. -- * Frank M. Siegert [frank@this.net] - Home http://www.this.net/~frank * NeXTSTEP, IRIX, Solaris, Linux, BeOS, PDF & PostScript Wizard * "The answer is vi, what was your question...?"
From: Dana.Delany@Nude.Here.com Newsgroups: comp.sys.next.programmer Subject: ALL OF THE HOTTEST NUDE CELEBS 35732 Date: Saturday, 23 May 1998 11:10:10 -0600 Organization: <no organization> Distribution: World Message-ID: <23059811.1010@Nude.Here.com> http://members.coolnet.net/~tyler12 THE ORIGINAL CELEBS SITE WE HAVE OVER 8,000 CELEB PICS INCLUDING OVER 1,000 PERSONALLY TAKEN "SPY" PICS AND VIDEO OF HOT CELEBS. SOME OF THESE PICS AND VIDEOS ARE BRAND NEW HOTEL ROOM FOOTAGE FROM CELEBS HOLIDAY VACATIONS YOU HAVE TO SEE ALL THESE STARS WE ARE ADDING NEW PICS EVERY WEEK **CAST OF THE MOVIE THE CRAFT SNEAK PIX** *Nicole Eggert, Samantha Fox, Hellen Hunt *MARISA TOMEI, TIA LEONIE, PAULA ABDULE *SPICE GIRLS, JEWEL, HOLLY ROBINSON, *U.S. OLYMPIC GYMNAST LOCKER ROOM PIX *NEVE CAMPBELL, FIONA APPLE, *ENTIRE BEVERLY HILLS 90120 CAST *ENTIRE FEMALE STAR TREK CAST *MTV's TABITHA SOREN and SARENA ALTSHULA http://members.coolnet.net/~tyler12 http://members.coolnet.net/~tyler12 MvmF
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <23059811.1010@Nude.Here.com> Control: cancel <23059811.1010@Nude.Here.com> Date: 23 May 1998 14:12:34 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.23059811.1010@Nude.Here.com> Sender: Dana.Delany@Nude.Here.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: 23 May 1998 14:34:34 GMT Organization: Idiom Communications Message-ID: <6k6mpq$avg$2@news.idiom.com> References: <6jvp3d$62e$2@news.idiom.com> <1998052307482001.DAA17251@ladder03.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: larrysb@aol.com LarrySB may or may not have said: -> <<<< -> Of course, if you take a bunch of those simple, common UNIX programs like sed -> and the like out of the product, then you'll have a needlessly crippled -> system. -> >>> -> -> What the hell is my mom gonna do with "sed"? She's going to ignore it, obviously. However, she might someday install a program that wants to alter the rc.local script (you know, something like a UPS controller) and that program is going to need sed to figure out where to insert itself in the boot sequence. Or, perhaps she's going to use a control panel to kill an app that's hung. Maybe that app uses ps | sed ... to find the PID of the app to kill, given its name. -> I mean seriously, if you want to compact OS-X, you can leave out the /usr/bin , -> /sbin /libexec and all the compilers, libraries, utilities, network daemons, -> vi, emacs, all the shared libraries and so forth. Here's an exercise for you: On any UNIX system, try removing stdlib. Then try rebooting. -> Only us nerds use or understand(?) that stuff anyway. It's not just us, it's us and the CODE we write. Code that can be used by people who coldn't write it. -> Leave it on the distribution CD, or even offer it as a seperate option -> purchase. It's about FOUR CENTS worth of disk space, for crying out loud. -> Good grief. Think it through. -jcr
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: how can I study? Date: 23 May 1998 14:39:00 GMT Organization: Idiom Communications Message-ID: <6k6n24$avg$3@news.idiom.com> References: <3565F3A1.794B@cs.utexas.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: sadahiro@cs.utexas.edu Makoto Sadahiro may or may not have said: -> Hi. I just got NeXT Cube with NS3.3dev. What would be a first -> recommended reading? I have bit experience programming C++ on unix, but -> have no idea how to develop software on NeXT. I kind of started to -> blowse around online docs in /lib. -> -> If you can directly me for some good reading, will be really -> apprecited. Start with /NextDeveloper/Bookshelves/Developer.bshlf Look for the interface builder tutorial, and the Objective-C introduction. -jcr
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 23 May 1998 14:52:09 GMT Organization: Idiom Communications Message-ID: <6k6nqp$avg$4@news.idiom.com> References: <6j84dp$vln$1@crib.bevc.blacksburg.va.us> <1998052308052000.EAA21989@ladder01.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: larrysb@aol.com LarrySB may or may not have said: -> <<< -> Better advice: if you're writing a new app, forget the Mac API. -> OpenStep is infinitely better. Might as well develop good habits from -> the beginning, and not waste your time. -> >>> -> -> Oh bullshit. -> -> There are plenty of cases where OpenStep just plain sux. I had to break it to -> you, as good as it is it suffers from the same problem *all* frameworks have: -> They are great as long as what you want to do fits with the frameworks -> designer's goals. Start coloring outside the lines a little and you wind up -> having to backtrack over what the framework has left out, or worse, has done -> wrong. To make matters worse, you can't see the source code, like you can with -> C++ frameworks. It's certainly better if you can see the code, but it's simply not true that all C++ framework vendors give you their code to read. As for fixing things that the framework has "done wrong", well that's exactly the same as any other situation where you want to alter the behaviour of given class. You subclass it and change what you don't like, or extend it to cover what you think is missing. OpenStep is the best environment I've seen yet for software development, with the *possible* exception of Smalltalk. -> If you want to write a Nextstep app today, and your market is already running -> OpenStep or Windows + OpenStep, well good. Do it. -> -> Also, OpenStep can be quite s l o w, even on fast hardware. I've run it on -> Pentium Pro 200's, PowerMac 604e/200's and G3/266's. It's slow. Not the Unix -> part, but the yellowbox part is slow. Even the current bluebox is more -> responsive than the Yellowbox. I had reserved all judgement until I got DR2, -> but that's my opinion. It positively crawls. So run it under a profiler, and fix it! -> What if you already have significant investment in application code, or even -> libraries of domain specific code built up over years of development? You link it in and you call it! Obj-C has no problems at all calling C and C++ code. If I were a Mac developer today, looking at the transition from the MacToolbox API to Carbon, I'd take the opportunity to re-do the GUI using the Appkit. It's fast, it's flexible, and it's a few hundred man-years of code I don't have to write. -jcr
Subject: Re: New display model in OS-X From: "Baskaran Subramaniam" <baskar@RemoveMe.snet.net> Newsgroups: comp.sys.next.programmer References: <6joa23$ik26@odie.mcleod.net> MIME-Version: 1.0 Message-ID: <B18C6B5E-2D6E6A@192.168.128.1> Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Date: Sat, 23 May 1998 16:08:12 GMT NNTP-Posting-Date: Sat, 23 May 1998 12:08:12 EDT On Sun, May 17, 1998 11:29 PM, Michelle L. Buck <mailto:buck.erik@mcleod.net> wrote: >There is another even more alarming problem. If a Carbon application takes >over the screen and does not give it back, no other application (Yellow or >Carbon) will be able to display! A dead or infinitely looping Carbon >application will prevent you from using the Process panel to kill it. This >question was asked and clarified several times. Carbon applications will be >able to take over the entire display and we must rely on their good nature >to give it back and let other applications draw. I used to deride Windows >for that flaw. As far as I know, in MacOS 10, Carbon is an API spec for the module that runs on top of the kernel, the same kernel that the YB API uses. Carbon is not the bluebox compatibility environment. Carbon cannot take over the machine in the way you described above. In fact they demoed a spinning color wheel in a carbon app while another QT movie was playing just to show that. Later on the spinning color wheel application was killed by invoking the process panel. I have been following the discussion in this group since WWDC. The majority of the poeple are just going crazy over either incomplete or down right wrong interpretation of the new display model described at WWDC. Apple website has the slides and the QT movies of all the sessions. If you lack information, I think it is worthwile to spend a little time looking at them to get a better idea. Please do not expect the rumour sites or the clueless media to provide the correct information for you. I agree that Apple screwed up in the way they communicated to their developers about the changes in their OS strategy. I do not know how NeXT was with respect to this, but if anyone knows how best to garble a clear message, I can guarantee you that Apple will beat them by a wide margin. It is just plain frustrating. Baskaran. PS: Please delete the string "RemoveMe." from my email address before replying to this message
Subject: Re: future of mac programming From: "Baskaran Subramaniam" <baskar@RemoveMe.snet.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer nntp://news.internetMCI.com/comp.sys.mac.programmer.codewarrior, nntp://news.internetMCI.com/comp.sys.next.programmer References: <SCOTT.98May11153549@slave.doubleu.com> MIME-Version: 1.0 Message-ID: <B18C6E34-2E193D@192.168.128.1> Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Date: Sat, 23 May 1998 16:20:18 GMT NNTP-Posting-Date: Sat, 23 May 1998 12:20:18 EDT On Mon, May 11, 1998 11:35 AM, Scott Hess <mailto:scott@doubleu.com> wrote: >My >suspicion is that Carbon is essentially a hugely improvd BlueBox, and >that something AppKit based is still where the medium to long term >future is. I don't think so. Bluebox is a compatibility environment for executing legacy MacOS binaries without a recompile. On the other hand Carbon is an API spec that allows software writers to recompile existing MacOS source after making minimal changes to bring the code Carbon-compliant. Baskaran PS: Please delete the string "RemoveMe." from my email address before replying to this message
Subject: Re: future of mac programming From: "Baskaran Subramaniam" <baskar@RemoveMe.snet.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer nntp://news.internetMCI.com/comp.sys.mac.programmer.codewarrior, nntp://news.internetMCI.com/comp.sys.next.programmer References: <MPG.fc14db42395aa139896a2@news.supernews.com> MIME-Version: 1.0 Message-ID: <B18C6F81-2E6780@192.168.128.1> Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Date: Sat, 23 May 1998 16:25:51 GMT NNTP-Posting-Date: Sat, 23 May 1998 12:25:51 EDT On Mon, May 11, 1998 8:03 PM, Donald Brown <mailto:don.brown@cesoft.com> wrote: >If you're not already working with Rhapsody/OpenStep/Yellow Box/etc., >hold off until we hear more. The word I thought I heard this morning is >that those are not the future, at least for client software. > That is plain wrong information. Yellow box is here to stay. You will see changes in the DPS-dependent parts of the AppKit, and you will see more objects added to it for any new functionality. Also, YB is THE API to use if you want to your app to run on multiple platforms. Baskaran PS: Please delete the string "RemoveMe." from my email address before replying to this message
From: "Chris Hegarty" <chris.hegarty@virgin.net> Newsgroups: comp.sys.next.programmer Subject: What is Carbon? Date: Sat, 23 May 1998 18:22:45 +0100 Organization: Virgin Net News Service Message-ID: <6k702m$a0l$1@nclient1-gui.server.virgin.net> I am new to NeXTSTEP/OPENSTEP coding and I want to know what this Carbon is, PLEASE??
From: "Michelle L. Buck" <buck.erik@mcleod.net> Newsgroups: comp.sys.next.programmer Subject: Re: New display model in OS-X Date: Sat, 23 May 1998 12:23:06 -0500 Organization: McleodUSA - http://www.mcleodusa.net Message-ID: <6k70pe$11q81@odie.mcleod.net> References: <6joa23$ik26@odie.mcleod.net> <B18C6B5E-2D6E6A@192.168.128.1> Baskaran Subramaniam wrote in message ... >On Sun, May 17, 1998 11:29 PM, Michelle L. Buck ><mailto:buck.erik@mcleod.net> wrote: >>There is another even more alarming problem. If a Carbon application takes >>over the screen and does not give it back, no other application (Yellow or >>Carbon) will be able to display! A dead or infinitely looping Carbon >>application will prevent you from using the Process panel to kill it. This >>question was asked and clarified several times. Carbon applications will >be >>able to take over the entire display and we must rely on their good nature >>to give it back and let other applications draw. I used to deride Windows >>for that flaw. > >As far as I know, in MacOS 10, Carbon is an API spec for the module that >runs on top of the kernel, the same kernel that the YB API uses. Carbon >is not the bluebox compatibility environment. Carbon cannot take over >the machine in the way you described above. In fact they demoed a >spinning color wheel in a carbon app while another QT movie was >playing just to show that. Later on the spinning color wheel application >was killed by invoking the process panel. > [Frustration with garbled message deleted] > >Baskaran. > > >PS: Please delete the string "RemoveMe." from my email address before > replying to this message > > I was THERE. I have all of the session notes and many of the videos. The demo in which they killed a hung carbon app has nothing to do with display issues. That carbon app had not locked down the screen. IF a carbon application chooses to lock down the entire screen it can. AND IF it then loops infinitely and or crashes WITHOUT releasing the screen, you better hope you have a terminal attached or can telnet in because the display is locked! You will not be able to see the Processes panel to kill the Carbon Application. YellowBox may gain this "feature" also in the new display architecture. Don't get me wrong. There are reasons to allow this. Chief among them is games and or single purpose installations like medical imaging. This is appropriate for a client/home computer. This is not appropriate for a server. Carbon goes a long way to making toolbox/quickdraw not-so-bad. As with everything else, this compromise has a down side also. - Erik M. Buck (From my wife's account)
From: "Chris Hegarty" <chris.hegarty@virgin.net> Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: Sat, 23 May 1998 18:28:21 +0100 Organization: Virgin Net News Service Message-ID: <6k70d6$a4o$1@nclient1-gui.server.virgin.net> References: <6junig$is5$2@ns3.vrx.net> <Pine.NXT.3.96.980520161846.26221T-100000@pathos> <6jvfa8$2pj$1@crib.bevc.blacksburg.va.us> <6k0fon$5pr@mochi.lava.net> <6k0haj$41v$1@crib.bevc.blacksburg.va.us> Nathan Urban wrote in message <6k0haj$41v$1@crib.bevc.blacksburg.va.us>... >In article <6k0fon$5pr@mochi.lava.net>, arti@lava.DOTnet (Art Isbell - remove "DOT") wrote: > >> nurban@crib.bevc.blacksburg.va.us (Nathan Urban) wrote: > >> > You're joking, right? Don't tell me that NT actually implements >> > whole-process _swapping_ over paging! I knew that NT VM performance >> > was bad, but not _that_ bad! This would explain much.. > >> Regardless of how NT implements VM, the disk seek noise and sluggishness >> when changing app focus seems incredible, especially considering that my PC >> has 96 MB of RAM. > >Oh, this is without doubt. I've got a 64 MB system at work and the >disk is _constantly_ grinding during various tasks.. and when I'm in >certain phases of compiling, it's sometimes all but impossible to even >switch apps! (Points to not just VM problems, but scheduler problems >too.) I don't know _how_ they could have implemented such a braindead >system, but my OPENSTEP for Mach box at home (on a slower system) blows >NT out of the water in terms of interactive performance. This surprises me, on a 80Mb Win95 box recompiling the 1Mb of source that is the Q2 dll takes no time at all and even with Word97,Excel97 & Acces97 all open my disk only ever spins when I open or save a file. Chris Hegarty
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: What is Carbon? Date: 23 May 1998 13:28:30 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6k70vu$ba9$1@crib.bevc.blacksburg.va.us> References: <6k702m$a0l$1@nclient1-gui.server.virgin.net> In article <6k702m$a0l$1@nclient1-gui.server.virgin.net>, "Chris Hegarty" <chris.hegarty@virgin.net> wrote: > I am new to NeXTSTEP/OPENSTEP coding and I want to know what this Carbon is, > PLEASE?? It is a stripped down and cleaned up subset of the old MacOS APIs, which will let developers of MacOS applications port their apps to MacOS X with minimal effort. The differences between the Carbon and MacOS APIs are sufficient to guarantee that Carbon apps, unlike MacOS apps, won't do anything too strange that prevents preemptive multitasking, protected memory, etc. from being applied. If you're interested in OpenStep coding then you're probably not too interested in Carbon, except for the fact that some of the useful APIs available in MacOS X might not have OpenStep equivalents.
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 23 May 1998 17:54:07 GMT Organization: Technical University of Berlin, Germany Message-ID: <6k72fv$m2q$1@news.cs.tu-berlin.de> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559EA34.B89B62D7@alumni.caltech.edu> <3559F204.3352@stanford.edu> <Pine.NXT.3.96.980513175733.12677B-100000@pathos> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Bill Bumgarner <bbum@codefab.com> writes: [Lots of interesting stuff on memory] >Threads suck. Threads suck when used like processes. In which case they really, really suck. Sadly this is the only use of threads widely documented. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 23 May 1998 18:01:20 GMT Organization: Technical University of Berlin, Germany Message-ID: <6k72tg$m45$1@news.cs.tu-berlin.de> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <6j7nlu$v51$1@crib.bevc.blacksburg.va.us> <35575DA8.B7A1148F@milestonerdl.com> <6j7te7$vah$1@crib.bevc.blacksburg.va.us> <6j9h1h$evm@fridge.shore.net> <1d8xvsi.14m93h87k4cz0N@pppa40-memphis9-2r431.saturn.bbn.com> <6jc50h$eh@fridge.shore.net> <3559DD44.3914@stanford.edu> <3559EE8D.DA8F1525@alumni.caltech.edu> <355A74AC.107@stanford.edu> <6je66b$pcj$1@news.idiom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit jcr.remove@this.phrase.idiom.com (John C. Randolph) writes: [handle discussion] >I don't know about OS's, but don't Postscript and Smalltalk both use >double-indirection to facilitate garbage collection? I haven't really kept up, but modern Smalltalks seem to use direct pointers instead of an object-table. Makes some operations much more expensive but many, many more operations slightly more efficient, which seems to come out as a win most of the time. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: don.brown@cesoft.com (Donald Brown) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Sat, 23 May 1998 13:27:39 -0500 Organization: Prairie Group Message-ID: <MPG.fd0d1146995e53798973f@news.supernews.com> References: <MPG.fc14db42395aa139896a2@news.supernews.com> <B18C6F81-2E6780@192.168.128.1> In article <B18C6F81-2E6780@192.168.128.1>, baskar@RemoveMe.snet.net says... > On Mon, May 11, 1998 8:03 PM, Donald Brown <mailto:don.brown@cesoft.com> > wrote: > >If you're not already working with Rhapsody/OpenStep/Yellow Box/etc., > >hold off until we hear more. The word I thought I heard this morning is > >that those are not the future, at least for client software. > > > > That is plain wrong information. Yellow box is here to stay. > You will see changes in the DPS-dependent parts of the AppKit, > and you will see more objects added to it for any new functionality. > Also, YB is THE API to use if you want to your app to run > on multiple platforms. Not if you want your app to run under MacOS 7.x or 8.x Donald
Subject: Re: New display model in OS-X From: "Baskaran Subramaniam" <baskar@RemoveMe.snet.net> Newsgroups: comp.sys.next.programmer References: <6k70pe$11q81@odie.mcleod.net> MIME-Version: 1.0 Message-ID: <B18D10FD-1EA95F@192.168.128.1> Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Date: Sun, 24 May 1998 03:54:52 GMT NNTP-Posting-Date: Sat, 23 May 1998 23:54:52 EDT On Sat, May 23, 1998 1:23 PM, Michelle L. Buck <mailto:buck.erik@mcleod.net> wrote: >IF a carbon >application chooses to lock down the entire screen it can. AND IF it then >loops infinitely and or crashes WITHOUT releasing the screen, you better >hope you have a terminal attached or can telnet in because the display is >locked! You will not be able to see the Processes panel to kill the Carbon >Application. Hmm... I don't understand how one would do that. If that is possible, then there should be some key combination that would bring up the process panel, sort of like how it is done in WindowsNT. Baskaran. PS: Please delete the string "RemoveMe." from my email address before replying to this message
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <10466895377625@digifix.com> Date: 24 May 1998 03:49:53 GMT Organization: Digital Fix Development Message-ID: <8873895982424@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1993. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: rex@smallandmighty.com (Eric King) Newsgroups: comp.sys.next.programmer Subject: Re: New display model in OS-X Date: Sun, 24 May 1998 03:47:59 -0400 Organization: The Small & Mighty Group Message-ID: <rex-2405980347590001@192.168.0.3> References: <6joa23$ik26@odie.mcleod.net> <6jp3hu$5mr$1@news.tuwien.ac.at> In article <6jp3hu$5mr$1@news.tuwien.ac.at>, wilkie@cg.tuwien.ac.at (Alexander Wilkie) wrote: :> PDF is NOT Postscript and true WYSIWYG is probably dead until PDF printers :> become available. : :Which will probably happen sometime after hell has frozen over solidly. Oh, they'll become available, but hell will freeze over before they're popular. Though you've got to remember, you're losing 'Postscript Everywhere' not WYSIWYG. The two are *not* the same. :> EPS will now only be supported as bitmap thumbnails. : :I'm speechless. Is Lawson English the new CTO at Apple? Of course not, otherwise the new display system would be based on GX ;) You see the fundamental problem is that you guys just didn't believe us GX advocates when we said that Adobe could be a most uncooperative company at times. They don't want to support anything that might compete with their product lines. From Adobe's standpoint, DPS did nothing but make it easier for folks to make Photoshop, Illustrator, and Pagemaker killers. Worse it would do it across their major platforms. I don't think Apple ever had a chance at coming to reasonable licensing terms. Adobe probably thought the risk was too high. : I mean, they've got a :good (read: not perfect, but quite decent) working display model right now. :And they go out and discard it in favour of _this_? Are they crazy? No, Apple just really bends over backwards for their large developers. With the exception of Quicktime, Adobe simply does not want powerful 2D graphics technology on the Mac unless it is theirs. How else can you explain Apple's dropping of GX, DPS, and Taligent's 2D graphics system, with no reasonable replacements? That's *three* powerful imaging systems down the drain. I think the odds of Apple creating a good fourth are pretty darn small. They'll make something that works, but I don't expect it to turn the heads of many graphics developers. -Eric
From: HollyRobinson@ThisSite.com Newsgroups: comp.sys.next.programmer Subject: SEE OLYMPIC SKATERS NUDE 41480 Date: Saturday, 23 May 1998 21:28:34 -0600 Organization: <no organization> Distribution: World Message-ID: <23059821.2834@ThisSite.com> http://members.coolnet.net/~tyler12 =======ALL FOR FREE======= YOU HEVER TO SEE THIS TO BELIEVE IT. 4 HOT COLLEGE GIRLS LIVING IN A HOUSE WITH CANS IN EVERY ROOM INCLUDING THE SHOWER. THE CAMS ARE ON 24 HOURS AND THERE IS ALSO A CHAT ROOM SO YOU CAN TALK TO THEM TOO... http://members.coolnet.net/~tyler12 http://members.coolnet.net/~tyler12 t6ny
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <23059821.2834@ThisSite.com> Control: cancel <23059821.2834@ThisSite.com> Date: 24 May 1998 15:18:40 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.23059821.2834@ThisSite.com> Sender: HollyRobinson@ThisSite.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: HollyRobinson@ThisSite.com Newsgroups: comp.sys.next.programmer Subject: SEE OLYMPIC SKATERS NUDE 82093 Date: Sunday, 24 May 1998 06:09:50 -0600 Organization: <no organization> Distribution: World Message-ID: <24059806.0950@ThisSite.com> http://members.coolnet.net/~tyler12 =======ALL FOR FREE======= YOU HEVER TO SEE THIS TO BELIEVE IT. 4 HOT COLLEGE GIRLS LIVING IN A HOUSE WITH CANS IN EVERY ROOM INCLUDING THE SHOWER. THE CAMS ARE ON 24 HOURS AND THERE IS ALSO A CHAT ROOM SO YOU CAN TALK TO THEM TOO... http://members.coolnet.net/~tyler12 http://members.coolnet.net/~tyler12 ZfAX
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <24059806.0950@ThisSite.com> Control: cancel <24059806.0950@ThisSite.com> Date: 24 May 1998 19:24:50 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.24059806.0950@ThisSite.com> Sender: HollyRobinson@ThisSite.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Shimpei Yamashita <shimpei+usenet+.mil+.gov@BOFH.patnet.caltech.edu> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: 24 May 1998 19:24:15 GMT Organization: Hummingbird Heaven Message-ID: <6k9s4v$r0e@gap.cco.caltech.edu> References: <see-below-1605981620340001@209.24.241.16> <joe.ragosta-1705981730440001@elk85.dol.net> <Pine.NXT.3.96.980520161114.26221R-100000@pathos> <6jvp3d$62e$2@news.idiom.com> Originator: shimpei@mirkwood.patnet.caltech.edu (Shimpei Yamashita) John C. Randolph <jcr.remove@this.phrase.idiom.com> writes: >Of course, if you take a bunch of those simple, common UNIX programs like sed >and the like out of the product, then you'll have a needlessly crippled >system. > >/bin on my OpenStep 4.2 system is 4.2 megs. Thats about four cents worth of >storage. The UNIX /bin directory on Rhapsody is similar. It ain't broke, >but if Apple tries to "fix" it by taking out a bunch of these tools, it will >be a hell of a lot of work. Hell? Why? All it takes is for one person with a gcc cross-compiler and access to the Rhapsody headers to compile a few crucial tools from GNU or NetBSD (gcc, binutils, textutils, make, sed, grep, etc.) and post the tarball somewhere. It won't be no work, but I hardly expect this to be hellish, especially if the Rhapsody kernel interface is really anything like 4.4BSD. -- Shimpei Yamashita <http://www.patnet.caltech.edu/%7Eshimpei/> perl -w -e '$_="not a perl hacker\n";$q=qq;(.);x9;$qq=qq;345123h896789,;;;$s= pack(qq;H6;,q;6a7573;);$qq=qq;s,^$q,$s$qq;;$qq=~s;(\d);\$$1;g;eval$qq;print;'
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: 24 May 1998 21:33:32 GMT Organization: Technical University of Berlin, Germany Message-ID: <6ka3nc$46c$1@news.cs.tu-berlin.de> References: <see-below-1605981620340001@209.24.241.16> <joe.ragosta-1705981730440001@elk85.dol.net> <Pine.NXT.3.96.980520161114.26221R-100000@pathos> <6jvp3d$62e$2@news.idiom.com> <6k9s4v$r0e@gap.cco.caltech.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Shimpei Yamashita <shimpei+usenet+.mil+.gov@BOFH.patnet.caltech.edu> writes: >John C. Randolph <jcr.remove@this.phrase.idiom.com> writes: >>Of course, if you take a bunch of those simple, common UNIX programs like sed >>and the like out of the product, then you'll have a needlessly crippled >>system. >> >>/bin on my OpenStep 4.2 system is 4.2 megs. Thats about four cents worth of >>storage. The UNIX /bin directory on Rhapsody is similar. It ain't broke, >>but if Apple tries to "fix" it by taking out a bunch of these tools, it will >>be a hell of a lot of work. >Hell? Why? All it takes is for one person with a gcc cross-compiler >and access to the Rhapsody headers to compile a few crucial tools from >GNU or NetBSD (gcc, binutils, textutils, make, sed, grep, etc.) and >post the tarball somewhere. It won't be no work, but I hardly expect >this to be hellish, especially if the Rhapsody kernel interface is >really anything like 4.4BSD. The point is LCD. Having the base set of tools installed on all machines means it will be there when you need it: emergency remote sysadmin tasks, using external filters in your apps. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: "William Colbert" <wcolbert@tiac.net> Newsgroups: comp.sys.next.programmer Subject: Emacs key bindings in Project Builder for OPENSTEP 4.2 Windows NT Date: Sun, 24 May 1998 20:13:54 -0400 Organization: The Internet Access Company, Inc. Message-ID: <6kact3$ut@news-central.tiac.net> I am having trouble figuring out how to get Project Builder to use Emacs keybindings in OPENSTEP 4.2 on WindowsNT. I have tried to set "LeftControl = Control" in NextLibrary/Frameworks/AppKit.framework/Resources/StandardKeyBinding-winnt.d ict and that had no effect. I have also tried setting this in several other .dict files, also with no effect. If in fact it is possible to configure the key bindings, I'm obviously missing some piece of the puzzle, and any further information would be much appreciated. Thanks, Will Colbert
From: xPornStars@FamousChicksxx.com Newsgroups: comp.sys.next.programmer Subject: SEE OLYMPIC SKATERS NUDE 66610 Date: Sunday, 24 May 1998 21:36:48 -0600 Organization: <no organization> Distribution: World Message-ID: <24059821.3648@FamousChicksxx.com> http://cyberrealm.net/~dr777/ =======ALL FOR FREE======= YOU HEVER TO SEE THIS TO BELIEVE IT. 4 HOT COLLEGE GIRLS LIVING IN A HOUSE WITH CANS IN EVERY ROOM INCLUDING THE SHOWER. THE CAMS ARE ON 24 HOURS AND THERE IS ALSO A CHAT ROOM SO YOU CAN TALK TO THEM TOO... http://cyberrealm.net/~dr777/ http://cyberrealm.net/~dr777/ LY4J
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <24059821.3648@FamousChicksxx.com> Control: cancel <24059821.3648@FamousChicksxx.com> Date: 25 May 1998 00:38:32 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.24059821.3648@FamousChicksxx.com> Sender: xPornStars@FamousChicksxx.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
Subject: sedmail 8.9 compile fails under OS4.2 Newsgroups: comp.sys.next.programmer,comp.mail.sendmail From: nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) Followup-To: comp.sys.next.programmer Message-ID: <_o3a1.478$ON2.3297033@news.rdc1.nj.home.com> Date: Mon, 25 May 1998 01:00:42 GMT NNTP-Posting-Date: Sun, 24 May 1998 18:00:42 PDT Organization: @Home Network Hrm... this doesn't look promising :-( Any hints appreciated.... here's what happened when I did 'make' without any modifications to the sendmail source for 8.9 cc -o sendmail alias.o arpadate.o clock.o collect.o conf.o convtime.o daemon.o deliver.o domain.o envelope.o err.o headers.o macro.o main.o map.o mci.o mime.o parseaddr.o queue.o readcf.o recipient.o safefile.o savemail.o snprintf.o srvrsmtp.o stab.o stats.o sysexits.o trace.o udb.o usersmtp.o util.o version.o -ldbm /bin/ld: warning alias.o has external relocation entries in non-writable section (__TEXT,__text) for symbols: _printf _strchr _free _strcasecmp _fgetc _fgets _ungetc __filbuf __ctype_ _errno _fstat _geteuid _stat _strpbrk _memset _fprintf _strcpy _strcat _strncmp and then later: cp /dev/null unistd.h echo "#include <sys/dir.h>" > dirent.h echo "#define dirent direct" >> dirent.h cc -O -I. -I../../src -DNeXT -Wno-precomp -pipe -c rmail.c rmail.c: In function `main': rmail.c:326: `STDIN_FILENO' undeclared (first use this function) rmail.c:326: (Each undeclared identifier is reported only once rmail.c:326: for each function it appears in.) rmail.c:368: request for member `w_S' in something not a structure or union rmail.c:368: request for member `w_T' in something not a structure or union *** Exit 1 Stop. *** Exit 1
From: root@127.0.0.1 (Charlie Root) Newsgroups: comp.sys.next.programmer Subject: Re: sedmail 8.9 compile fails under OS4.2 Date: 25 May 1998 01:56:56 GMT Message-ID: <6kaj58$ig6$0@207.212.27.63> References: <_o3a1.478$ON2.3297033@news.rdc1.nj.home.com> On Mon, 25 May 1998 01:00:42 GMT, Timothy Luoma <nospam+yes-this-is-a-valid_address@luomat.peak.org> wrote: > >Hrm... this doesn't look promising :-( > >Any hints appreciated.... here's what happened when I did 'make' >without any modifications to the sendmail source for 8.9 The file src/README clearly states: ********************* !! DO NOT USE MAKE !! in this directory to compile sendmail -- ********************* instead, use the "Build" script located in the src directory. It will build an appropriate Makefile, and create an appropriate obj.* subdirectory so that multiplatform support works easily. It builds fine on NEXTSTEP 3.3 and OPENSTEP 4.2.
Subject: Re: sedmail 8.9 compile fails under OS4.2 Newsgroups: comp.sys.next.programmer References: <_o3a1.478$ON2.3297033@news.rdc1.nj.home.com> <6kaj58$ig6$0@207.212.27.63> In-Reply-To: <6kaj58$ig6$0@207.212.27.63> From: nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) Message-ID: <wP4a1.481$ON2.3332424@news.rdc1.nj.home.com> Date: Mon, 25 May 1998 02:37:16 GMT NNTP-Posting-Date: Sun, 24 May 1998 19:37:16 PDT Organization: @Home Network On 05/24/98, Charlie Root wrote: >On Mon, 25 May 1998 01:00:42 GMT, Timothy Luoma ><nospam+yes-this-is-a-valid_address@luomat.peak.org> wrote: >> >>Hrm... this doesn't look promising :-( >> >>Any hints appreciated.... here's what happened when I did 'make' >>without any modifications to the sendmail source for 8.9 > >The file src/README clearly states: > >********************* >!! DO NOT USE MAKE !! in this directory to compile sendmail -- >********************* instead, use the "Build" script located in >the src directory. It will build an appropriate Makefile, and >create an appropriate obj.* subdirectory so that multiplatform >support works easily. > >It builds fine on NEXTSTEP 3.3 and OPENSTEP 4.2. The README is also a little outdated, as the Makefile is basically a front-end to the Build script now. However, just to see, I tried it again under 4.2 using Build # ./Build Configuration: os=NeXT, rel=4.2, rbase=4, rroot=4.2, arch=I386, sfx= Making in obj.NeXT.4.2.I386 cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c conf.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c convtime.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c daemon.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c deliver.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c domain.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c envelope.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c err.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c headers.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c macro.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c main.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c map.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c mci.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c mime.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c parseaddr.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c queue.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c readcf.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c recipient.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c safefile.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c savemail.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c snprintf.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c srvrsmtp.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c stab.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c stats.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c sysexits.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c trace.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c udb.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c usersmtp.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c util.c cc -O -I. -DNDBM -DNIS -DNETINFO -DNeXT -Wno-precomp -pipe -c version.c cc -o sendmail alias.o arpadate.o clock.o collect.o conf.o convtime.o daemon.o deliver.o domain.o envelope.o err.o headers.o macro.o main.o map.o mci.o mime.o parseaddr.o queue.o readcf.o recipient.o safefile.o savemail.o snprintf.o srvrsmtp.o stab.o stats.o sysexits.o trace.o udb.o usersmtp.o util.o version.o -ldbm /bin/ld: Undefined symbols: _hosts_ctl _putenv _unsetenv *** Exit 1 Stop. so that's fewer errors, to be sure, but it still fails. It will compile under NS3.3 without any errors.... However I want to use some Frameworks which are only available under OpenStep.... Then again, I can't even figure out how to do '-DTCPWRAPPERS ' so maybe I ought to just quit while I'm behind....
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <130942233171449856@mailexcite.com> Control: cancel <130942233171449856@mailexcite.com> Date: 25 May 1998 03:06:56 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.130942233171449856@mailexcite.com> Sender: nobody@gatekeeper.transre.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Ranjit Deshpande <ranjitd@pacbell.net> Newsgroups: comp.sys.next.hardware,comp.sys.next.programmer Subject: NeXTstation TurboColor Documentation Date: Sun, 24 May 1998 22:10:19 +0000 Organization: Pacific Bell Internet Services Message-ID: <35689ACA.E4CB4E1B@pacbell.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; x-mac-type="54455854"; x-mac-creator="4D4F5353" Content-Transfer-Encoding: 7bit Is there any possibility of me getting hold of some hardware documentation for the NeXTstation TurboColor ? I need the low down dirty stuff... General architecture, chip level stuff (how to program the TMC, register addresses of the NCR SCSI chip, etc.) that sort of thing. I have looked really hard and come up empty handed. The only promising lead was the "NeXT Bible" by Doug Clapp, but that book's out of print so... I would appreciate any kind of help from all you NeXT gurus. Thanks, Ranjit
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <1998052507544800.DAA25183@ladder01.news.aol.com> Date: 25 May 1998 07:54:48 GMT Organization: AOL http://www.aol.com References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <<< All you produced was a bunch of whining about how you can't do anything outside the lines, which is of course patently ludicrous, not to mention completely unsubstantiated. >>> Ok, how about something as simple as manipulating raw image pixels in an offscreen buffer? (say a median filter) It's a pretty basic thing to want to do in image processing apps, isn't it? However, it's not exactly simple or straightforward getting to the pixels in the Yellowbox, now is it? Sure you can get there, and I did, but there are plenty of other instances I ran into in working with it that left me wondering why bother. Every time someone tries to sell me on the yellowbox, they show me how you can build a "Full featured text editor" with three lines of code! I see it for exactly what it is: Bullshit. I can do the same damn thing in PowerPlant too, just as fast. (Or MacApp, MFC, or Visual Basic for that matter) But who needs a text editor? How useful is a canned demo using the prebuilt-classes and stitching them together? The NS classes are all built on a single tree model with no multiple inheritance. I like multiple inheritance and mix in classes, they make my life a lot easier. Sure, there are some ways to overcome it in Objective-C but it's not the same thing. Also, where do I find an Objective C programmer to add on my staff. Most of the world speaks C++ these days, with a minority of other languages. The other stickling issue with YellowBox is the fact that it may *never* hit the shelves. It wouldn't be the first time Apple reversed course leaving developers in the lurch. I for one am taking an attitude of indifference toward it. When I see it ship, I will consider writing for it. Last year Apple says the YB for Windows will be royalty free. This year, it's "minimal cost." What will it be next year? As far as speed goes, I ran OpenStep on PPro200 with plenty of RAM and NT was much snappier. Sorry, but that's the truth. I ran the OpenStep for NT and found that the app's were often notably slower feeling and less responsive than the native Win32 apps. I ran DR1 and DR2 on Apple hardware. I was willing to forgive DR1 completely. However I am not satisfied with DR2's runtime performance. Applications take forever to launch and the system is often unresponsive leaving the little "spinning rainbow" cursor up for several minutes at a time. This is on a 200mhz604e with 64megs of ram. It also thrashed the disk a lot too. Now, if you are gonna tell me that 64megs is ram starvation, I'll tell you that Rhapsody is a dead dog and a complete failure. OK, it is a DR/2 release, but even running OpenStep release for Intel on a well equipped PC left me very underwhelmed. The MacOS API's may be sloppy and antiquated and I won't argue that. However it's the only way to target the Mac today and quite possibly ever. -- Dr. Nuketopia Read the Blue Glow in Tubes FAQ at http://www.persci.com/~larrysb This address gets lots-o-spam. Please note that your letter is *not* spam in the subject line.
From: xPornStars@FamousChicksxx.com Newsgroups: comp.sys.next.programmer Subject: SEE OLYMPIC SKATERS NUDE 52171 Date: Monday, 25 May 1998 05:52:58 -0600 Organization: <no organization> Distribution: World Message-ID: <25059805.5258@FamousChicksxx.com> http://cyberrealm.net/~dr777/ ================================ *****TOTALLY FREE MEMBERSHIP**** ================================ ***MUST SEE SPYCAM CELEB PICS*** **CAST OF THE MOVIE THE CRAFT SNEAK PIX** *Cameron Diaz, Mira Sorvino, Rosie Perez* *Nicole Eggert, Samantha Fox, Hellen Hunt* *MARISA TOMEI, TIA LEONIE, PAULA ABDULE* *SPICE GIRLS, JEWEL, HOLLY ROBINSON* *U.S. OLYMPIC GYMNAST LOCKER ROOM PIX* *NEVE CAMPBELL, FIONA APPLE* *ENTIRE BEVERLY HILLS 90120 CAST* PLUS HIDDEN SPY CAMS FROM ALL OVER THE WORLD ================================ *****TOTALLY FREE MEMBERSHIP**** ================================ http://cyberrealm.net/~dr777/ http://cyberrealm.net/~dr777/ ?L'=
From: xPornStars@FamousChicksxx.com Organization: <no organization> Newsgroups: comp.sys.next.programmer Subject: rm <25059805.5258@FamousChicksxx.com> Message-ID: <cancel.25059805.5258@FamousChicksxx.com> Control: cancel <25059805.5258@FamousChicksxx.com> References: <25059805.5258@FamousChicksxx.com> Date: Mon, 25 May 1998 09:23:44 +0100 By this advisory message, hweede@berlin.snafu.de recommends the local removal of a spam whose Breidbart index is 120. The above header does not state who has issued this message or the spam itself. DO NOT REPLY TO ANY OF THESE ADDRESSES. See report "cyberrealmnetdr7-05250923" in news.lists.filters. Subject was: SEE OLYMPIC SKATERS NUDE 52171
From: jurgen@oic.de (Juergen Moellenhoff) Newsgroups: comp.sys.next.programmer Subject: Re: sedmail 8.9 compile fails under OS4.2 Date: 25 May 1998 10:55:24 GMT Organization: OIC, Bochum, Germany Message-ID: <6kbims$127$1@nexus.oic.de> References: <_o3a1.478$ON2.3297033@news.rdc1.nj.home.com> <6kaj58$ig6$0@207.212.27.63> <wP4a1.481$ON2.3332424@news.rdc1.nj.home.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit In <wP4a1.481$ON2.3332424@news.rdc1.nj.home.com> Timothy Luoma wrote: > However, just to see, I tried it again under 4.2 using Build > cc -o sendmail alias.o arpadate.o clock.o collect.o conf.o > convtime.o daemon.o deliver.o domain.o envelope.o err.o headers.o > macro.o main.o map.o mci.o mime.o parseaddr.o queue.o readcf.o > recipient.o safefile.o savemail.o snprintf.o srvrsmtp.o stab.o > stats.o sysexits.o trace.o udb.o usersmtp.o util.o version.o -ldbm > /bin/ld: Undefined symbols: > _hosts_ctl > _putenv > _unsetenv > *** Exit 1 > Stop. > > so that's fewer errors, to be sure, but it still fails. > > It will compile under NS3.3 without any errors.... However I want to > use some Frameworks which are only available under OpenStep.... 'sh Build' works for me. I compiled it yesterday and had no problems (OPENSTEP 4.2). Bye, Jürgen
From: schuerig@acm.org (Michael Schuerig) Newsgroups: comp.sys.next.programmer Subject: Embedding scripts in applications? Date: Mon, 25 May 1998 13:51:55 +0200 Organization: Completely Disorganized Message-ID: <1d9l70g.1n2j4061m6adm0N@rhrz-isdn3-p42.rhrz.uni-bonn.de> Mail-Copies-To: never I'm still an utter newbie regarding Openstep/Rhapsody/YB programming. Once upon a time I went through the first example in the developer tutorial. Now, as soon as RDR2 arrives, I'd like to do something real. My idea involves that I need to embed scripts in my app. That is, users will be able to specify scripts that are triggered by events in the app. I don't want to prescribe any particular scripting language, though, but leave it up to the users' preferences. On Mac OS this is handled by the Open Scripting Architecture (OSA), that provides generic ways to interact with so-called scripting components that implement a concrete scripting language (e.g., AppleScript, Userland Frontier). Is there a recommended way to do this in the Yellow Box? Michael -- Michael Schuerig mailto:schuerig@acm.org http://www.uni-bonn.de/~uzs90z/
From: Kevin Kratzke <news.remove@this.phrase.rwe.com> Newsgroups: comp.sys.next.programmer Subject: Re: how can I study? Date: Mon, 25 May 1998 08:58:05 -0700 Organization: Robert William Environmental, Inc. Message-ID: <3569950D.1A9BD081@this.phrase.rwe.com> References: <3565F3A1.794B@cs.utexas.edu> <6k6n24$avg$3@news.idiom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: jcr.remove@this.phrase.idiom.com I have "Nextstep Programming" by Simson Garfilkel & Michael Mahoney for Nextstep 2 & 3.0. Having no experience in C programming and a multitude of other things to work on, I never got that far and may be willing to part with it. Currently I have 3.3 running on Intel. Kevin Kratzke John C. Randolph wrote: > Makoto Sadahiro may or may not have said: > -> Hi. I just got NeXT Cube with NS3.3dev. What would be a first > -> recommended reading? I have bit experience programming C++ on unix, but > -> have no idea how to develop software on NeXT. I kind of started to > -> blowse around online docs in /lib. > -> > -> If you can directly me for some good reading, will be really > -> apprecited. > > Start with /NextDeveloper/Bookshelves/Developer.bshlf > > Look for the interface builder tutorial, and the Objective-C introduction. > > -jcr
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: rhapsody for intel and "carbon" Date: 25 May 1998 17:34:24 GMT Organization: Idiom Communications Message-ID: <6kca30$hb1$1@news.idiom.com> References: <see-below-1605981620340001@209.24.241.16> <joe.ragosta-1705981730440001@elk85.dol.net> <Pine.NXT.3.96.980520161114.26221R-100000@pathos> <6jvp3d$62e$2@news.idiom.com> <6k9s4v$r0e@gap.cco.caltech.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: shimpei+usenet+.mil+.gov@BOFH.patnet.caltech.edu Shimpei Yamashita may or may not have said: -> John C. Randolph <jcr.remove@this.phrase.idiom.com> writes: -> >Of course, if you take a bunch of those simple, common UNIX programs like sed -> >and the like out of the product, then you'll have a needlessly crippled -> >system. -> > -> >/bin on my OpenStep 4.2 system is 4.2 megs. Thats about four cents worth of -> >storage. The UNIX /bin directory on Rhapsody is similar. It ain't broke, -> >but if Apple tries to "fix" it by taking out a bunch of these tools, it will -> >be a hell of a lot of work. -> -> Hell? Why? All it takes is for one person with a gcc cross-compiler -> and access to the Rhapsody headers to compile a few crucial tools from -> GNU or NetBSD (gcc, binutils, textutils, make, sed, grep, etc.) and -> post the tarball somewhere. It won't be no work, but I hardly expect -> this to be hellish, especially if the Rhapsody kernel interface is -> really anything like 4.4BSD. Putting the UNIX utils back isn't the work I was talking about. I was talking about making the system boot, run its rc's, run its cron jobs, etc. without them. -jcr
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 25 May 1998 09:55:13 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6kbt81$1da$1@crib.bevc.blacksburg.va.us> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> In article <1998052507544800.DAA25183@ladder01.news.aol.com>, larrysb@aol.com (LarrySB) wrote: > <<< > All you produced was a bunch of whining about how you can't do anything > outside the lines, which is of course patently ludicrous, not to mention > completely unsubstantiated. > >>> > Ok, how about something as simple as manipulating raw image pixels in an > offscreen buffer? (say a median filter) How about NSDirectScreen and the Interceptor classes? > It's a pretty basic thing to want to do in image processing apps, isn't it? > However, it's not exactly simple or straightforward getting to the pixels in > the Yellowbox, now is it? Sure you can get there, and I did, but there are > plenty of other instances I ran into in working with it that left me wondering > why bother. Like what? > Every time someone tries to sell me on the yellowbox, they show me how > you can build a "Full featured text editor" with three lines of code! > I see it for exactly what it is: Bullshit. Sounds like you've got a lot of ignorance mixed in with a seriously large chip on your shoulder. Ever use OpenStep to build a big app? > I can do the same damn thing in PowerPlant too, just as fast. (Or MacApp, MFC, > or Visual Basic for that matter) With _nowhere near_ the same functionality. > But who needs a text editor? How useful is a > canned demo using the prebuilt-classes and stitching them together? If you're doing something with text manipulation, it's pretty useful! If not, there are plenty of other useful classes. It's not like text editors are the only thing you can build! > The NS classes are all built on a single tree model with no multiple > inheritance. Thank God. > I like multiple inheritance and mix in classes, they make my > life a lot easier. Sure, there are some ways to overcome it in Objective-C > but it's not the same thing. "It's different and therefore bad." > Also, where do I find an Objective C programmer to add on > my staff. Most of the world speaks C++ these days, with a minority of other > languages. You know how long it took me to learn all the subtleties of Objective-C? About a week. > The other stickling issue with YellowBox is the fact that it may *never* hit > the shelves. Ah yes, the convenient out of anyone arguing against an Apple technology. > Last year Apple says the YB for Windows will be royalty free. This year, it's > "minimal cost." What will it be next year? Free. > As far as speed goes, I ran OpenStep on PPro200 with plenty of RAM and NT was > much snappier. Sorry, but that's the truth. Well, that's awfully funny, I run NT on my PPro200 at work and OPENSTEP for Mach 4.2 on my Pentium/120 at home, and _OPENSTEP_ is much snappier, at least for interactive use. > I ran the OpenStep for NT and found that the app's were often notably slower > feeling and less responsive than the native Win32 apps. I've noticed OpenStep on NT being kind of sluggish. I think it's some sort of weird interaction with the window server. OPENSTEP for Mach, the operating system, doesn't suffer from this problem. > I ran DR1 and DR2 on Apple hardware. I was willing to forgive DR1 completely. > However I am not satisfied with DR2's runtime performance. Applications take > forever to launch and the system is often unresponsive leaving the little > "spinning rainbow" cursor up for several minutes at a time. What the heck are you doing to it??? My 4.2 system doesn't _ever_ do that. Maybe they've still got some bugs to work out. > The MacOS API's may be sloppy and antiquated and I won't argue that. However > it's the only way to target the Mac today and quite possibly ever. If you want to get left behind and outproduced, that's your problem. Meanwhile, I'll keep telling new developers to make the intelligent choice.
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 25 May 1998 18:07:30 GMT Organization: Idiom Communications Message-ID: <6kcc12$hb1$3@news.idiom.com> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: larrysb@aol.com LarrySB may or may not have said: -> The NS classes are all built on a single tree model with no multiple -> inheritance. I like multiple inheritance and mix in classes, they make my life -> a lot easier. Sure, there are some ways to overcome it in Objective-C but it's -> not the same thing. Also, where do I find an Objective C programmer to add on -> my staff. Most of the world speaks C++ these days, with a minority of other -> languages. First point: Multiple Inheritance is a bug, not a feature. It introduces ambiguity in method lookup, and can force subclasses to figure out which of multiple parents a method has been inherited from. MI may make the life of a lazy designer easier, but it complicates the task of anyone who has to maintain your code. Learn to use containment and delegation. Second point: A competent C programmer should be able to learn Obj-C in three to five days. There is much less intrusion on the C language to add the OO support of Obj-C, than there is to add the cruft of C++. -> The other stickling issue with YellowBox is the fact that it may *never* -> hit the shelves. It wouldn't be the first time Apple reversed course -> leaving developers in the lurch. I for one am taking an attitude of -> indifference toward it. When I see it ship, I will consider writing for -> it. It has already "hit the shelves." I have it, my customers have it, all of NeXT's old customers have it, and you can buy as many copies of OpenStep Enterprise as you want, right now. There are some classes added in RDR, but none have gone away from OpenStep 4.2 to Rhapsody that I know of. Keep in mind, that the Apple you know to be an unreliable business partner has been taken over by NeXT. -jcr
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 25 May 1998 15:25:12 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6kc2go$5tr$3@ns3.vrx.net> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: nurban@crib.bevc.blacksburg.va.us In <6kbt81$1da$1@crib.bevc.blacksburg.va.us> Nathan Urban claimed: > > I can do the same damn thing in PowerPlant too, just as fast. (Or MacApp, MFC, > > or Visual Basic for that matter) > > With _nowhere near_ the same functionality. Which is the real issue. With a near-future version you'll get for free... AppleScript ColorSync support _everywhere_ Strong text manipulation _everywhere_ no tasking issues to worry about great table support (could be better) tie-ins to EOF which is completely untouchable You get none of these with any of the frameworks he mentions. Not even close. Every single one of them is based on thin wrappers around the outdated OS packages - and they offer this as a "feature". > > The NS classes are all built on a single tree model with no multiple > > inheritance. > > Thank God. Hear hear. > > life a lot easier. Sure, there are some ways to overcome it in Objective-C > > but it's not the same thing. > > "It's different and therefore bad." In fact it's worth noting that the ONLY implementation available that uses MI on the Mac is (at least last time I looked) PowerPlant. PP has a lot of nice features but it's absolutely nothing whatsoever like YB's objects. However worth noting is the PP uses mixins in the same fashion Basically he's complaining that the object classes are single-inherited. Apparently this is supposed to be bad by itself I suppose. Of course any framework sucks if you design it poorly whether or not is has MI, but at the same time I've never heard YB described as having a bad design. > > Also, where do I find an Objective C programmer to add on > > my staff. Most of the world speaks C++ these days, with a minority of other > > languages. > > You know how long it took me to learn all the subtleties of Objective-C? > About a week. Five days for me. Never coded a line of _any_ C-a-like prior to day 1 either. > > As far as speed goes, I ran OpenStep on PPro200 with plenty of RAM and NT was > > much snappier. Sorry, but that's the truth. > > Well, that's awfully funny, I run NT on my PPro200 at work and OPENSTEP > for Mach 4.2 on my Pentium/120 at home, and _OPENSTEP_ is much snappier, > at least for interactive use. Me too. Can't figure out what he's talking about. NT for me always goes into swapping (even moving the mouse does it some times) and that never happens on my OS box. > If you want to get left behind and outproduced, that's your problem. > Meanwhile, I'll keep telling new developers to make the intelligent > choice. Keep it up! Maury
From: chris Newsgroups: comp.sys.next.programmer Subject: Rhapsody and jdk classpaths... Date: 25 May 1998 15:18:48 -0700 Organization: Newsguy News Service [http://www.newsguy.com] Message-ID: <6kcqo8$943@drn.newsguy.com> Hi, I am trying to write some java applications on rhapsody. DR2 seemed to ship without java release notes, or I cant find them. Where do you set the classpath for java, meaning where are the libraries at? This is driving me nuts. chris ------------------ Spam free Usenet news http://www.newsguy.com
From: mpaque@wco.com (Mike Paquette) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 25 May 1998 16:39:52 -0700 Organization: Electronics Service Unit No. 16 Message-ID: <1d9le4y.149xnlg656vwgN@carina25.wco.com> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.com> <6j84dp$vln$1@crib.bevc.blacksburg.va.us> <1998052308052000.EAA21989@ladder01.news.aol.com> <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> LarrySB <larrysb@aol.com> wrote: > Ok, how about something as simple as manipulating raw image pixels in an > offscreen buffer? (say a median filter) > > It's a pretty basic thing to want to do in image processing apps, isn't it? > However, it's not exactly simple or straightforward getting to the pixels in > the Yellowbox, now is it? All you need to do is build an NSBitmapImageRep. This cam be initialized with data from a file, or from a view (region) of a window, just by locking focus on the view and loading the bitmap rep. You can bang on the bits all you want, and them put them on the display or save to a file with one line of code. Got a file format that isn't built-in? Just construct a NSImageRep for that format, and all the stock tricks will work with your new format. Also, coming soon, the QTMLImageRep, which will leverage QuickTime to support anything that QTML supports. -- Mike Paquette mpaque@wco.com "Troubled Apple Computer" and the "Troubled Apple" logo are trade and service marks of Apple Computer, Inc.
Subject: Re: sedmail 8.9 compile fails under OS4.2 Newsgroups: comp.sys.next.programmer References: <_o3a1.478$ON2.3297033@news.rdc1.nj.home.com> <6kaj58$ig6$0@207.212.27.63> <wP4a1.481$ON2.3332424@news.rdc1.nj.home.com> <6kbims$127$1@nexus.oic.de> In-Reply-To: <6kbims$127$1@nexus.oic.de> From: nospam+yes-this-is-a-valid_address@luomat.peak.org (Timothy Luoma) Message-ID: <x6pa1.553$ON2.3853330@news.rdc1.nj.home.com> Date: Tue, 26 May 1998 01:42:53 GMT NNTP-Posting-Date: Mon, 25 May 1998 18:42:53 PDT Organization: @Home Network On 05/25/98, Juergen Moellenhoff wrote: > 'sh Build' works for me. > > I compiled it yesterday and had no problems (OPENSTEP 4.2). Weird.... Fresh install here.... I wonder what's up.... TjL
From: arti@lava.DOTnet (Art Isbell - remove "DOT") Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: 26 May 1998 04:36:57 GMT Organization: LavaNet, Inc. Distribution: world Message-ID: <6kdgt9$cn2@mochi.lava.net> References: <6junig$is5$2@ns3.vrx.net> <Pine.NXT.3.96.980520161846.26221T-100000@pathos> <6jvfa8$2pj$1@crib.bevc.blacksburg.va.us> <6k0fon$5pr@mochi.lava.net> <6k0haj$41v$1@crib.bevc.blacksburg.va.us> <6k70d6$a4o$1@nclient1-gui.server.virgin.net> "Chris Hegarty" <chris.hegarty@virgin.net> wrote: > > Nathan Urban wrote in message <6k0haj$41v$1@crib.bevc.blacksburg.va.us>... > >I've got a 64 MB system at work and the > >disk is _constantly_ grinding during various tasks.. and when I'm in > >certain phases of compiling, it's sometimes all but impossible to even > >switch apps! (Points to not just VM problems, but scheduler problems > >too.) I don't know _how_ they could have implemented such a braindead > >system, but my OPENSTEP for Mach box at home (on a slower system) blows > >NT out of the water in terms of interactive performance. > > This surprises me, on a 80Mb Win95 box recompiling the 1Mb of source that is > the Q2 dll takes no time at all and even with Word97,Excel97 & Acces97 all > open my disk only ever spins when I open or save a file. Maybe the difference in experiences can be explained by the amount of file system fragmentation on each system. Those of us accustomed to using a UNIX file system aren't accustomed to the need to defragment file systems so probably don't do so nearly as often as those familiar with Windows. Frankly, file system design should deal with fragmentation rather than requiring users to deal with it. UNIX file systems have been doing this for years, so I was shocked when Windows "New Technology" offered old file system design, even with NTFS. -- Art Isbell NeXT/MIME Mail: arti@lavaDOTnet IDX Systems Corporation Voice/Fax: +1 808 394 0511 (for whom I don't speak) Voice Mail: +1 808 394 0495 Healthcare Info Technology US Mail: Honolulu, HI 96825-2638
From: dunham@cs.tulane.edu (Andrea Dunham) Newsgroups: comp.sys.next.programmer Subject: C++ Libraries on Developer 3.3 Date: 26 May 1998 06:49:28 GMT Organization: Tulane University Message-ID: <6kdolo$hoe$1@rs10.tcs.tulane.edu> Hello All, I have two possibly related questions about running C++ on Dev3.3. ===== QUESTION #1 I am a newbie to Developer 3.3 and I am unable to run this simple program #include <stream.h> main(){ cout << "Hello World"; } I tried % cc++ hello.cpp and I get the error messages: ld: Undefined symbols: ostream::operator<<(char const *) ostream::operator<<(int) cout ===== QUESTION #2 I hope you can help. I'm running Developer 3.3 and tried to build the GNU source but after about 30 minutes... == building shlib version `specs' is up to date. rm -f tmplibgcc1.a for name in _mulsi3 _udivsi3 _divsi3 _umodsi3 _modsi3 _lshrsi3 _lshlsi3 _ashrsi3 _ashlsi3 _divdf3 _muldf3 _negdf2 _adddf3 _subdf3 _fixdfsi _fixsfsi _floatsidf _floatsisf _truncdfsf2 _extendsfdf2 _addsf3 _negsf2 _subsf3 _mulsf3 _divsf3 _eqdf2 _nedf2 _gtdf2 _gedf2 _ltdf2 _ledf2 _eqsf2 _nesf2 _gtsf2 _gesf2 _ltsf2 _lesf2; do echo ${name}; rm -f ${name}.o; ./xgcc -B./ -DSHLIB -O -DSHLIB -I. -I/NextDeveloper/Source/GNU/cc -I/NextDeveloper/Source/GNU/cc/config -c -DL${name} /NextDeveloper/Source/GNU/cc/libgcc1.c; if [ $? -eq 0 ] ; then true; else exit 1; fi; mv libgcc1.o ${name}.o; ar qc tmplibgcc1.a ${name}.o; rm -f ${name}.o; done _mulsi3 /NextDeveloper/Source/GNU/cc/libgcc1.c:35: header file 'libsys/shlib.h' not found *** Exit 1 Stop. *** Exit 1 Stop. Any suggestions on what I should do or where I can begin reading to solve these problems? Thanks for your help, * AndREa * (dunham@eecs.tulane.edu)
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <130942234221523456@mailcity.com> Control: cancel <130942234221523456@mailcity.com> Date: 26 May 1998 06:01:45 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.130942234221523456@mailcity.com> Sender: nobody@gatekeeper.transre.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: spammers@ruin.the.internet.channelu.com Newsgroups: comp.sys.next.hardware,comp.sys.next.programmer Subject: Re: NeXTstation TurboColor Documentation Date: 26 May 1998 05:09:56 GMT Organization: Michigan State University Message-ID: <6kdir4$fl0$1@msunews.cl.msu.edu> References: <35689ACA.E4CB4E1B@pacbell.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: ranjitd@pacbell.net In <35689ACA.E4CB4E1B@pacbell.net> Ranjit Deshpande wrote: > Is there any possibility of me getting hold of some hardware > documentation for the NeXTstation TurboColor ? I need the low down dirty > stuff... General architecture, chip level stuff (how to program the TMC, > register addresses of the NCR SCSI chip, etc.) that sort of thing. I > have looked really hard and come up empty handed. The only promising > lead was the "NeXT Bible" by Doug Clapp, but that book's out of print > so... > > I would appreciate any kind of help from all you NeXT gurus. > > Thanks, > Talk to NCR about the 53C90A :( General architecture stuff would be found in various articles (Byte, and some brochures), but I'm sure won't be as complete as you would like. Check out the following http://www.ratol.fi/~jisokung/NeXT/brochures/index.html (based on) http://iris.dissvcs.uga.edu/~archive/NeXT/NeXT.html and http://www.slip.net/~emclean/next/article.html For some preliminary info. There is a Linux page for Black hardware that basically only has the addresses of some of the hardware devices. (nothing else I could see). I can't seem to find the url at this moment. Someday I hope to have these and just a bit more under http://www.channelu.com/NeXT/Black There is a htmlified brochure (that should download quickly) @ http://www.channelu.com/NeXT/History/NeXTPub/N6030/index.html Other goodies I have will pop up periodically. Keep an eye out on my Log http://www.channelu.com/Log/index.html (i.e when ScanOmatic gets Astra 600S support, and I get used to Tiffany and Create for creating the images, and the web pages. Then hopefully I will get a little more productive.) Randy rencsok at channelu dot com argus dot cem dot msu dot edu spammers works also :) Randy Rencsok General UNIX, NeXTStep, IRIX Admining, Turbo Software Consulting, Programming, etc.)
Message-ID: <356ABB02.C14AE66B@iphysiol.unil.ch> Date: Tue, 26 May 1998 14:52:18 +0200 From: Sean Hill <noone@iphysiol.unil.ch> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer To: pdo@omnigroup.com Subject: Problem with Frameworks dynamic linking on PDO Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I'm using PDO 4.x on Solaris. I have installed the frameworks I've built into ~/Library/Frameworks/. I specify this path in the -F argument in the Makefile. I then specify the frameworks I need using the -framework argument. However during linking it warns that the frameworks are not found. It compiles successfully so it is finding the header files in those frameworks in the right place. And when I check the linked shared libraries using ldd, it shows the framework as /MyFramework.framework/Versions/A/MyFramework => not found When I launch the program it says (twice): ld.so.1: ./MyProgram: fatal: /MyFramework.framework/Versions/A/MyFramework: can't open file: errno=2. Why is it looking at the root directory for the framework? How can I make it truly link against the right shared library (framework)? Thanks very much for your help. Sean Hill
From: cms@macisp.net (cms) Newsgroups: comp.sys.next.programmer Subject: Wanted: ISP's using Next Comptuer System !! Date: Fri, 22 May 1998 18:30:41 -0500 Organization: CMS Message-ID: <cms-2205981830560001@209.26.71.138> We are looking for any ISP that are using the Next Computer in a Internet Service Business. please send email. Richard CMS cms@macisp.net
Message-ID: <356ABAE8.BA423A19@iphysiol.unil.ch> Date: Tue, 26 May 1998 14:51:52 +0200 From: Sean Hill <noone@iphysiol.unil.ch> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer To: pdo@omnigroup.com Subject: Problem with Frameworks dynamic linking on PDO Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I'm using PDO 4.x on Solaris. I have installed the frameworks I've built into ~/Library/Frameworks/. I specify this path in the -F argument in the Makefile. I then specify the frameworks I need using the -framework argument. However during linking it warns that the frameworks are not found. It compiles successfully so it is finding the header files in those frameworks in the right place. And when I check the linked shared libraries using ldd, it shows the framework as /MyFramework.framework/Versions/A/MyFramework => not found When I launch the program it says (twice): ld.so.1: ./MyProgram: fatal: /MyFramework.framework/Versions/A/MyFramework: can't open file: errno=2. Why is it looking at the root directory for the framework? How can I make it truly link against the right shared library (framework)? Thanks very much for your help. Sean Hill
From: embuck@palmer.cca.rockwell.com (Erik M. Buck) Newsgroups: comp.sys.next.programmer Subject: Re: NSHosting and MacOS X Date: 26 May 1998 13:33:33 GMT Organization: Rockwell International Distribution: world Message-ID: <6kegbd$m612@onews.collins.rockwell.com> References: <6k4d8f$ge0$1@news.cs.tu-berlin.de> <6k5evi$l6j$1@agate.berkeley.edu> <6k6duu$qou$1@news.seicom.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: frank@this.NO_SPAM.net In <6k6duu$qou$1@news.seicom.net> Frank M. Siegert wrote: > In <6k5evi$l6j$1@agate.berkeley.edu> Izumi Ohzawa wrote: > > ... > > Sadly, this is not what we got. Why not, why not, why not? > > Mh, just speculating... > > - The original model is not based on DPDF alone but for performance (or > whatever) sake Quickdraw and QT can draw to physical regions on the screen > thus remote hosting wouldn't work for (some?) Carbon based applications. > > - Because it is not considered to be of any importance. > > Unless Apple changes its mind, vector based remote display is gone. The key to vector based remote display is that vector drawing commands are serialized over an inter-process/network connection between a client and a server. Apple's new display model removes all vector rendering from the server thus making it ONLY a "region" handler. With that decision, remote device independence is lost, remote display is lost, and remote event handling is lost. The only feasible remote display mechanism suggested by Apple's stated graphics architecture would be network copying of device dependant bitmaps.
From: Tim Triemstra <nospam@nospam.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Tue, 26 May 1998 10:17:51 -0700 Organization: PM Global Foods, LLC Message-ID: <356AF93F.CB6E1927@nospam.com> References: <MPG.fc14db42395aa139896a2@news.supernews.com> <B18C6F81-2E6780@192.168.128.1> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Baskaran Subramaniam wrote: > > On Mon, May 11, 1998 8:03 PM, Donald Brown <mailto:don.brown@cesoft.com> > wrote: > >If you're not already working with Rhapsody/OpenStep/Yellow Box/etc., > >hold off until we hear more. The word I thought I heard this morning is > >that those are not the future, at least for client software. > > > > That is plain wrong information. Yellow box is here to stay. > You will see changes in the DPS-dependent parts of the AppKit, > and you will see more objects added to it for any new functionality. > Also, YB is THE API to use if you want to your app to run > on multiple platforms. I'd say you'd be hard pressed to show anything released from Apple since WWDC that shows his statement to be completely wrong. Although Apple may not have said "Yellow Box API's are dead" they haven't made much of a commitment to it lately either. When you consider that this was the "World Wide Developer's Convention" - think about that name - you'd think SOMETHING would have come to the press regarding future yellow development. Perhaps you haven't had experience with NeXT before, but this is somewhat par for the course. The "conversion" tools used for NeXT 3.x -> OpenStep 4.x were miserable, leaving many apps in the dust when you consider the amount of work and limited market. Now that Carbon is being pushed to the primary ISVs, how many do you really think will switch staffing over to learn Yellow Box? If companies like Adobe and Microsoft aren't writing new Mac apps using Yellow, who will and for how long. How hard will Apple try to support Yellow developers when their primary money making ISVs are using Carbon. I'm not saying there aren't compelling reasons to write for Yellow (I was looking forward to it on Intel and Macs) but there have been compelling reasons to use Objective C or NeXTSTEP for a long time too. Without Apple really pushing them their won't be the support, and Apple is far from "pushing" Yellow APIs as of late. 'Course, I still haven't received my Intel version of DR2, now two weeks after the PowerPC version... -- Tim Triemstra . TimT at PMGLOBAL dot COM PM Global Foods, LLC ... Atlanta GA USA My home page: www.mindspring.com/~timtr
Date: Tue, 26 May 1998 09:54:06 -0400 From: wmoss@harland.net (William Moss) Newsgroups: comp.sys.next.programmer Subject: Lotus Notes and Enterprise Objects Message-ID: <wmoss-2605980954060001@bstephens.harland.net> Organization: Harland (The Studio) Greetings all, I've been lurking on this newsgroup for a while and soaking in the collective knowledge; I am totally fascinated with the potential of YellowBox but as of yet have no version of Rhapsody or OpenStep with which to play. So I've resigned myself to be a mere "scholarly observer" at this point. My questions are addressed to those who are intimately familiar with the Enterprise Objects Framework. I'm the "computer guy" in the creative department of a rather large corporation whose I.T. department has standardized on Lotus Notes, Peoplesoft, and Win95. Being so small and of such alien computer needs, my department been pretty much overlooked and been allowed to self-manage our own computer needs... until now. The directive has come down to interface our data with the corporate Lotus Notes databases. With our current DBMS, this is not possible. So either we will need to switch database managers or we should give our needs over to the I.T. department. I'd rather not do this. My users consists of artists and creative people so the user interface and printed reports are of tatamount importance to ensure the integrity of the data (if they think it's a pain to enter hours, they'll put it off for weeks and weeks and make it up at the last minute). From what I've seen of other departments, the sensitivity to user needs had been lacking. Would the Enterprise Object Framework facilitate the communication of yellow box apps with a Lotus Notes database? Has this been done before or is this "experimental ground"? Are there any Yellowbox developed databases that are "EOF-enabled"? Beyond Stepwise and the Apple website, are there other resources available describing some EOF capabilities and/or success stories? I did already pick up the book "Developing Business Applications with OpenStep" by Gervae and Clark. What else might I want to try to hunt down? Thanks William
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: Re: NSHosting and MacOS X Newsgroups: comp.sys.next.programmer References: <6k4d8f$ge0$1@news.cs.tu-berlin.de> <6k5evi$l6j$1@agate.berkeley.edu> <6k6duu$qou$1@news.seicom.net> <6kegbd$m612@onews.collins.rockwell.com> Distribution: world Message-ID: <356ae230.0@news.depaul.edu> Date: 26 May 98 15:39:28 GMT Erik M. Buck <embuck@palmer.cca.rockwell.com> wrote: >Unless Apple changes its mind, vector based remote display is gone. The key >to vector based remote display is that vector drawing commands are serialized >over an inter-process/network connection between a client and a server. >Apple's new display model removes all vector rendering from the server thus >making it ONLY a "region" handler. With that decision, remote device >independence is lost, remote display is lost, and remote event handling is >lost. The only feasible remote display mechanism suggested by Apple's stated >graphics architecture would be network copying of device dependant bitmaps. While Apple's not planning to use the client-server WindowServer drawing scheme, they need not use a monolithic drawing system. I suppose that Apple could insert a 'disconnect' inside the graphics system. It won't be client-server between processes, but there could be a similar separation within the graphics subsystem, within an application. That would at least provide some potential for remote display. Each application could have a front-end graphics engine, and a back-end graphics engine which would manage the screen memory and mark it. They would normally both be in the same process. If you put the back-end engine on a different machine, then it could manage screen memory and draw on a different machine's screen. This is similar to NSHosting, but different, because you'd actually need an executing process on the display machine. Perhaps this could be handled by a generic back-end process? If you call a drawing function in the new graphics API, why shouldn't it internally be sent off to another machine? It becomes tricky if you're drawing directly on screen bits, but if not, it shouldn't be a problem. Wouldn't this be similar to the way DPS worked on a Dimension? The DPS interpreter on the 040 would boil the DPS down to primitives, and send those to the Dimension for drawing, IINM. MacOS X could use a similar system. - Jon -- "... and subpoenas for all." - Ken Starr
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <1998052616533300.MAA26118@ladder01.news.aol.com> Date: 26 May 1998 16:53:33 GMT Organization: AOL http://www.aol.com References: <6kbt81$1da$1@crib.bevc.blacksburg.va.us> nurban@crib.bevc.blacksburg.va.us (Nathan Urban) wrote: :Sounds like you've got a lot of ignorance mixed in with a seriously :large chip on your shoulder. Ever use OpenStep to build a big app? Nope. Have you? I have however shipped big applications based on MFC and PowerPlant, among others. You may call them ugly, but you can't ignore the fact that they do actually work. :> I can do the same damn thing in PowerPlant too, just as fast. (Or MacApp, MFC, :> or Visual Basic for that matter) :With _nowhere near_ the same functionality. All the functionality except that which is unique to display postscript. However, I can assure you that the text edit classware in PowerPlant easily implements extensive style options. :> The NS classes are all built on a single tree model with no multiple :> inheritance. :Thank God. Says you. But that's a whole 'nother holy war isn't it. :> The other stickling issue with YellowBox is the fact that it may *never* hit :> the shelves. :Ah yes, the convenient out of anyone arguing against an Apple technology. Obviously, you don't understand the principle of business risk. Right now, YB is vaporware with no assurance that it will even ship as a mainstream product. It's been promised, yes, but then again, so were GX graphics, Newton, OpenDoc, Copland and so on. All leaving behind them a wake of early adopters busted and broke. :> Last year Apple says the YB for Windows will be royalty free. This year, it's :> "minimal cost." What will it be next year? :Free. Hey bucko, I *was* at WWDC and I saw the slides that clearly stated that YB for Windows would be available for "minimum cost." Indeed last year, they were royalty free. So which is it? You obviously don't have the answer. The truth is that we won't know until we see it. :> The MacOS API's may be sloppy and antiquated and I won't argue that. However :> it's the only way to target the Mac today and quite possibly ever. :If you want to get left behind and outproduced, that's your problem. :Meanwhile, I'll keep telling new developers to make the intelligent :choice. In other words, you don't have a freakin clue and you wish to spread your ignorance around. Good for you. YB may be the greatest thing since sliced bread. But until it's wrapped up and on the shelves and the end-user's computers it has no value or relevance.
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 26 May 1998 10:15:10 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6keipe$35e$1@crib.bevc.blacksburg.va.us> References: <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <1998052616533300.MAA26118@ladder01.news.aol.com> In article <1998052616533300.MAA26118@ladder01.news.aol.com>, larrysb@aol.com (LarrySB) wrote: > nurban@crib.bevc.blacksburg.va.us (Nathan Urban) wrote: > :Sounds like you've got a lot of ignorance mixed in with a seriously > :large chip on your shoulder. Ever use OpenStep to build a big app? > Nope. Have you? Not extremely large, no. Others have and can comment on the time it saved them. > I have however shipped big applications based on MFC and PowerPlant, among > others. You may call them ugly, but you can't ignore the fact that they do > actually work. <yawn> I do MFC at work. Elsewhere I posted an _enormous_ list of Yellow Box technologies that would have saved us man-years of time in our application if we had had them. > :> I can do the same damn thing in PowerPlant too, just as fast. (Or MacApp, > :> MFC, or Visual Basic for that matter) > :With _nowhere near_ the same functionality. > All the functionality except that which is unique to display postscript. Not even remotely! You are _very_ ignorant. Go study the NSText* classes again. I think Maury mentioned a feature comparison elsewhere. > :> The other stickling issue with YellowBox is the fact that it may *never* > :> hit the shelves. > :Ah yes, the convenient out of anyone arguing against an Apple technology. > Obviously, you don't understand the principle of business risk. Right now, YB > is vaporware with no assurance that it will even ship as a mainstream product. It's already shipped in the form of OPENSTEP Enterprise and is used extensively in a number of large MCCA applications. > It's been promised, yes, but then again, so were GX graphics, Newton, OpenDoc, > Copland and so on. I love it how everyone lumps all those technologies together. They were all different technologies, and dumped for varying reasons. You have to examine the factors involved rather than painting them all with the same broad brush. And Yellow Box makes an extreme amount of business sense for Apple. > :> Last year Apple says the YB for Windows will be royalty free. This year, > :> it's "minimal cost." What will it be next year? > :Free. > Hey bucko, I *was* at WWDC and I saw the slides that clearly stated that YB > for Windows would be available for "minimum cost." Indeed last year, they > were royalty free. They said that they were going to make it free once the new graphics system is in place. > :If you want to get left behind and outproduced, that's your problem. > :Meanwhile, I'll keep telling new developers to make the intelligent > :choice. > In other words, you don't have a freakin clue and you wish to spread your > ignorance around. I have a far better clue than you, as you have so conveniently proved. > YB may be the greatest thing since sliced bread. But until it's wrapped up > and on the shelves and the end-user's computers it has no value or relevance. Wait two years to adopt it and that's just two more years of lost time for you.
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 26 May 1998 18:10:37 GMT Organization: Idiom Communications Message-ID: <6kf0it$7pn$1@news.idiom.com> References: <6kcc12$hb1$3@news.idiom.com> <1998052617163600.NAA23784@ladder03.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: larrysb@aol.com LarrySB may or may not have said: -> :Keep in mind, that the Apple you know to be an unreliable business partner -> :has been taken over by NeXT. -> -> Now that's a twist I hadn't heard. Now let me get this straight, you say that -> NeXT bought out Apple? I didn't say bought out, I said taken over. Look at the roster of top Apple executives: VP for software development: Avie Tevanian VP of Sales: Mitch Mandich VP for Hardware development: Jon Rubinstein (interim) CEO: Steve Jobs And so on. -> The MacOS API's may represent inertia, but sometimes that's the smartest place -> to put your time and money. If it's inertia you want, try writing COBOL business apps for MVS. That market is still huge, and still growing at better than 10% yearly. For my part, I prefer to use tools that don't make me want to puke. That's why I went to the Mac in 1984, and it's why I went to NeXTStep in 1989. If NeXT and Apple hadn't merged, I'd probably be a Smalltalker right now. -jcr
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <1998052617163600.NAA23784@ladder03.news.aol.com> Date: 26 May 1998 17:16:36 GMT Organization: AOL http://www.aol.com References: <6kcc12$hb1$3@news.idiom.com> :First point: Multiple Inheritance is a bug, not a feature. It introduces :ambiguity in method lookup, and can force subclasses to figure out which of :multiple parents a method has been inherited from. MI may make the life of a :lazy designer easier, but it complicates the task of anyone who has to :maintain your code. Learn to use containment and delegation. That's the stupidest fucking thing I've ever heard. Properly used, MI greatly simplifies design and maintainence. Having used both Power Plant and MFC, I can tell you firsthand that MI is much cleaner and easier to deal with. I've used it with great advantage and improved update cycles in *production* code. At least C++ *allows* the option of either. :Second point: A competent C programmer should be able to learn Obj-C in three :to five days. There is much less intrusion on the C language to add the OO :support of Obj-C, than there is to add the cruft of C++. I didn't find Obj-C terribly difficult to get a grasp of. However, learning a language's subtleties takes a long time. Many many people know C++, as bad as it can be. But it's a devil we know. I can't tell you all the buga-boos of Objective C; that takes time. Take a productive programmer and tell him to learn Obj-C and the YB frameworks and you've got six months of learning curve. That's reality. Add that to the total uncertainty of YB ever being a real product and you've got an unacceptable level of risk. :It has already "hit the shelves." I have it, my customers have it, all of :NeXT's old customers have it, and you can buy as many copies of OpenStep :Enterprise as you want, right now. There are some classes added in RDR, but :none have gone away from OpenStep 4.2 to Rhapsody that I know of. But Rhapsody isn't shipping, yet. When it does there won't be any compelling reasons for MacOS users to buy it. Basically, a Mac person gets less from Rhapsody 1.0 than from MacOS 8.5. :Keep in mind, that the Apple you know to be an unreliable business partner :has been taken over by NeXT. Now that's a twist I hadn't heard. Now let me get this straight, you say that NeXT bought out Apple? From all the SEC filings, it would appear that Amelio's Apple bought NeXT. Jobs is running the show at Apple now and hopefully some sanity will return. (Jobs and sanity in the same sentence?) But how long does he remain in charge? When will he get tired or bored and take his marbles to play elesewhere? Then what happens? The MacOS API's may represent inertia, but sometimes that's the smartest place to put your time and money.
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: NSHosting and MacOS X Date: 22 May 98 09:10:04 Organization: Is a sign of weakness Message-ID: <SCOTT.98May22091004@slave.doubleu.com> References: <6k1uic$6dp$1@crib.bevc.blacksburg.va.us> <6k1vqo$itb@shelob.afs.com> In-reply-to: Greg_Anderson@afs.com's message of 21 May 1998 19:38:00 GMT In article <6k1vqo$itb@shelob.afs.com>, Greg_Anderson@afs.com (Gregory H. Anderson) writes: Nathan Urban writes > In article <SCOTT.98May18215210@slave.doubleu.com>, > scott@doubleu.com (Scott Hess) wrote: > > And _PLEASE_ do not tell me that in order to draw with my > > advanced object-oriented environment, I have to revert to a > > palpimset function-based API, > > I've run through all of the words, acronyms, and potential > misspellings I can think of, to no avail. So tell me.. what the > heck is "palpimset"?? I believe Scott meant to say "palimpsest", which according to Webster is "writing material (as a parchment or tablet) used one or more times after earlier writing has been erased". Yes! I _knew_ that someone would find me a correct spelling if only I posted. [Usually Webster's manages that for me :-).] As long as I'm here, I intended to refer to the fact that you can generally see that there was previous writing, but it's generally very hard to figure out what the writing _was_. In this case, I'm suggesting that systems like MacOS and Windows95/NT have many levels of API which address overlapping problems, with the reasoning behind each API often obscured by later revisions. Worse, under some systems the later APIs don't superset the similar earlier APIs, so often programming becomes a process of weeding out your choices. NeXT has been very good with a two-revisions-and-you're-out policy, painful for people using the older APIs, but keeping the APIs manageable for everyone else, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: 22 May 98 09:33:51 Organization: Is a sign of weakness Message-ID: <SCOTT.98May22093351@slave.doubleu.com> References: <6jvb8u$2do$1@crib.bevc.blacksburg.va.us> <B189C84A-3C2CE@141.214.128.36> <6k1o9p$66j$1@crib.bevc.blacksburg.va.us> In-reply-to: nurban@crib.bevc.blacksburg.va.us's message of 21 May 1998 13:29:29 -0400 In article <6k1o9p$66j$1@crib.bevc.blacksburg.va.us>, nurban@crib.bevc.blacksburg.va.us (Nathan Urban) writes: In article <B189C84A-3C2CE@141.214.128.36>, "Robert A. Decker" <comrade@umich.edu> wrote: > On Wed, May 20, 1998 2:34 PM, Nathan Urban > <mailto:nurban@crib.bevc.blacksburg.va.us> wrote: > > For some reason I seem to have really bad luck with OpenStep, > > though. :( > Mine goes down about twice a month. It's always OmniWeb that does > it. The entire machine locks up and I can't even telnet into > it... While the things that produce my lockups seem to be kind of unpredictable, it _does_ seem that more than the usual share of them occur with OmniWeb (even considering the fact that I use it more). I wonder if there's something to that.. It could be related to the fact that OmniWeb is multithreaded and banging on the windowserver heavily, which combination might provide one of the heavier loads on your CPU. It could be one of those things where the system is marginal, except when running a certain set of operations. Similar to how many people had zero problems in normal usage but got crashes when running one of the mprimes programs (which exercised the FPU), -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 26 May 1998 10:25:21 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6kejch$378$1@crib.bevc.blacksburg.va.us> References: <MPG.fc14db42395aa139896a2@news.supernews.com> <B18C6F81-2E6780@192.168.128.1> <356AF93F.CB6E1927@nospam.com> In article <356AF93F.CB6E1927@nospam.com>, Tim Triemstra <nospam@nospam.com> wrote: > Although Apple > may not have said "Yellow Box API's are dead" they haven't made much of > a commitment to it lately either. How do you justify that claim? > When you consider that this was the > "World Wide Developer's Convention" - think about that name - you'd > think SOMETHING would have come to the press regarding future yellow > development. Sheesh. You know as well as I do that the press only covered the opening keynote and one or two other things. You know that the whole Rhapsody thing was scaring Mac developers, hence the MacOS X strategy, hence Rhapsody was downplayed in the keynote. But if you actually look at the rest of the sessions, aimed at developers, you'd see that Apple is being _very_ active in updating and upgrading Yellow technologies, and wants new development to take place using it. > Now that Carbon > is being pushed to the primary ISVs, how many do you really think will > switch staffing over to learn Yellow Box? That is indeed a valid question, but it remains to be seen. I don't think that any of the major apps are going to be ported to Yellow, but _new_ apps may be written to it. Depends on how well developers can be educated.
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: NSHosting and MacOS X Date: 22 May 98 15:05:11 Organization: Is a sign of weakness Message-ID: <SCOTT.98May22150511@slave.doubleu.com> References: <1d93iik.1l170dz8etsaxN@carina47.wco.com> <01bd80c6$be64d100$a4e82080@mizuki.HIP.Berkeley.EDU> <6k4d8f$ge0$1@news.cs.tu-berlin.de> In-reply-to: marcel@cs.tu-berlin.de's message of 22 May 1998 17:39:27 GMT In article <6k4d8f$ge0$1@news.cs.tu-berlin.de>, marcel@cs.tu-berlin.de (Marcel Weiher) writes: Re-read Mike's post. He was very clear in separating (a) NSHost capability, the current protocol which requires DPS (b) remote display capability, which does not Supporting option (a) is not possible because there is no Postscript interpreter included with MacOS X. There seems to be no problem in supporting (b) except for available time and priorities. Additionally, there may be support for third parties to provide this capability. The "problem", such as it is, is that to support remote display, you effectively have to reinvent the ability to transmit the drawing and event data over some stream connection. DPS gets this basically for free, due to how it's designed (sending a stream of ASCII data over a socket is the height of trivial. Doing it efficiently is where the trouble is). The thing is, there's no real reason why they couldn't duplicate the DPS transmission layer, and redirect it to a different windowserver (which is where the _real_ Adobe stuff is). I'm willing to bet that you could even retain binary compatibility for 95% of the apps out there for NeXTSTEP/OpenStep. The things that you couldn't retain compatilibity with would be the windowPackage.ps, the AppKit, and certain specialized drawing apps. Heck, given the way that the DPS client library works, you wouldn't even have to have much in the way of a parser in the server - and at that, what's hard about an RPN parser? I'd estimate that if you started with a decent drawing subsystem, you could write a client/server comm layer modelled on DPS's client library in six months or so. The hard part isn't in the communications layer, it's in the drawing layer. Gosling did it with NeWS, after all, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: stefan.boehringer_REMOVE_ME_@ruhr-uni-bochum.de (Stefan Boehringer) Newsgroups: comp.sys.next.programmer Subject: Re: access to [NSString cString] functionality? Date: 26 May 1998 19:41:10 GMT Organization: Ruhr-Universitaet Bochum, Rechenzentrum Message-ID: <6kf5sm$c73$1@sun579.rz.ruhr-uni-bochum.de> References: <slrn6mlrr8.4s8.rog@talisker.ohm.york.ac.uk> >the -(const char *)cString method of NSString manages >to put a bare C string into an autorelease pool. >i would very much like to be able to do the same thing, >to return a UTF-8 string as a char *. does anyone >out there know how it's done? (surely they couldn't have >selfishly restricted this functionality to their own objects, >could they? :-)) [..] > [NSAutoreleasePool addNonObject:r]; // what should go here? Don't do that. Do the following instead: [NonObjectHolder createWithNonObject:r]; The NonObjectHolder class must return an autoreleased object which will then be deallocated with the next autoreleasepool cleaning. It will free 'r' just then. Best wishes, Stefan
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <1998052620073200.QAA13289@ladder01.news.aol.com> Date: 26 May 1998 20:07:32 GMT Organization: AOL http://www.aol.com References: <slrn6mlpv9.4s8.rog@talisker.ohm.york.ac.uk> rog@ohm.york.ac.uk (Roger Peppe) wrote: :On Mon, 25 May 1998 16:39:52 -0700, Mike Paquette <mpaque@wco.com> wrote: :> LarrySB <larrysb@aol.com> wrote: :> > Ok, how about something as simple as manipulating raw image pixels in an :> > offscreen buffer? (say a median filter) :> All you need to do is build an NSBitmapImageRep. This cam be :> initialized with data from a file, or from a view (region) of a window, :> just by locking focus on the view and loading the bitmap rep. You can :> bang on the bits all you want, and them put them on the display or save :> to a file with one line of code. :that's not as easy to do as you're making out. :sure, you can get access to the data, but there's no standard way of :addressing the pixels of data that you've got from an NSBitmapImageRep, :because they arrive in one of many possible formats - you can't know :which until you get it, and the manual page doesn't enumerate all the :possibilities. :for instance, if you're getting pixels in this way, your code has :to handle: : planar and non-planar data : any number of bits per pixel (where does it say that i might not be : given 13 bit per pixel data?) : any number of bytes per row :that's quite a lot of bit twiddling and nasty special cases :just to get at screen bitmap data in a portable form. :if they just provided a "- (NSColour *)getPixelAt:(NSPoint)position" :method, it would make a lot of things much easier, or a standard :way of indexing into data as returned from an NSBitmapImageRep, :it would make things a lot easier! :i think LarrySB has a point (if only a small one, because i'd hazard :that *most* applications have no need to do this) Thank you Roger. I was waiting for someone to pipe in with the correct NS terminology. You are correct, maybe most applications don't do image processing - but mine do. My bread and butter is in image processing, which means getting the raw pixels and doing math on them. I think you will find a lot of MacOS software which does image processing. The point about "- (NSColour *)getPixelAt:(NSPoint)position" would make it easier to program, but I strongly fear it would introduce undue overhead. Take a median filter for example. In the most basic scenario you collect values from a 3x3 pixel area, sort them and take the median value and replace the center pixel with that new value. Increasing the radius by one pixel to a 5x5 filter changes the problem from 9 pixels sampled and sorted to 25 pixels. A 7x7 kernal raises the stakes to 49 pixels. Multiply that by an average 800x600 image and you see how much overhead is suddenly introduced. A 5x5 median filter on a 800x600 image involves 12,000,000 pixel reads. (actually, there are better ways to do this, but lets keep it fairly simple) Consider the overhead of going through Objective C's runtime binding and dispatch mechanism and you've got the recipe for an unduly slow image processing task. The performance hit is simply unnacceptable. For reasonable performance even on fast processors like the G3, I have to keep my pixels in buffers in memory and directly access them with pointers. But that's not a big deal, since all my image acquisition and file handling code put images into a plain old Mac GWorld. So what if the MacOS api is procedural. The point is that it works. I just wrap up things like GWorlds in wrapper classes to make my life easier. Sure, it's not elegant from an academic standpoint, but it is completely functional. Since my wrapper class inherits from nothing else, I can port it back into my existing applications without dragging PowerPlant with it. What's neat is that my thinly wrapped class is completely compatible with the Toolbox too, so I can draw in it, copy to it, blit it to the screen, export to quicktime codecs and so on. I still don't know exactly how to keep my image data as raw pixels and yet make it 100% compatible with the rest of NS. I want to save them as tiff format data, with multiple images per file, along with other data. This is a pretty simple example of coloring outside the lines and how a framework can actually make things more difficult. I have run into plenty of other examples in my line of work too. For instance, I often work in color spaces outside of RGB, so I need to convert from one format to another, hammer the raw data and then convert it back. Here's another tough one: Direct hardware support. Yup, it's ugly, but I have no choice in the matter. Niche market? Yup. But then again, the Mac is a niche machine and OpenStep is a niche player too. I mean, how many major applications that shipped more than 10,000 seats were based on OpenStep? Before someone pips off, a *lot* of existing software touches hardware. Communication software uses the serial port. Games go direct to screen very often, get input from game controllers and use the sound hardware. Lots of graphics programs support scanners and digital cameras. Lots of data acquisition software directly interfaces with GPIB systems. It's not a rare thing. Even stuff like anti-piracy dongles need hardware support. There are a number of instances where I've used MI to great advantage too. You have to exercise good judgement when applying MI, but it does work very nicely. I've never run into conflicts, or had any trouble with other programmers figuring out what was going on. It's actually nice doing things like mixing an LView, LListener and an LPeriodical together. If anything, the code is cleaner and simpler to maintain. I don't advocate complex MI schemes with deep hierarchy. Good use of MI can greatly reduce the amount of duplicate code, which in turn greatly eases maintainence. Whether that agrees with someones OO-religion is beside the point. It is reality.
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 26 May 1998 16:03:48 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6mlpv9.4s8.rog@talisker.ohm.york.ac.uk> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.c <1d9le4y.149xnlg656vwgN@carina25.wco.com> On Mon, 25 May 1998 16:39:52 -0700, Mike Paquette <mpaque@wco.com> wrote: > LarrySB <larrysb@aol.com> wrote: > > Ok, how about something as simple as manipulating raw image pixels in an > > offscreen buffer? (say a median filter) > All you need to do is build an NSBitmapImageRep. This cam be > initialized with data from a file, or from a view (region) of a window, > just by locking focus on the view and loading the bitmap rep. You can > bang on the bits all you want, and them put them on the display or save > to a file with one line of code. that's not as easy to do as you're making out. sure, you can get access to the data, but there's no standard way of addressing the pixels of data that you've got from an NSBitmapImageRep, because they arrive in one of many possible formats - you can't know which until you get it, and the manual page doesn't enumerate all the possibilities. for instance, if you're getting pixels in this way, your code has to handle: planar and non-planar data any number of bits per pixel (where does it say that i might not be given 13 bit per pixel data?) any number of bytes per row that's quite a lot of bit twiddling and nasty special cases just to get at screen bitmap data in a portable form. if they just provided a "- (NSColour *)getPixelAt:(NSPoint)position" method, it would make a lot of things much easier, or a standard way of indexing into data as returned from an NSBitmapImageRep, it would make things a lot easier! i think LarrySB has a point (if only a small one, because i'd hazard that *most* applications have no need to do this) cheers, rog.
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: access to [NSString cString] functionality? Date: 26 May 1998 16:35:48 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6mlrr8.4s8.rog@talisker.ohm.york.ac.uk> the -(const char *)cString method of NSString manages to put a bare C string into an autorelease pool. i would very much like to be able to do the same thing, to return a UTF-8 string as a char *. does anyone out there know how it's done? (surely they couldn't have selfishly restricted this functionality to their own objects, could they? :-)) i.e. something like the following: @implementation NSString(UTF8CStringExtension) - (const char *)utf8string { char *r; int len; NSData *d = [self dataUsingEncoding:NSUTF8StringEncoding allowLossyConversion:YES]; len = [d length]; r = malloc(len + 1); [d getBytes:r]; r[len] = 0; [NSAutoreleasePool addNonObject:r]; // what should go here? return r; } @end thanks for any help, rog.
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: My _first_ reboot! Date: 22 May 98 09:30:07 Organization: Is a sign of weakness Distribution: world Message-ID: <SCOTT.98May22093007@slave.doubleu.com> References: <6junig$is5$2@ns3.vrx.net> <6jvb8u$2do$1@crib.bevc.blacksburg.va.us> <Pine.NXT.3.96.980520161846.26221T-100000@pathos> <6jvfa8$2pj$1@crib.bevc.blacksburg.va.us> <6k0fon$5pr@mochi.lava.net> In-reply-to: arti@lava.DOTnet's message of 21 May 1998 05:57:43 GMT In article <6k0fon$5pr@mochi.lava.net>, arti@lava.DOTnet (Art Isbell - remove "DOT") writes: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) wrote: > You're joking, right? Don't tell me that NT actually implements > whole-process _swapping_ over paging! I knew that NT VM > performance was bad, but not _that_ bad! This would explain > much.. Regardless of how NT implements VM, the disk seek noise and sluggishness when changing app focus seems incredible, especially considering that my PC has 96 MB of RAM. One thing that has always annoyed me about NT is lack of backing store. So when you close that maximized Netscape window, every other app on-screen has to be pulled back into memory to redraw it's window. I feel like I'm back in the 80's on a 16M cube swapping to optical. Worse, NT seems to be "tuned" for 32M or less memory size. All I have is my own experience to go on, here, starting from when I upgraded from 64M to 128M on my PPro. Linux got faster, NS/OS got faster, NT didn't change a bit. So far as I can tell, it seems that NT actively swaps secondary processes out to meet some free-memory goal, so that you effectively have only the active processes in memory at once. My guess is that this would allow launching a new process, or new allocations by the current process, to happen quickly. Unfortunately, when you have 128M, you generally have enough memory to keep _all_ processes in memory, even those you haven't used for awhile. I'd guess it's a filesystem buffering thing, except that most of the time I'm not manipulating more than a couple meg of data... Then I installed Diskkeeper Lite, a disk defragmenter. It reported 54% fragmentation! fsck typically reports 0.3% on my Mach systems. There is a really interesting site out there which elaborated on certain Windows NT VM and filesystem fragmentation problems. The fragmentation page is at http://coop.wec.ufl.edu/hydew/comp/NT/ntfsfrag.htm And remember, folks, NT is Microsoft's high-performance operating system :-) And they did it _all_ by themselves! Isn't that amazing? I mean, imagine the horrors if they had built on 20 years of operating system research. They'd have a secure, high-performance system... and instead of most everyone using NT, _everyone_ would be using NT. Of course, using the Windows UI on such an OS would be like running a Porsche on rims, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 26 May 1998 10:20:43 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6kej3r$366$1@crib.bevc.blacksburg.va.us> References: <6kcc12$hb1$3@news.idiom.com> <1998052617163600.NAA23784@ladder03.news.aol.com> In article <1998052617163600.NAA23784@ladder03.news.aol.com>, larrysb@aol.com (LarrySB) wrote: > :First point: Multiple Inheritance is a bug, not a feature. It introduces > :ambiguity in method lookup, and can force subclasses to figure out which of > :multiple parents a method has been inherited from. MI may make the life of a > :lazy designer easier, but it complicates the task of anyone who has to > :maintain your code. Learn to use containment and delegation. > That's the stupidest fucking thing I've ever heard. Properly used, MI greatly > simplifies design and maintainence. Having used both Power Plant and MFC, I > can tell you firsthand that MI is much cleaner and easier to deal with. Like he said: learn to use containment and delegation. You get pretty much the same thing, without all the conflict resolution hassles. > I've > used it with great advantage and improved update cycles in *production* code. > At least C++ *allows* the option of either. Yet another unnecessary complication of the language. Why do you think C++ is such a complex language? They keep adding all kinds of features that you don't really need. > Take a productive programmer and tell him to learn Obj-C and the YB frameworks > and you've got six months of learning curve. It didn't take me that long to be able to do most things. > That's reality. Add that to the > total uncertainty of YB ever being a real product and you've got an > unacceptable level of risk. Only if you think that there's "total uncertainty" involved, which is of course complete paranoia. > :Keep in mind, that the Apple you know to be an unreliable business partner > :has been taken over by NeXT. > Now that's a twist I hadn't heard. Now let me get this straight, you say that > NeXT bought out Apple? From all the SEC filings, it would appear that > Amelio's Apple bought NeXT. The interim CEO, the VP of software, and the VP of hardware are all ex-NeXT. Apple has NeXT employees in most of the high positions of power. > Jobs is running the show at Apple now and hopefully some > sanity will return. (Jobs and sanity in the same sentence?) But how long does > he remain in charge? When will he get tired or bored and take his marbles to > play elesewhere? Then what happens? I think it's unlikely that Jobs would get replaced with someone with a significantly different vision of the future anytime soon. > The MacOS API's may represent inertia, but sometimes that's the smartest place > to put your time and money. And sometimes it's not.
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 26 May 1998 18:21:33 GMT Organization: Idiom Communications Message-ID: <6kf17d$7pn$2@news.idiom.com> References: <6kcc12$hb1$3@news.idiom.com> <1998052617163600.NAA23784@ladder03.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: larrysb@aol.com LarrySB may or may not have said: -> :First point: Multiple Inheritance is a bug, not a feature. It introduces -> :ambiguity in method lookup, and can force subclasses to figure out which of -> :multiple parents a method has been inherited from. MI may make the life of a -> :lazy designer easier, but it complicates the task of anyone who has to -> :maintain your code. Learn to use containment and delegation. -> -> That's the stupidest fucking thing I've ever heard. Properly used, MI greatly -> simplifies design and maintainence. Well, if you believe that, then I'll make quite sure I never take a gig where I have to read any of your code. ->Having used both Power Plant and MFC, Ah! Sterling credentials for OO software development, indeed. Now, if you had trotted out some of the pro-MI arguments from the Smalltalk community, I might have given you some credence. -> I can tell you firsthand that MI is much cleaner and easier to deal with. -> I've used it with great advantage and improved update cycles in -> *production* code. At least C++ *allows* the option of either. And it allows the *option* of writing utterly incomprehensible, unmaintainable code. I've read a great deal of it, and often had to recommend a re-write. -> :Second point: A competent C programmer should be able to learn Obj-C in three -> :to five days. There is much less intrusion on the C language to add the OO -> :support of Obj-C, than there is to add the cruft of C++. -> -> I didn't find Obj-C terribly difficult to get a grasp of. However, learning a -> language's subtleties takes a long time. Many many people know C++, as bad as -> it can be. But it's a devil we know. I can't tell you all the buga-boos of -> Objective C; that takes time. C++ is ridiculously complex for the limited benefits it offers. I don't even consider it an OO language, and Stroustrup admitted as much when he first presented it to the ACM. C++ is C with a facility for adding typedefs that include overloading the aritmetic operators. Try using it for real OO development, and you quickly run headlong into its morass of accreted "features". Obj-C is as simple as it appears to be. I've used it for over seven years, and it's never thrown me a curve like when Stroustroup suddenly decided that under certain circumstances, your cleverly overloaded "=" operator won't get called if both operands are the same type. "C++ is to C, as Lung Cancer is to Lung." -> The MacOS API's may represent inertia, but sometimes that's the smartest place -> to put your time and money. Enjoy your intertia. I don't know what kinds of products your outfit does, but watch over your shoulder for a competitor who uses better tools. He'll eat your lunch. -jcr
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 26 May 1998 17:08:15 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6kfavv$3mg$1@crib.bevc.blacksburg.va.us> References: <slrn6mlpv9.4s8.rog@talisker.ohm.york.ac.uk> <1998052620073200.QAA13289@ladder01.news.aol.com> In article <1998052620073200.QAA13289@ladder01.news.aol.com>, larrysb@aol.com (LarrySB) wrote: > The point about "- (NSColour *)getPixelAt:(NSPoint)position" would make it > easier to program, but I strongly fear it would introduce undue overhead. It would be interesting to see Mike Paquette's comments on efficiency of doing direct drawing using the Yellow APIs. > This is a pretty simple example of coloring outside the lines and how a > framework can actually make things more difficult. Yes, but you're missing the point that I originally challenged you on: the framework doesn't _prevent_ you from coloring outside the lines. If the framework doesn't do what you need, then don't use it; not using it doesn't make your job any _harder_ than if the framework wasn't there at all! In fact, with the MacOS X imaging model, you'll be able to use _identical_ graphics code as what you're doing in MacOS, even from a Yellow app, because the app draws directly into a buffer instead of communicating with the window server to draw. Then you could use Yellow for all the other aspects of the app. > Here's another tough one: Direct hardware support. Writing your app in Yellow doesn't prevent you from doing that. It just means that you have to go outside the Yellow APIs to do it, and it won't be portable. > There are a number of instances where I've used MI to great advantage too. > [...] > Good use of MI can greatly reduce the amount of duplicate code, > which in turn greatly eases maintainence. Whether that agrees with someones > OO-religion is beside the point. It is reality. All true, but the fact of the matter is that you don't _need_ MI to accomplish those things. You can gain the code reuse and maintenance advantages without it, using things like interface ineritance (protocols) and delegation.
From: Bill Bumgarner <bbum@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Tue, 26 May 1998 17:23:22 -0400 Organization: Spacelab.net Internet Access Message-ID: <Pine.NXT.3.96.980526171858.9803A-100000@pathos> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.c <1d9le4y.149xnlg656vwgN@carina25.wco.com> <slrn6mlpv9.4s8.rog@talisker.ohm.york.ac.uk> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII To: Roger Peppe <rog@ohm.york.ac.uk> In-Reply-To: <slrn6mlpv9.4s8.rog@talisker.ohm.york.ac.uk> On 26 May 1998, Roger Peppe wrote: > On Mon, 25 May 1998 16:39:52 -0700, Mike Paquette <mpaque@wco.com> wrote: > > LarrySB <larrysb@aol.com> wrote: > > > Ok, how about something as simple as manipulating raw image pixels in an > > > offscreen buffer? (say a median filter) > > > All you need to do is build an NSBitmapImageRep. This cam be > > initialized with data from a file, or from a view (region) of a window, > > just by locking focus on the view and loading the bitmap rep. You can > > bang on the bits all you want, and them put them on the display or save > > to a file with one line of code. > > that's not as easy to do as you're making out. > > sure, you can get access to the data, but there's no standard way of > addressing the pixels of data that you've got from an NSBitmapImageRep, > because they arrive in one of many possible formats - you can't know > which until you get it, and the manual page doesn't enumerate all the > possibilities. Actually, the same holds true for a screen... is it 555, 444, 565, 888, or whatever? There are numerous different screen buffers... ...the key is writing routines that can effeciently twiddle the bits at the most common configurations. [... more interesting observations deleted ...] When working with NSBitmapImageReps, the key is to convert any given bitmap you are handed into whatever format your bit twiddling algorithms can deal with. Just because an image arrives as planar with five channels doesn't mean you can't easily derive a second bitmap image rep that is interleaved with three channels-- the API is all there and works quite well [if I remember correctly]. > i think LarrySB has a point (if only a small one, because i'd hazard > that *most* applications have no need to do this) Having built NUMEROUS extremely large and, in some cases, very successful OpenStep apps, the need to twiddle bits has never really been something I have needed to do. This includes apps that were 3D modelers and other graphically intense apps. In the few cases where I did have to twiddle bits, using NSBitmapImageReps to both convert data to a usable format and then twiddle away worked great. b.bum
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: Best development environment, a nitpick Date: 26 May 1998 18:52:40 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6kfh3o$40h$1@crib.bevc.blacksburg.va.us> References: <6kfcm3$gq$1@nnrp1.dejanews.com> In article <6kfcm3$gq$1@nnrp1.dejanews.com>, quinlan@intergate.bc.ca wrote: > Just a comment for those of you who are claiming the YB is the best > development environment on the market: you should really take a look at a > good smalltalk implementation and see if you still feel that way. You're right, this is something most of us (sometimes intentionally) ignore in our YB advocacy; we tend to compare YB to the _popular_ frameworks, to which it generally far superior. I have no doubt that Smalltalk environments are better in terms of the development tools and environment -- and in many respects the language, though sometimes I think it's a bit _too_ inefficient. (Normally I don't mind a little intrinsic inefficiency too much because in practice the development advantages outweigh it.. but of course you can go too far. OTOH, I've never programmed it directly, just heard some things secondhand from some Smalltalk people, so I don't know how bad it really is.) However, I don't know how the standard available Smalltalk frameworks compare to the design and functionality of YB.
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <1998052623235100.TAA06628@ladder01.news.aol.com> Date: 26 May 1998 23:23:51 GMT Organization: AOL http://www.aol.com References: <6kfavv$3mg$1@crib.bevc.blacksburg.va.us> nurban@crib.bevc.blacksburg.va.us (Nathan Urban) wrote: :> The point about "- (NSColour *)getPixelAt:(NSPoint)position" would make it :> easier to program, but I strongly fear it would introduce undue overhead. :It would be interesting to see Mike Paquette's comments on efficiency of :doing direct drawing using the Yellow APIs. It isn't a question of drawing API's, it's a question of accessing the pixel data. Obviously, it can be done or programs like Tiffany wouldn't work now would they? How else can you sharpen, blur, smooth, stretch and enhance image data? That's the crux of what I deal with. I also work with scientific-grade cameras which produce 16-bit deep monochrome data. I do quantitative analysis. :> This is a pretty simple example of coloring outside the lines and how a :> framework can actually make things more difficult. :Yes, but you're missing the point that I originally challenged you on: :the framework doesn't _prevent_ you from coloring outside the lines. If :the framework doesn't do what you need, then don't use it; not using it :doesn't make your job any _harder_ than if the framework wasn't there at :all! Where did I say impossible? The problem with YB is the API isn't an API, it's a framework. The authors of the framework decided that no one needs an easy way to get to raw pixel data and didn't put any way in to do it. The problem is it *does* make life harder. There's not an easy way to leverage the YB for this task, which means you have to do it all yourself. Advantage nil. :All true, but the fact of the matter is that you don't _need_ MI to :accomplish those things. You can gain the code reuse and maintenance :advantages without it, using things like interface ineritance (protocols) :and delegation. I have my doubts about supporting protocols, which is Obj-C's work-around for mix-ins. Don't you have to implement the protocol in each class that supports it?
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <1998052620100700.QAA13560@ladder01.news.aol.com> Date: 26 May 1998 20:10:07 GMT Organization: AOL http://www.aol.com References: <6keipe$35e$1@crib.bevc.blacksburg.va.us> nurban@crib.bevc.blacksburg.va.us (Nathan Urban) wrote: <<< <yawn> I do MFC at work. Elsewhere I posted an _enormous_ list of Yellow Box technologies that would have saved us man-years of time in our application if we had had them. >>> Why, pray tell, aren't you using Yellow Box then? There must be some reason.
From: stefan.boehringer_REMOVE_ME_@ruhr-uni-bochum.de (Stefan Boehringer) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 26 May 1998 20:31:56 GMT Organization: Ruhr-Universitaet Bochum, Rechenzentrum Message-ID: <6kf8rs$fbl$1@sun579.rz.ruhr-uni-bochum.de> References: <MPG.fc14db42395aa139896a2@news.supernews.com> <B18C6F81-2E6780@192.168.128.1> <356AF93F.CB6E1927@nospam.com> <6kejch$378$1@crib.bevc.blacksburg.va.us> nurban@crib.bevc.blacksburg.va.us (Nathan Urban) wrote: >In article <356AF93F.CB6E1927@nospam.com>, Tim Triemstra <nospam@nospam.com> wrote: >> Although Apple >> may not have said "Yellow Box API's are dead" they haven't made much of >> a commitment to it lately either. > >How do you justify that claim? While they haven't said so and I personally believe that YB will stay at least some 3 years from now on there is a another point to be learned. Those long term Next advocates have learned that when money becomes a problem the concepts begin to wobble as well (IMHO). Those who are left have accepted this behaviour until now and I personally have lived quite comfortably with it. However now Apple kills my favorite dev environment and forces me to proprietary hardware which is unacceptable to me (yes this is a NeXT experience). It's not a matter of buying a Mac but that the scope of my products, my choice of hardware, and the perception of the OS will be limited again and will fall short. Additionally I feel that YB is no longer a key technology for Apple. It may or may not survive and may be sacrificed on the altar of profit. While it's okay for Apple to be profitable it's not okay for me who doesn't get what he wants. This makes us separate (just me personally). Never believe that this is an easy step, there is still nothing which comes close and by now I can't see where to go. Don't get me wrong - it's my totally personal humble opinion. It's only posted because there is the very little hope that those who are responsible may change their mind to retain Rhapsody/MacOS X/whatever on intel. I promise to be indulgent and stay with Apple for the future to come. However some points must be respected (detailed above). For those who are happy with Mac hardware: cheer up, you'll get the best OS on the planet. Best wishes, Stefan
From: quinlan@intergate.bc.ca Newsgroups: comp.sys.next.programmer Subject: Best development environment, a nitpick Date: Tue, 26 May 1998 21:37:07 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6kfcm3$gq$1@nnrp1.dejanews.com> Just a comment for those of you who are claiming the YB is the best development environment on the market: you should really take a look at a good smalltalk implementation and see if you still feel that way. I'm not saying the YB is bad, just that working in a smalltalk environment is better. Why? My top three reasons are: 1) All types are objects. 2) No primitive source/include file model 3) Code modifications can be made and take effect while the code is running -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 26 May 1998 17:00:26 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6kfaha$3la$1@crib.bevc.blacksburg.va.us> References: <6keipe$35e$1@crib.bevc.blacksburg.va.us> <1998052620100700.QAA13560@ladder01.news.aol.com> In article <1998052620100700.QAA13560@ladder01.news.aol.com>, larrysb@aol.com (LarrySB) wrote: > nurban@crib.bevc.blacksburg.va.us (Nathan Urban) wrote: > <yawn> I do MFC at work. Elsewhere I posted an _enormous_ list of > Yellow Box technologies that would have saved us man-years of time in > our application if we had had them. > Why, pray tell, aren't you using Yellow Box then? There must be some reason. Because I came into the project two years after it had begun, no one there had even heard of Yellow Box until I told them about it, and two years ago the development costs of OpenStep development were too high anyway ($5000/seat for developer tools, plus an exorbitant Windows runtime fee; $20 would have been acceptable). If we were beginning this project today, I could quite likely convince management to approve YB development on the basis of a demo I could hack up in a couple of weeks. I could quite likely even convince my fellow development team, despite their investment in MFC; they, honestly, are very fed up with MFC development, and during our daily Microsoft gripe-sessions I point out YB would have solved our given problem-of-the-day more easily. In fact, a lot of our best designs are now based on things I've copied from OpenStep and explained to the team -- which we've had to implement ourselves, of course. Unfortunately, we cannot throw away years of code; even if it might be better in the long run, it's just completely unfeasible business-wise to stop all development of new features for the time it would require to do a port (even if that would be less than the two years we've put into it). Which is why something like Carbon was needed for Mac developers. You'll note that I'm not advocating porting existing apps to Yellow Box, but that developers of _new_ applications use it. Porting is generally not worth the time (though on MacOS X you'll be able to slowly start transitioning portions to Yellow, even though it won't really give you all of the benefits), but if you take full advantage of YB from the ground up, it _will_ save you a lot of development time for most apps. (There are, of course, exceptions; some APIs are better at some things than others, but YB seems to be easily the best _overall_ based on the other things I've used or studied.)
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 26 May 1998 17:38:55 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6keunf$c5r$8@ns3.vrx.net> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.c <1d9le4y.149xnlg656vwgN@carina25.wco.com> <slrn6mlpv9.4s8.rog@talisker.ohm.york.ac.uk> <Pine.NXT.3.96.980526171858.9803A-100000@pathos> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: bbum@codefab.com In <Pine.NXT.3.96.980526171858.9803A-100000@pathos> Bill Bumgarner claimed: > Actually, the same holds true for a screen... is it 555, 444, 565, 888, or > whatever? There are numerous different screen buffers... Another issue the NSColor use tends to avoid completely, and likely completely with the introduction of ColorSync everywhere. > ...the key is writing routines that can effeciently twiddle the bits at > the most common configurations. Yup. > doesn't mean you can't easily derive a second bitmap image rep that is > interleaved with three channels-- the API is all there and works quite > well [if I remember correctly]. Works well, but I always wanted a post convert API. IE, once you ask it for the BMP version of a graphic (for instance) you have to run a filter to get it into anything else. I'd love a... myImageRepAsATIFF = [NSImage imageRepInTIFFFormat:myImage]; or something like that. I'd save at least, oh, 20 lines of code in my app! Maury
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 26 May 1998 17:35:03 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6keug7$c5r$7@ns3.vrx.net> References: <slrn6mlpv9.4s8.rog@talisker.ohm.york.ac.uk> <1998052620073200.QAA13289@ladder01.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: larrysb@aol.com In <1998052620073200.QAA13289@ladder01.news.aol.com> LarrySB claimed: > :sure, you can get access to the data, but there's no standard way of > :addressing the pixels of data that you've got from an NSBitmapImageRep, > :because they arrive in one of many possible formats - you can't know > :which until you get it, and the manual page doesn't enumerate all the > :possibilities. *blink* THAT is what you're whining about? That you _actually_have_to_write_code_?? Boo hoo! What the heck do you get paid for? > The point about "- (NSColour *)getPixelAt:(NSPoint)position" would make it > easier to program, but I strongly fear it would introduce undue overhead. Fine, don't use it. > are better ways to do this, but lets keep it fairly simple) Consider the > overhead of going through Objective C's runtime binding and dispatch mechanism > and you've got the recipe for an unduly slow image processing task. The > performance hit is simply unnacceptable. No no, you THINK it's "simply unnacceptable". But when the system runs faster and has less latency, and runs real world apps many times faster (like Mathematica for instance), then I wonder what you base your comment on. Are you simply guessing as I suspect? > For reasonable performance even on fast processors like the G3, I have to keep > my pixels in buffers in memory and directly access them with pointers. But > that's not a big deal, since all my image acquisition and file handling code > put images into a plain old Mac GWorld. Sigh. theBitMap = [[NSData alloc] initWithContentsOfFile:theFile]; newImageRep = [NSImageRep newImageRep = [NSImageRep imageRepClassForData:theData]; bitsPerPixel = [newImageRep bitsPerSample]; // in case you need this for something // now insert your code here, working via pointers into the NSData // which you can get with the horribly complex call... thePointer = [theBitMap bytes]; // When you're done, put it into the newImageRep, put that // in a NSImage, and display > The point is that it works. So does this. > I just wrap up things like GWorlds in wrapper > classes to make my life easier. *snores* > blit it to the screen, export to quicktime codecs and so on. I still don't > know exactly how to keep my image data as raw pixels and yet make it 100% > compatible with the rest of NS. Have you bothered to actually try reading the docs sometime? > I want to save them as tiff format data, with > multiple images per file, along with other data. Does NS's TIFF have a problem with multi-part TIFF's? > This is a pretty simple example of coloring outside the lines and how a > framework can actually make things more difficult. Yeah, great example! Now let me ask you to describe how to make your code work when the image is larger than the physical memory of the machine? Or want it to work on a server machine that does the filters? Share document space? I do know how to under YB, I do exactly NOTHING. > I have run into plenty of other examples in my line of work too. And I'm sure at least one of them needs more than three lines of code to figure out, right? > For instance, I often work in color > spaces outside of RGB, so I need to convert from one format to another, hammer > the raw data and then convert it back. Ahhh, do you mean like... [myColor getRed:theRed green:the Green blue:theBlue alpha:theAlpha]; mashedColor = [self hammerOnTheColors:theRed:theBlue:theGreen]; [mashedColor colorUsingColorSpaceName:NSDeviceCMYKColorSpace]; I can't believe you picked NSColor as an example _problem_. NSColor is one of the most god-like classes in the entire YB! > Here's another tough one: Direct hardware support. Surely you jest. Needing direct access to hardware is a perfect sign of a lousy OS that no one uses any more. > someone pips off, a *lot* of existing software touches hardware. Communication > software uses the serial port. Not one on the Mac that I am aware of has ever gone directly to the serial port, they all use the serial drivers. > Games go direct to screen very often Not any more. Like not in the last 10 years that I know of. All the ones I know of use GWorlds, they have to or they break. > get input from game controllers and use the sound hardware. Nope and nope. At a minimum they talk to ADB, at a max ISp, most typically via a driver written by the vendor. Sounds ALWAYS goes to sound manager or above. Can't think of a single counter example, although I'm sure it's possible. > Lots of graphics programs support scanners and digital cameras. No, lots of graphics programs talk to the vendor supplied hardware drivers. > Lots of data acquisition software > directly interfaces with GPIB systems. No, lots of data acquisition software talks to the GPIB driver. > It's not a rare thing. Even stuff like > anti-piracy dongles need hardware support. Via the ABD Manager. These are even dumber examples than picking NSColor! What, do you think you can't writers under OpenStep? I/OKit anyone?!? > figuring out what was going on. It's actually nice doing things like mixing an > LView, LListener and an LPeriodical together. Or just subclassing from NSResponder, not needingto mixin Listener because the language handles that for you because you can send any message to any object, and not needing a real periodical framework because you block waiting for input and in the other cases (cursor blinking and such) the same reason as the second. More bad examples. The middle one is perfect for instance, because when you look at the syntax you get to use in NS with it's NotificationCenters it's WAYYYY better looking code. I mean here's how you get a mouseDown... - mouseDown: Can't get much easier than that. > If anything, the code is cleaner and simpler to maintain. Har! Maury
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: Best development environment, a nitpick Date: 26 May 1998 17:40:35 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6keuqj$c5r$9@ns3.vrx.net> References: <6kfcm3$gq$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: quinlan@intergate.bc.ca In <6kfcm3$gq$1@nnrp1.dejanews.com> quinlan@intergate.bc.ca claimed: > Just a comment for those of you who are claiming the YB is the best > development environment on the market: you should really take a look at a > good smalltalk implementation and see if you still feel that way. I have to some degree and still do. Why? a) Unix tools b) it's all GNU c) it's not a captive enviornment to any degree > 1) All types are objects. > 2) No primitive source/include file model > 3) Code modifications can be made and take effect while the code is running Absolutely, all great examples. The last one is apparently possible under OpenStep (on Mach anyway) but the person doing it left NeXT and it died (or so the story goes). Maury
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 26 May 1998 20:25:00 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6kfmgs$45u$1@crib.bevc.blacksburg.va.us> References: <6kfavv$3mg$1@crib.bevc.blacksburg.va.us> <1998052623235100.TAA06628@ladder01.news.aol.com> In article <1998052623235100.TAA06628@ladder01.news.aol.com>, larrysb@aol.com (LarrySB) wrote: > nurban@crib.bevc.blacksburg.va.us (Nathan Urban) wrote: > :> The point about "- (NSColour *)getPixelAt:(NSPoint)position" would make it > :> easier to program, but I strongly fear it would introduce undue overhead. > :It would be interesting to see Mike Paquette's comments on efficiency of > :doing direct drawing using the Yellow APIs. > It isn't a question of drawing API's, it's a question of accessing the pixel > data. Which obviously has something to do with the Yellow APIs, doesn't it? Read Maury's post for a discussion of how this is done. > :Yes, but you're missing the point that I originally challenged you on: > :the framework doesn't _prevent_ you from coloring outside the lines. If > :the framework doesn't do what you need, then don't use it; not using it > :doesn't make your job any _harder_ than if the framework wasn't there at > :all! > Where did I say impossible? You didn't say impossible. You said that frameworks hinder you if you try to "color outside the lines", which is false. Frameworks don't do everything. But if they don't do what you need, then they don't make it any _harder_ to do than if you weren't using a framework at all. > I have my doubts about supporting protocols, which is Obj-C's work-around for > mix-ins. Don't you have to implement the protocol in each class that supports > it? You write a "mixin" as a class which the things that mix it in compose. They conform to the protocol (in a syntax similar to mixing in a new class), and have then implement stub methods which call the mixin delegate's methods. There are various alternatives, such as forwarding messages in a protocol to the delegate, which are more convenient. Incidentally, protocols were not designed as a "work-around for mix-ins", they were designed to make distributed objects efficient.
From: louis@orb.isdn.emory.edu (Louis Leon) Newsgroups: comp.sys.next.programmer Subject: Re: Lotus Notes and Enterprise Objects Date: 27 May 1998 00:54:09 GMT Organization: Emory University Message-ID: <6kfo7h$1lq$1@lendl.cc.emory.edu> References: <wmoss-2605980954060001@bstephens.harland.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: wmoss@harland.net In <wmoss-2605980954060001@bstephens.harland.net> William Moss wrote: > Greetings all, > [stuff skipped] > > The directive has come down to interface our data with the corporate Lotus > Notes databases. With our current DBMS, this is not possible. So either we > will need to switch database managers or we should give our needs over to > the I.T. department. I'd rather not do this. I am not a Notes/Domino app developer, but for the last couple of weeks I have been looking at a similiar question. In my case, the requirement calls for exporting from the Notes database, via their APIs and then importing into my document management system (DMS) via its APIs. Your case sounds like it needs to go in the other direction.(?) > [stuff skipped] > > Would the Enterprise Object Framework facilitate the communication of > yellow box apps with a Lotus Notes database? Has this been done before or > is this "experimental ground"? It sounds like the powers that be are not requiring that the Notes database be the repository of record, so it doesn't seem that you need to write a yellowbox app that acts as a frontend to it. If you did write it as a frontend and intended to use EOF, you would need an "adaptor"; and I never hear of one commercially available. > > Are there any Yellowbox developed databases that are "EOF-enabled"? It sounds like you could get away with updating the Notes database periodically, in which case, you will need to export from your "EOF-enabled Yellowbox database" and then import into the Notes database. I am aware of EOF adapters for Sybase, Oracle, and OpenBase. > > Thanks > > William > -- Louis Leon
From: "Baskaran Subramaniam" <baskar@RemoveMe.snet.net> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 26 May 98 22:49:16 -0400 Organization: "SNET dial access service" Message-ID: <B190F772-50B09@192.168.128.1> References: <MPG.fd0d1146995e53798973f@news.supernews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit nntp://news.snet.net/comp.sys.mac.programmer.codewarrior, nntp://news.snet.net/comp.sys.next.programmer On Sat, May 23, 1998 2:27 PM, Donald Brown <mailto:don.brown@cesoft.com> wrote: >> Also, YB is THE API to use if you want to your app to run >> on multiple platforms. >Not if you want your app to run under MacOS 7.x or 8.x >Donald You are quite right on this. What will make life easy for me is Carbon support in YB for Windows. Baskaran PS: Please delete the string "RemoveMe." from my email address before replying to this message
From: wori0011@rz03.FH-Karlsruhe.DE (Richard Woeber) Newsgroups: comp.sys.next.programmer Subject: Re: Best development environment, a nitpick Date: 27 May 1998 06:30:36 GMT Organization: Fachhochschule of Karlsruhe, Germany Message-ID: <6kgbuc$8s6$1@nz12.rz.uni-karlsruhe.de> References: <6kfcm3$gq$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit quinlan@intergate.bc.ca wrote: : Just a comment for those of you who are claiming the YB is the best : development environment on the market: you should really take a look at a : good smalltalk implementation and see if you still feel that way. I'm not : saying the YB is bad, just that working in a smalltalk environment is better. : Why? My top three reasons are: : 1) All types are objects. : 2) No primitive source/include file model : 3) Code modifications can be made and take effect while the code is running ever written a device-driver in smalltalk? -- ============================================================================== Richard Woeber | login: Bill G Student of Cartography @ FH Karlsruhe | password: 97816 Lohr am Main /Germany | $msword | ed is the standard UNIX Text editor rich@qtvr.com NeXTMail ok | Your disk quota has been set to 50k http://www.qtvr.com/rich | $_ ==============================================================================
From: *tqayumi@familyware.com Newsgroups: comp.sys.next.programmer Subject: looking for openstep developer in vancouver Date: Tue, 26 May 1998 19:03:13 -0700 Organization: The University of British Columbia Sender: qayumi@206.186.113.167 Message-ID: <*tqayumi-2605981903130001@206.186.113.167> e-mail me at ****tarique@@cyberactivetech.com**** please remove star signs and extra @ sign from e-mail
From: "Baskaran Subramaniam" <baskar@RemoveMe.snet.net> Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 26 May 98 22:47:10 -0400 Organization: "SNET dial access service" Message-ID: <B190F6F4-4ED8D@192.168.128.1> References: <6k6nqp$avg$4@news.idiom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit On Sat, May 23, 1998 10:52 AM, John C. Randolph <mailto:jcr.remove@this.phrase.idiom.com> wrote: >If I were a Mac developer today, looking at the transition from the >MacToolbox API to Carbon, I'd take the opportunity to re-do the GUI using the >Appkit. It's fast, it's flexible, and it's a few hundred man-years of code I >don't have to write. > > This is no good if you want to target MacOS 8.x derivative or MacOS 7.5.x. If you stick with Carbon, you can target all versions of MacOS. Remember, MacOS X is not going to run on most of the installed base of the Mac hardware out there. What we really need is support for Carbon on the Windows version of the Yellow Box. That way the majority of the commercial vendors developing software for the Mac and Windoze will benefit. Baskaran
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 27 May 1998 00:24:17 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6kg4hh$4g3$1@crib.bevc.blacksburg.va.us> References: <6kfavv$3mg$1@crib.bevc.blacksburg.va.us> <1998052623235100.TAA06628@ladder01.news.aol.com> <6kfmgs$45u$1@crib.bevc.blacksburg.va.us> <6kg0td$6cp$1@news.digifix.com> In article <6kg0td$6cp$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: > From this discussion, I get that a protocol isn't what is > needed, but a Category on NSBitmapImageRep that implements a > getPixel:: in such a manner that it can deal with a all the variants > of the pixel stuff required. Yeah, there were actually _two_ separate topics being discussed: the direct image manipulation stuff, and whether Objective-C has a viable substitute for multiple inheritance.
From: quinlan@intergate.bc.ca Newsgroups: comp.sys.next.programmer Subject: Re: Best development environment, a nitpick Date: Wed, 27 May 1998 05:08:05 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6kg73l$1ot$1@nnrp1.dejanews.com> References: <6kfcm3$gq$1@nnrp1.dejanews.com> <6keuqj$c5r$9@ns3.vrx.net> In article <6keuqj$c5r$9@ns3.vrx.net>, maury@remove_this.istar.ca (Maury Markowitz) wrote: > Absolutely, all great examples. The last one is apparently possible > under OpenStep (on Mach anyway) but the person doing it left NeXT and it > died (or so the story goes). I think that the technology was called Fix-and-Go. Talk to Andrew if you want to know more about it. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: Ranjit Deshpande <ranjitd@pacbell.net> Newsgroups: comp.sys.next.hardware,comp.sys.next.programmer Subject: Re: NeXTstation TurboColor Documentation Date: Tue, 26 May 1998 23:30:31 +0000 Organization: Pacific Bell Internet Services Message-ID: <356B5095.43FFA722@pacbell.net> References: <35689ACA.E4CB4E1B@pacbell.net> <6kdir4$fl0$1@msunews.cl.msu.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; x-mac-type="54455854"; x-mac-creator="4D4F5353" Content-Transfer-Encoding: 7bit Randy, Thanks for your response. I do have information on the NCR chip and on several bits and pieces of silicon in the NeXTstation, but have no idea on how all of this fits together. Strange that you should mention the Black Linux group, they don't seem to have much either. I guess NeXT never released anything on their hardware. Typical of Steve Jobs I guess. Unfortunately the pointers you gave don't help much. I've been through those already. I really think that the 68k line of processors are awesome and I'd love to tinker with my new (well... used) NeXTstation. Maybe I'll just have to poke around. Anybody care to tell me if the NeXT programming manuals provide any kind of low level info on the NeXTstation ? Thanks, Ranjit
From: kay@buddhist.com (Kay Schulz (The Bumpui)) Newsgroups: comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.lang..java.gui Subject: GNUStep and Java Date: Wed, 27 May 1998 01:26:00 GMT Organization: National University of Singapore Message-ID: <356b6ace.511302914@news.nus.edu.sg> Hi is there anyone out there thinking about programming the OpenStep API in Java? Is there anyone who build the look and feel og OpenStep using Java? I mean the menues, the filebrowser? I would love to have the possibility to have a Next-look and feel API for Java Kay Schulz
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 27 May 1998 03:22:21 GMT Organization: Digital Fix Development Message-ID: <6kg0td$6cp$1@news.digifix.com> References: <6kfavv$3mg$1@crib.bevc.blacksburg.va.us> <1998052623235100.TAA06628@ladder01.news.aol.com> <6kfmgs$45u$1@crib.bevc.blacksburg.va.us> In-Reply-To: <6kfmgs$45u$1@crib.bevc.blacksburg.va.us> On 05/26/98, Nathan Urban wrote: <snip> >You write a "mixin" as a class which the things that mix it in compose. >They conform to the protocol (in a syntax similar to mixing in a new >class), and have then implement stub methods which call the mixin >delegate's methods. There are various alternatives, such as forwarding >messages in a protocol to the delegate, which are more convenient. > >Incidentally, protocols were not designed as a "work-around for >mix-ins", they were designed to make distributed objects efficient. > From this discussion, I get that a protocol isn't what is needed, but a Category on NSBitmapImageRep that implements a getPixel:: in such a manner that it can deal with a all the variants of the pixel stuff required. As I wrote to Mike Paquette, I'd like to see an NSBitMapImagePixelEnumerator so that you could easily iterate over every pixel in the data set. Mind you, I know that its not that big a deal to implement if you need to on your own, especially if you cheat and promote all your graphics for that need to 24-bit in a known storage format using NSBitMapImage.. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: jesjones@halcyon.com (Jesse Jones) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Tue, 26 May 1998 21:43:17 -0700 Organization: Edmark Message-ID: <jesjones-ya02408000R2605982143170001@news.accessone.com> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <6kc2go$5tr$3@ns3.vrx.net>, maury@remove_this.istar.ca (Maury Markowitz) wrote: > In fact it's worth noting that the ONLY implementation available that uses > MI on the Mac is (at least last time I looked) PowerPlant. PP has a lot of > nice features but it's absolutely nothing whatsoever like YB's objects. MacApp has moved rather heavily into MI: they've dropped their base class TObject and added a bunch of mixin classes. Also Taligent's CommonPoint frameworks are built from the ground up around MI (Taligent published an excellent C++ style guide that includes a nice section on using MI in C++ without stumbling into the virtual base class nightmare). -- Jesse
From: jesjones@halcyon.com (Jesse Jones) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Tue, 26 May 1998 22:18:49 -0700 Organization: Edmark Message-ID: <jesjones-ya02408000R2605982218490001@news.accessone.com> References: <6kcc12$hb1$3@news.idiom.com> <1998052617163600.NAA23784@ladder03.news.aol.com> <6kej3r$366$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <6kej3r$366$1@crib.bevc.blacksburg.va.us>, nurban@vt.edu wrote: > In article <1998052617163600.NAA23784@ladder03.news.aol.com>, larrysb@aol.com (LarrySB) wrote: > > > :First point: Multiple Inheritance is a bug, not a feature. It introduces > > :ambiguity in method lookup, and can force subclasses to figure out which of > > :multiple parents a method has been inherited from. MI may make the life of a > > :lazy designer easier, but it complicates the task of anyone who has to > > :maintain your code. Learn to use containment and delegation. I've been using the constrained form of MI advocated in _Taligent's Guide to Well Mannered Programs_ for years now and this has very rarely been a problem. In fact the only place I can remember having a problem is with my templated broadcast/listener (ie observer) classes because an object may want to listen to or broadcast more than one message. But even here the compiler will let you know if there's an ambiguity and it's easy to work around. I've seen an awful lot of people disparage MI including some big name authors, but I've never seen a rationale that made sense. The only argument that I've seen against MI in C++ that carries weight consists of pointing to the ugly and complex semantics of virtual base classes. The answer, of course, is not to use virtual base classes: it's much cleaner to use light weight mixin classes so that you don't wind up with a diamond shaped inheritance tree. Note that both PowerPlant and (recently) MacApp use MI rather heavily. In both cases they seem considerably better then they would be without MI. > > That's the stupidest fucking thing I've ever heard. Properly used, MI greatly > > simplifies design and maintainence. Having used both Power Plant and MFC, I > > can tell you firsthand that MI is much cleaner and easier to deal with. > > Like he said: learn to use containment and delegation. You get pretty > much the same thing, without all the conflict resolution hassles. Unfortunately I don't know Objective-C well enough to judge how well it compensates for lacking MI. Let me give you an example from my own framework Raven: the document class TDocument. This is the model in the MVC pattern. It descends from: 1) MCommander which enables it to respond to menu commands (chain of reponsibility pattern). 2) MStreamable so that the document can be streamed in and out. 3) MReferenceCounted so that the document can be deleted automatically when the last view goes away (typically when the user closes a window). 4) MBroadcaster<SDocumentMessage> so that the document can broadcast to interested objects as the model state changes (MVC and Observer patterns). MBroadcaster and MListener are templatized since different objects can broadcast lots of different things (they descend from non-template base classes to minimize code bloat). This is a very nice design: because virtual base classes are avoided the inheritance semantics are very simple and instead of a single tree with rather fat objects near the top you wind up with very small easily composable objects and a forest of trees. I've used MacApp back when it lacked MI and IMO decent frameworks with MI are easier to understand and to use. How would something like my TDocument class be written in Objective-C? -- Jesse
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 27 May 1998 02:06:36 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6kgahc$4lv$1@crib.bevc.blacksburg.va.us> References: <6kcc12$hb1$3@news.idiom.com> <1998052617163600.NAA23784@ladder03.news.aol.com> <6kej3r$366$1@crib.bevc.blacksburg.va.us> <jesjones-ya02408000R2605982218490001@news.accessone.com> In article <jesjones-ya02408000R2605982218490001@news.accessone.com>, jesjones@halcyon.com (Jesse Jones) wrote: > Let me give you an example from my own > framework Raven: the document class TDocument. This is the model in the MVC > pattern. It descends from: > > 1) MCommander which enables it to respond to menu commands (chain of > reponsibility pattern). > 2) MStreamable so that the document can be streamed in and out. > 3) MReferenceCounted so that the document can be deleted automatically when > the last view goes away (typically when the user closes a window). > 4) MBroadcaster<SDocumentMessage> so that the document can broadcast to > interested objects as the model state changes (MVC and Observer patterns). > How would something like my TDocument class be written in Objective-C? I mentioned this elsewhere in the thread, but.. you would generally make each of those mixin classes into regular Objective-C classes, and then _compose_ them into the derived object rather than _inheriting_ them. Composition is more flexible anyway. You also typically create protocols (like Java interfaces, or akin to C++ pure virtual abstract interface mixins) for the derived class to conform to (i.e., telling the compiler and runtime system that the object implements that interface, without it actually inheriting any implementations). So the TDocument class would conform to those four protocols, meaning that all of the methods are known to exist within the class, but it doesn't inherit any implementation from the protocols; that is done separately by composition. Designs in Objective-C do tend towards the "fatter" higher-level objects that you don't like, however. For example, the OpenStep analogue to your TDocument example would be: (1) The class would inherit from NSResponder (like your MCommander), which itself inherits from NSObject -- this is its "primary" behavior. (2) The archiving stuff is defined in NSObject and overridden by subclasses. (3) Reference counting is used by all objects for the sake of consistency, so that too is implemented in NSObject. (4) Is unnecessary using the NSNotification scheme in Foundation Kit; an object needs to conform to no particular interface in order to be an observer. No composition is used in this case. Composition _is_ used extensively elsewhere in OpenStep, but not as often in the lightweight MI sense you are using; the collaborators are, as I mentioned, "fatter". Categories are also sometimes used where you would use MI. To be fair, I have advocated in the past the addition of a certain kind of MI-like behavior in Objective-C which was discussed in the past for a while (either here or on comp.lang.objective-c) -- it has actually been implemented in GNU Objective-C as "behaviors" as well as in the TOM language -- which consists of being able to associate an implementation with a protocol which can be added to any existing protocol (and hence to anything that implements that protocol) as a sort of "protocol category", as long as its implementation does not access any instance variables or methods outside of the protocols it inherits. Various people have argued back and forth about whether this is good in general; it does lead to some of the same problems (method collisions) that C++ MI has.
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 27 May 1998 11:03:13 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6mnsnk.55h.rog@talisker.ohm.york.ac.uk> References: <slrn6mlpv9.4s8.rog@talisker.ohm.york.ac.uk> <1998052620073200.QAA13289@ladder01.n <6keug7$c5r$7@ns3.vrx.net> On 26 May 1998 17:35:03 GMT, Maury Markowitz <maury@remove_this.istar.ca> wrote: > > For reasonable performance even on fast processors like the G3, I have to keep > > my pixels in buffers in memory and directly access them with pointers. But > > that's not a big deal, since all my image acquisition and file handling code > > put images into a plain old Mac GWorld. > > Sigh. > > theBitMap = [[NSData alloc] initWithContentsOfFile:theFile]; > newImageRep = [NSImageRep newImageRep = [NSImageRep > imageRepClassForData:theData]; > bitsPerPixel = [newImageRep bitsPerSample]; // in case you need this for > something > > // now insert your code here, working via pointers into the NSData > // which you can get with the horribly complex call... > thePointer = [theBitMap bytes]; > > // When you're done, put it into the newImageRep, put that > // in a NSImage, and display > > > The point is that it works. > > So does this. well, you've done the easiest bit. (and even that's not even right, because under openstep a) the method is called "bitmapData", not "bytes" b) that method will only get *some* of the data if the data is held in planar form. "working via pointers" into the NSData is not that easy, as i explained before. a colleague of mine has in fact written some code to unpack an NSBitmapImageRep. this code is just under 300 lines of fairly involved bit twiddling, and it *still* doesn't cater for all possible cases. from looking at the code, it has to work with all combinations of the following: o number of bits per sample o number of samples per pixel o number of bits per pixel o number of bytes per row o planar and non-planar data o has alpha or not o at least 4 different colour spaces all of these can vary arbitrarily. i hope you'll agree that "working via pointers into the NSData" is not really a sufficient description... sometime i will write a category of NSBitmapImageRep, something like: - (int *)getDataInRect:(NSRect)r; which would return a block of data for a rectangular block of pixels in a completely standard form. (24+8 bit rgb+alpha, perhaps) this would get around the efficiency problems of invoking a method for every pixel. in the meantime, it is *not* an easy problem and you're being ingenuous if you maintain that it is... > > blit it to the screen, export to quicktime codecs and so on. I still don't > > know exactly how to keep my image data as raw pixels and yet make it 100% > > compatible with the rest of NS. > > Have you bothered to actually try reading the docs sometime? you have to read a lot of fairly poorly worded documentation to get a good idea of what's entailed in blitting raw pixels to screen efficiently under openstep. i don't think his confusion is entirely his fault. cheers, rog.
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: Best development environment, a nitpick Date: 27 May 1998 02:18:58 GMT Organization: Idiom Communications Message-ID: <6kft6i$23d$1@news.idiom.com> References: <6kfcm3$gq$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: quinlan@intergate.bc.ca quinlan@intergate.bc.ca may or may not have said: -> Just a comment for those of you who are claiming the YB is the best -> development environment on the market: you should really take a look at a -> good smalltalk implementation and see if you still feel that way. I'm not -> saying the YB is bad, just that working in a smalltalk environment is better. -> -> Why? My top three reasons are: -> -> 1) All types are objects. -> 2) No primitive source/include file model -> 3) Code modifications can be made and take effect while the code is running Smalltalk is certainly "purer" than Objective-C, and its library of classes is richer and has had a few decades of debugging. If I couldn't use OpenStep, I'd probably be a Smalltalker. The advantage of Obj-C over Smalltalk though, is performance. What I would consider the holy grail though, would be an environment where I could write any method in any language, with no compile-edit-debug cycle, where every object could be queried for its source code, and so forth. Also, I'm a bit more inclined towards prototypes over classes. -jcr
From: cb@bartlet.df.lth.se (Christian Brunschen) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 27 May 1998 14:45:15 +0200 Organization: The Computer Society at Lund Message-ID: <6kh1sr$uej$1@bartlet.df.lth.se> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.c <1d9le4y.149xnlg656vwgN@carina25.wco.com> <slrn6mlpv9.4s8.rog@talisker.ohm.york.ac.uk> NNTP-Posting-User: cb In article <slrn6mlpv9.4s8.rog@talisker.ohm.york.ac.uk>, Roger Peppe <rog@ohm.york.ac.uk> wrote: >On Mon, 25 May 1998 16:39:52 -0700, Mike Paquette <mpaque@wco.com> wrote: >> LarrySB <larrysb@aol.com> wrote: >> > Ok, how about something as simple as manipulating raw image pixels in an >> > offscreen buffer? (say a median filter) > >> All you need to do is build an NSBitmapImageRep. This cam be >> initialized with data from a file, or from a view (region) of a window, >> just by locking focus on the view and loading the bitmap rep. You can >> bang on the bits all you want, and them put them on the display or save >> to a file with one line of code. > >that's not as easy to do as you're making out. > >sure, you can get access to the data, but there's no standard way of >addressing the pixels of data that you've got from an NSBitmapImageRep, >because they arrive in one of many possible formats - you can't know >which until you get it, and the manual page doesn't enumerate all the >possibilities. > >for instance, if you're getting pixels in this way, your code has >to handle: > planar and non-planar data > any number of bits per pixel (where does it say that i might not be > given 13 bit per pixel data?) > any number of bytes per row Ummm ... not necessarily. Say you want to handle only 24-bits-per-pixel RGB, no alpha, pixels in meshed configuration (ie, not planar) - (NSBitmapImageRep *) doSomethingWithPixelsInImageRep:(NSBitmapImageRep *)pixelsInAnyFormat { int pixelsWide = [pixelsInAnyFormat pixelsWide]; int pixelsHigh = [pixelsInAnyFormat pixelsHigh]; int i, j; NSBitmapImageRep *pixelsInMyPreferredFormat = [[NSBitmapImageRep alloc] initWithBitmapDataPlanes:NULL pixelsWide:pixelsWide pixelsHigh:pixelsHigh bitsPerSample:8 samplesPerPixel:3 hasAlpha:NO isPlanar:NO colorSpaceName:NSCalibratedRGBColorSpace bytesPerRow:0 bitsPerPixel:0]; NSImage *tmpImage = [[NSImage alloc] initWithSize:NSMakeSize(pixelsWide, pixelsHigh)]; unsigned char *bitmapData; [tmpImage addRepresentation:pixelsinMyPreferredFormat]; [tmpImage lockFocusOnRepresentation:pixelsInMyPreferredFormat]; [pixelsInAnyFormat draw]; [tmpImage unlockFocus]; [tmpImage release]; // At this point, 'pixelsInMyPreferredFormat' represents the same image // as 'orig', but in your preferred pixel format // fetch a pointer to the pixel data bitmapData = [pixelsInMyPreferredFormat bitmapData]; // Loop over the pixels for (i = 0; i < pixelsHigh; i++) { for (j = 0; j < pixelsWide; j++) { unsigned char red = *bitmapData++; unsigned char green = *bitmapData++; unsigned char blue = *bitmapData++; // do something with the pixel here ... } } return [pixelsInMyPreferredFormat autorelease]; } You could even write a method in a category for NSBitmapImageRep, like this: - (id)initWithImageRep:(NSImageRep *)imageInAnyFormat pixelsWide:(int)width pixelsHigh:(int)height bitsPerSample:(int)bps samplesPerPixel:(int)spp hasAlpha:(BOOL)alpha isPlanar:(BOOL)isPlanar colorSpaceName:(NSString *)colorSpaceName bytesPerRow:(int)rowBytes bitsPerPixel:(int)pixelBits { [self initWithBitmapDataPlanes:NULL pixelsWide:width pixelsHigh:height bitsPerSample:bps samplesPerPixel:spp hasAlpha:alpha isPlanar:isPlanar colorSpaceName:colorSpaceName bytesPerRow:rowBytes bitsPerPixel:pixelBits]; NSImage *tmpImage = [[NSImage alloc] initWithSize:NSMakeSize(width, height)]; [tmpImage addRepresentation:self]; [tmpImage lockFocusOnRepresentation:self]; [pixelsInAnyFormat drawInRect:NSMakeRect(0, 0, width, height)]; [tmpImage unlockFocus]; [tmpImage release]; return self; } and you can use it to create a new bitmapImageRep in any format you like ... from any sort if NSImageRep, including any vector-based NSImageRep subclasses you might get from somewhere ... > >that's quite a lot of bit twiddling and nasty special cases >just to get at screen bitmap data in a portable form. Well, there you go ... no bit-twiddling necesary ... try cut-and-paste:ing the above code snippets, they _should_ work. > >if they just provided a "- (NSColour *)getPixelAt:(NSPoint)position" >method, it would make a lot of things much easier, or a standard >way of indexing into data as returned from an NSBitmapImageRep, >it would make things a lot easier! Yes it would. But things as they are now are not _that_ much worse; and of course, if you really do want to iterate over all pixels in an image, a startup cost for translating all your pixels to a standard format such as the above code snippet may well be preferrable to having overhead associated with fetching each pixel. Ie, you may gain more by manipulating the raw pixels you access by walking through the 'bitmapData' you get from the NSBitmapImageRep compared to accessing the pixels through the imaginary "- (NSColor *)getPixelAt:(NSPoint)position" method for each pixel, than you lose by having to call the conversion process above compared to not having to do any pre-initialization. OK, I just reread the paragraph I just wrote, and _I_ understand it ... barely :) If it's unclear to you what I meant: t_total = t_initialization + w * h * (t_pixel_access + t_pixel_work) If we use '- (NSColor *)getPixelAt:(NSPoint)position', we have t_total = 0 + w * h * (t_'getPixelAt:' + t_pixel_work) If we use the stuff I outlined above, we have t_total = t_'convertPixels' + w * h * ( 0 + t_pixel_work) (Well, vaguely speaking anyway). And which one of these is better is not trivial - it depends on each of the factors involved, and will likely vary from application to application. That said, I would very much like to have a method like 'getPixelAt:', except I'd probably call it - (NSColor *)colorAtPoint:(NSPoint)point perhaps with a matching - (NSColor *)colorAtPixelX:(int x) y:(int)y Why ? Well, 'colorAtPoint' could use fractional x and y values, and actually do interpolation or some form of anti-aliasing .... but I'm getting far off the subject :) > >i think LarrySB has a point (if only a small one, because i'd hazard >that *most* applications have no need to do this) Well, as I have tried to show, banging at pixels > > cheers, > rog. > Best regards // Christian Brunschen
From: "Jeremy Bettis" <jeremy@hksys.com> Newsgroups: comp.sys.next.programmer Subject: Re: access to [NSString cString] functionality? Date: Wed, 27 May 1998 10:55:14 -0500 Organization: Internet Nebraska Message-ID: <6khd25$p6n$1@owl.inetnebr.com> References: <slrn6mlrr8.4s8.rog@talisker.ohm.york.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit -----BEGIN PGP SIGNED MESSAGE----- >the -(const char *)cString method of NSString manages >to put a bare C string into an autorelease pool. They use an undocumented class, _NSAutoreleasedMallocMemory (From gdb) - -[_NSAutoreleasedMallocMemory initWithCapacity:] - -[_NSAutoreleasedMallocMemory remainingCapacity] - -[_NSAutoreleasedMallocMemory mallocMemoryWithCapacity:] - -[_NSAutoreleasedMallocMemory dealloc] +[_NSAutoreleasedMallocMemory mallocMemoryWithCapacity:] It is a simple concept, make a class that takes a void pointer in init: then frees it on dealloc. But in your case, you are creating an autoreleased NSData object, why not just use -[NSData bytes] to get an autoreleased pointer to the bytes? -----BEGIN PGP SIGNATURE----- Version: PGP 5.5.5 iQBVAwUBNWw3YRmtiimFYrNNAQFECgIAnZ6vpFpZhSyndf9G5BxYXB3kl+09oAs3 UpWDB2AdZvJFG9h3Dp3E5lelIdgXlMWz4qdjJqIgmqfHcfNefq/vdQ== =GNTO -----END PGP SIGNATURE-----
From: embuck@palmer.cca.rockwell.com (Erik M. Buck) Newsgroups: comp.sys.next.programmer Subject: Re: Best development environment, a nitpick Date: 27 May 1998 15:28:51 GMT Organization: Rockwell International Message-ID: <6khbfj$loc2@onews.collins.rockwell.com> References: <6kfcm3$gq$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: quinlan@intergate.bc.ca In <6kfcm3$gq$1@nnrp1.dejanews.com> quinlan@intergate.bc.ca wrote: > Just a comment for those of you who are claiming the YB is the best > development environment on the market: you should really take a look at a > good smalltalk implementation and see if you still feel that way. I'm not > saying the YB is bad, just that working in a smalltalk environment is better. > I am a big fan of Smalltalk also. I think you will find that many Objective-C users are. > Why? My top three reasons are: > > 1) All types are objects. > 2) No primitive source/include file model > 3) Code modifications can be made and take effect while the code is running > 1) This is possible in Objective-C if you are disciplined and are willing to give up some performance and have appropriate frameworks available. With the introduction of NSString etc. in NeXT Objective-C, the use of non-scalar non-object types is rare. 2) Actually, many people think the "artificial" separation of interface from implementation is a GOOD thing. The "primitive" include model is a file system issue. It is really not all that primitive and class browsers etc. are available. 3) This would be nice to have There are however some very large problems that have traditionally plagued the "superior" smalltalk environments. The biggest issue has always been supporting more than one developer per project. If you have 20 programmers on a team, they will be stepping all over each other in smalltalk. The "shoot from the hip" tradition in smalltalk can be a disaster in some cases. How can anyone figure out where and when the new bug was introduced ? Distribution of smalltalk apps has been a problem too. Is there a runtime available ? Do I have to distribute source ? The program just produced an error, how do I continue ?
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 27 May 1998 11:40:38 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6kgu3m$nc0$10@ns3.vrx.net> References: <6k6nqp$avg$4@news.idiom.com> <B190F6F4-4ED8D@192.168.128.1> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: baskar@RemoveMe.snet.net In <B190F6F4-4ED8D@192.168.128.1> "Baskaran Subramaniam" claimed: > This is no good if you want to target MacOS 8.x derivative or MacOS 7.5.x. > If you stick with Carbon ... this is no good if you wish to target Win32. Hmmm, Win32 vs. MacOS... > What we really need is support for Carbon on the Windows version of the > Yellow Box. *coff* Ummm, you don't do Carbon "in" or "on" anything - it talks to the kernel directly. What you suggest is likely impossible (or close enough) MacOS -> YB -> Win32, and seems to stem from a potential misunderstanding of how the current solution is to be implemented. Maury
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 27 May 1998 14:12:59 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6mo7rf.55h.rog@talisker.ohm.york.ac.uk> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.c <6kh1sr$uej$1@bartlet.df.lth.se> On 27 May 1998 14:45:15 +0200, Christian Brunschen <cb@bartlet.df.lth.se> wrote: > In article <slrn6mlpv9.4s8.rog@talisker.ohm.york.ac.uk>, > Roger Peppe <rog@ohm.york.ac.uk> wrote: > >for instance, if you're getting pixels in this way, your code has > >to handle: > > planar and non-planar data > > any number of bits per pixel (where does it say that i might not be > > given 13 bit per pixel data?) > > any number of bytes per row > > Ummm ... not necessarily. > > Say you want to handle only 24-bits-per-pixel RGB, no alpha, pixels in > meshed configuration (ie, not planar) > > - (NSBitmapImageRep *) > doSomethingWithPixelsInImageRep:(NSBitmapImageRep *)pixelsInAnyFormat [snipped code to convert NSBitmapImageRep to standard bit depth] yes, you're right (and others have suggested it too) - that is indeed a reasonable solution. the only thing i don't like about it is that it at least doubles the space needed for the bitmap. if you're dealing with large bitmaps, (i've had to deal with *huge* (200MB+) images before) then this is not a trivial overhead. as NSImageReps don't allow rendering of a particular portion of themselves, this would seem to be an unavoidable problem with this way of doing things. c'est la vie. cheers, rog.
From: uzs62s@uni-bonn.de (Axel "Mikesch" Katerbau) Newsgroups: comp.sys.next.programmer Subject: Re: GNUStep and Java Date: Wed, 27 May 1998 17:58:37 +0200 Organization: Uni Bonn Message-ID: <1d9p1q5.66tnon15t0o0eN@rhrz-isdn3-p3.rhrz.uni-bonn.de> References: <356b6ace.511302914@news.nus.edu.sg> Kay Schulz (The Bumpui) <kay@buddhist.com> wrote: > Hi > is there anyone out there thinking about programming the OpenStep API > in Java? > Is there anyone who build the look and feel og OpenStep using Java? > I mean the menues, the filebrowser? > I would love to have the possibility to have a Next-look and feel API > for Java > > Kay Schulz I don't really know if you want to have a OpenStep Java-Swing look&feel or if you want to program the OpenStep Framework. The latter can be done now with the current Rhapsody DR2. Axel -- Microsoft - Blue screens for a blue planet.
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 27 May 1998 11:38:18 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6kgtva$nc0$9@ns3.vrx.net> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: jesjones@halcyon.com In <jesjones-ya02408000R2605982143170001@news.accessone.com> Jesse Jones claimed: > MacApp has moved rather heavily into MI: Fair enough, haven't looked at it since 3.0. > TObject and added a bunch of mixin classes. Also Taligent's CommonPoint > frameworks are built from the ground up around MI (Taligent published an > excellent C++ style guide that includes a nice section on using MI in C++ > without stumbling into the virtual base class nightmare). However I don't believe it ever released on the Mac, no? Maury
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 27 May 1998 16:41:26 GMT Organization: Technical University of Berlin, Germany Message-ID: <6khfnm$af3$1@news.cs.tu-berlin.de> References: <6j7me9$so3$1@newsfep4.sprintmail.com> <MPG.fc14db42395aa139896a2@news.supernews.c <6kh1sr$uej$1@bartlet.df.lth.se> <slrn6mo7rf.55h.rog@talisker.ohm.york.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit rog@ohm.york.ac.uk (Roger Peppe) writes: >On 27 May 1998 14:45:15 +0200, Christian Brunschen <cb@bartlet.df.lth.se> wrote: >> In article <slrn6mlpv9.4s8.rog@talisker.ohm.york.ac.uk>, >> Roger Peppe <rog@ohm.york.ac.uk> wrote: >> >for instance, if you're getting pixels in this way, your code has >> >to handle: >> > planar and non-planar data >> > any number of bits per pixel (where does it say that i might not be >> > given 13 bit per pixel data?) >> > any number of bytes per row >> >> Ummm ... not necessarily. >> >> Say you want to handle only 24-bits-per-pixel RGB, no alpha, pixels in >> meshed configuration (ie, not planar) >> >> - (NSBitmapImageRep *) >> doSomethingWithPixelsInImageRep:(NSBitmapImageRep *)pixelsInAnyFormat >[snipped code to convert NSBitmapImageRep to standard bit depth] >yes, you're right (and others have suggested it too) - that is indeed a >reasonable solution. >the only thing i don't like about it is that it at least doubles the >space needed for the bitmap. if you're dealing with large bitmaps, >(i've had to deal with *huge* (200MB+) images before) then this is not >a trivial overhead. My solution to this has been a scan-line oriented filter system. Works great for most image-processing tasks except for arbitrary geometry transformations, though I've been sketching a simple tiling scheme with dependencies to do these efficently as well. Too bad I don't need geometry transforms for my work. :-) One big advantage is that multiple filters will operate sequentially on a single scan-line (fitting in L2, sometimes even L1) instead of each one iterating over available memory (+ disk). Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 27 May 1998 11:52:37 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6kguq5$nc0$11@ns3.vrx.net> References: <6kfavv$3mg$1@crib.bevc.blacksburg.va.us> <1998052623235100.TAA06628@ladder01.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: larrysb@aol.com In <1998052623235100.TAA06628@ladder01.news.aol.com> LarrySB claimed: > It isn't a question of drawing API's, it's a question of accessing the pixel > data. Data... NSData... data... NSData. Hmmmmm, could it be that NSData is used for working with big chunks of data? No, couldn't be that. > Where did I say impossible? The problem with YB is the API isn't an API, it's > a framework. Uhhh, ok. Sister! daughter! sister! daughter! > The authors of the framework decided that no one needs an easy > way to get to raw pixel data and didn't put any way in to do it. Whereas the MacOS doesn't either. The fact that you can use a GWorld to do it is basically the same as using a NSData to do it, with the exception that the pointer math is done for you. > The problem is it *does* make life harder. How? > There's not an easy way to leverage the YB for > this task, which means you have to do it all yourself. Advantage nil. Disadvantage nil. A very big difference from your last statement, which was that there is a big disadvantage. > I have my doubts about supporting protocols, which is Obj-C's work-around for > mix-ins. Don't you have to implement the protocol in each class that supports > it? *blink* Maury
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 27 May 1998 11:56:28 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6kgv1c$nc0$12@ns3.vrx.net> References: <6kfavv$3mg$1@crib.bevc.blacksburg.va.us> <1998052623235100.TAA06628@ladder01.news.aol.com> <6kfmgs$45u$1@crib.bevc.blacksburg.va.us> <6kg0td$6cp$1@news.digifix.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: sanguish@digifix.com In <6kg0td$6cp$1@news.digifix.com> Scott Anguish claimed: > From this discussion, I get that a protocol isn't what is > needed, but a Category on NSBitmapImageRep that implements a > getPixel:: in such a manner that it can deal with a all the variants > of the pixel stuff required. Seems that way. I think the solution could be memory intensive though, unless we can get at the imageRep via some sort of bytes call. Basically the code is easy, ask the imageRep for it's bitsPerSample, then use that as a pointer size to go into the data and read it. Five lines, maybe less. The issue as I see it is that currently imageRep doesn't have a "trivial" way to deal with it as a NSData (or similar) which is what you'd way. > As I wrote to Mike Paquette, I'd like to see an > NSBitMapImagePixelEnumerator so that you could easily iterate over > every pixel in the data set. Hmmm, neat idea. In general are enums fast though? I tend to use for's. Maury
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 27 May 1998 12:34:04 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6kh17s$nc0$13@ns3.vrx.net> References: <slrn6mlpv9.4s8.rog@talisker.ohm.york.ac.uk> <1998052620073200.QAA13289@ladder01.n <6keug7$c5r$7@ns3.vrx.net> <slrn6mnsnk.55h.rog@talisker.ohm.york.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: rog@ohm.york.ac.uk In <slrn6mnsnk.55h.rog@talisker.ohm.york.ac.uk> Roger Peppe claimed: > > theBitMap = [[NSData alloc] initWithContentsOfFile:theFile]; > > newImageRep = [NSImageRep newImageRep = [NSImageRep > > imageRepClassForData:theData]; > > bitsPerPixel = [newImageRep bitsPerSample]; // in case you need this for > > something > > > > // now insert your code here, working via pointers into the NSData > > // which you can get with the horribly complex call... > > thePointer = [theBitMap bytes]; > > > > // When you're done, put it into the newImageRep, put that > > // in a NSImage, and display > > > > > The point is that it works. > > > > So does this. > > well, you've done the easiest bit. (and even that's not even > right, because under openstep My thought was to... a) load the image into a NSData b) init a imageRep from that NSData c) use that to get the "pixel size" for the pointer math d) use that on the original NSData for the pixel data. However... > a) the method is called "bitmapData", not "bytes" Whoa, do you mean I have access to it as a byte stream already?!? Doesn't that just make it all that much easier? > b) that method will only get *some* of the data if the data > is held in planar form. Sorry, can you expand on that a bit? Oh, I see what you mean. Considering the original post was working with a GWorld, that portion can be assumed. GWorlds are simple pixmaps of depth x. > "working via pointers" into the NSData is not that easy, as i explained > before. a colleague of mine has in fact written some code to unpack an > NSBitmapImageRep. this code is just under 300 lines of fairly involved > bit twiddling, and it *still* doesn't cater for all possible cases. But would it work in this case? Against a "simple" bitmap? > o number of bits per sample > o number of samples per pixel > o number of bits per pixel > o number of bytes per row > o planar and non-planar data > o has alpha or not > o at least 4 different colour spaces Ok, now I understand why it's 300 lines. Sure, this is doing a LOT of work though. This seems to be one of those general issues though - it's harder under OS because OS can do so much more with it. If NSBitmapImageRep were basically equivalent to a GWorld, which the orignal post used, it would likely be some 20 lines. > i hope you'll agree that "working via pointers into the NSData" > is not really a sufficient description... It is for this *particular* case though. > sometime i will write a category of NSBitmapImageRep, something like: > > - (int *)getDataInRect:(NSRect)r; > > which would return a block of data for a rectangular block > of pixels in a completely standard form. (24+8 bit rgb+alpha, perhaps) Oh please please please plese pretty please!!! > you have to read a lot of fairly poorly worded documentation In general I find the docs in question to be the _best_ I've ever seen, with the exception of the "blow by blow" API listings. Maury
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.lang..java.gui Subject: Re: GNUStep and Java Date: 27 May 1998 18:24:17 GMT Organization: Idiom Communications Message-ID: <6khloh$3cm$2@news.idiom.com> References: <356b6ace.511302914@news.nus.edu.sg> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: kay@buddhist.com Kay Schulz (The Bumpui) may or may not have said: -> Hi -> is there anyone out there thinking about programming the OpenStep API -> in Java? Yes, Apple. -> Is there anyone who build the look and feel og OpenStep using Java? -> I mean the menues, the filebrowser? -> I would love to have the possibility to have a Next-look and feel API -> for Java I'm afraid you'd have to settle for the Mac Look and Feel. -jcr
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: Re: future of mac programming Newsgroups: comp.sys.next.programmer References: <slrn6mlpv9.4s8.rog@talisker.ohm.york.ac.uk> <6keug7$c5r$7@ns3.vrx.net> <slrn6mnsnk.55h.rog@talisker.ohm.york.ac.uk> Message-ID: <356c5f0b.0@news.depaul.edu> Date: 27 May 98 18:44:27 GMT Roger Peppe <rog@ohm.york.ac.uk> wrote: > all of these can vary arbitrarily. > i hope you'll agree that "working via pointers into the NSData" > is not really a sufficient description... > sometime i will write a category of NSBitmapImageRep, something like: > - (int *)getDataInRect:(NSRect)r; > which would return a block of data for a rectangular block > of pixels in a completely standard form. (24+8 bit rgb+alpha, perhaps) > this would get around the efficiency problems of invoking a method > for every pixel. Another approach would be to have an ImageDataCursor class, which works similarly to an NSEnumerator. You'd tell it to move to the next pixel, (or move n pixels). It would have methods for accessing the color data at the pixel. It might also have methods like performSelector:onRange: or performFunction:onRange:. The ImageDataCursor would then take that selector, grab a pointer to the method implementation (avoiding repeated lookups) and call it for each pixel in the given range. Or, it would just use a function pointer and do the same thing. That would be a lot faster than iterating over the image and asking for and setting NSColors. Concrete subclasses of ImageDataCursor would know how to handle different image formats. -- "... and subpoenas for all." - Ken Starr
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <1998052719541700.PAA15634@ladder01.news.aol.com> Date: 27 May 1998 19:54:16 GMT Organization: AOL http://www.aol.com References: <6khfnm$af3$1@news.cs.tu-berlin.de> Ok, I've let my temper get the best of me in my previous posts. I will endeavor to de-escalate now. Let me explain some basic image processing concepts for those who are unfamiliar with common techniques. The basic representation of data for photographic images is a 2D array of samples, each representing an image value at a particular point in the image. In the case of monochrome images, it is usually an integer value of a bitlength sufficient to represent the sample data. Often, it's a single byte. In my case, I have image aquisition hardware that is capable of 12-16 bits of significant information per pixel. Color image representations are usually a packed structure of values, often 3 bytes representing RGB. Another common form is aRGB which is four bytes, with the first either unused or used for some purpose as transpearancy. Another common type of RGB is two bytes where the components are 1,5,5,5 bits. There are other representations where the RGB information is held in seperate planes. Color images may also exist in spaces other than RGB. CMYK is typically the form for color printing. HSV, or hue saturation and value is another common form, HSL is similar. There are a number of color space models available, such as LAB, YUV, and so on. A good introduction is Apple's "Advanced Color Imaging On the Macintosh." The graphics and the first few chapters serve as a fast education on color spaces. Anyway, back to the meat of image processing: A common technique in image processing is something called a convolution kernel. The idea is quite simple: Take a pixel, and all of it's neighbors within radius R. If R equals 1, you have a 3x3 neghborhood of pixels. For each particular pixel within that neighborhood, multiply it by a particular constant according to it's position and sum all of the values. Replace the target pixel with the result. Consider: xxx xox xxx This is a basic 3x3 sampling kernel. In this example, (2,2) is the target pixel, (1,1) is the northwest neighbor, (3,3) is the southeast neighbor. -1 1 -1 1 0 1 -1 1 -1 This is a weighting kernel. By changing the values in the weighting kernel, a number of useful image processing functions can be applied. One can also change the size of the kernel for various effects. In one instance, I have had to apply a 13x13 kernel for peculiar task. There are other functions beyond simple summing which also provide useful IP functions. One example is the median filter. The median filter takes the list of sample points, sorts them according to value and replaces the target with the median value. The median filter is particular useful for removing noise in images. There are many other functions, such high-pass (sharpen), LaPlace, Gaussian blur, Sobel, minimum and maximum, mean-least variance and so on, which all use some variation of a pixel neighborhood. As you might imagine, when working with very large kernels, or particularly complex evaluations of the pixels, speed becomes a serious factor, even on very fast processor. Anyone who has to sit while photoshop applies some filter can attest to that fact. In actual code, one usually has a source buffer and an empty target buffer. You shouldn't be working on already altered pixels obvioulsy. Now, having written a substantial amount of this kind of code, I know where the bottlenecks crop up. To get acceptable speed, you have to be careful and avoid things like integer->float conversions, keep as much of the image as possible in memory, do direct pointer access to the pixels. For more sophisticated filters like median, it's not real smart to read in a whole kernel, sort it and pick the median. It's much better to use things like self-sorting linked lists. For images which will not fit in memory, you have to take a banding approach. What I have seen of the NSImage family of classes is that they are extremely easy to use for opening image files and displaying them and storing the data. However, they are not well suited without a *lot* of intervening work in order to get bitmap representations of image data in a well-conditioned form. If you have ever implemented image processing code, it is incredibly troublesome to try and special-case every possible image data format within the processing code. It's just about always easier to put the data into a known format and convert it back later if need be. Also, be aware that sharpening the RGB channels of a color image individually is *not* the same as sharpening the V channel of the same image in HSV space. Also the order of black-to-white makes a difference. Sometimes black is a zero value, although, sometimes 00x is actually the max white value. There have been some half-baked replies about how to accomplish this task, which substantially under-represent the difficulty of the problem. More reasonable responses tend to agree with what I found a few months ago with Rhapsody DR/1 - that it is possible, but not without a lot of work. In my mind, this is the fundmental problem with frameworks. For what the authors intended, the NS classes are awesome. But, if you need to do something they didn't anticipate, the Framework can make life more difficult. For YB, this is exacerbated by the fact that you only get "black-box" classes to work with. Years of experience tell me that "clear-box" classes are far more robust and useful in real life. Again that can lead back to another OO Holy War(tm), but it is a fact that a lot of other programmers have come face to face with.
From: Bill Bumgarner <bbum@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 27 May 1998 16:57:32 -0400 Organization: Spacelab.net Internet Access Message-ID: <Pine.NXT.3.96.980527164001.19969A-100000@pathos> References: <6khfnm$af3$1@news.cs.tu-berlin.de> <1998052719541700.PAA15634@ladder01.news.aol.com> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII To: LarrySB <larrysb@aol.com> In-Reply-To: <1998052719541700.PAA15634@ladder01.news.aol.com> OK.. can't resist. On 27 May 1998, LarrySB wrote: > Ok, I've let my temper get the best of me in my previous posts. I will endeavor > to de-escalate now. > [ ... lots of useful image processing theory deleted ... ] > > What I have seen of the NSImage family of classes is that they are extremely > easy to use for opening image files and displaying them and storing the data. > However, they are not well suited without a *lot* of intervening work in order > to get bitmap representations of image data in a well-conditioned form. If you > have ever implemented image processing code, it is incredibly troublesome to > try and special-case every possible image data format within the processing > code. It's just about always easier to put the data into a known format and > convert it back later if need be. Ask yourself this. Why is there so much intervening work? Is it because NSImage makes it hard or because it supports so many image formats in such a transparent manner for loading/saving/imaging that supporting a similar set for processing is a lot of work? I.e. would you rather have a kit that can open/save/render a bazillion different image formats or one that limits you to open/save/render only a few? Better question; Would you want the NSImage classes to automatically convert *any* image into a limited set of formats that make processing easy even when 90% of the developers are just going to blast the bits on teh screen? As you say-- it is just about always easier to toss the bits into a known format and then convert back later if needed. Nothing about NSBitmapImageRep prevents you from doing that! Someone previously posted a very useful code snipped that will do exactly that! So, convert to a useful format... munge, munge, munge... convert back. > In my mind, this is the fundmental problem with frameworks. For what the > authors intended, the NS classes are awesome. But, if you need to do something > they didn't anticipate, the Framework can make life more difficult. For YB, > this is exacerbated by the fact that you only get "black-box" classes to work > with. Years of experience tell me that "clear-box" classes are far more robust > and useful in real life. So, just because the designers of the framework DIDN'T make a particular, relatively esoteric [yes, image processing is common in your world-- but it is not something that most developers will have regular need to do] feature, an easy and highly supported thing to do makes the ENTIRE framework useless? I don't understand how the frameworks make it any more difficult-- for that matter, they seem to make it a lot *less* difficult.... Regardless of the system, the bits still have to be twiddled into a format that your image processing algorithms can deal with... if it happens inside your code before being passed off to the frameworks, as an extension to the frameworks, or external to the application before being loaded by the app, it STILL HAS TO BE DONE. With the NS* frameworks, at least you do the conversion to a useful format ONCE-- raising exceptions for any formats that you can't/don't deal with (not can't as incapable, can't as in not enough return on investment to spend the time doing so)-- the framework objects then make imaging, etc, very easy.... Now, for particularly large images are particularly memory-hungry processing, you mention the need to stripe or otherwise subdivide the image processing mechanism... Fine-- go for it-- it can certainly can be [and has been] done. There is nothing about NSBitmapImageRep that makes this task any more/less difficult than it would be otherwise; multiple NSBitmapImageRep's can be instantiated against a single NSData object-- each pointing to a different hunk of memory... Alternatively, you can create sub-data objects against a single data object... Considering that a data object is just a pointer to a piece of memory with an associated length, there is nothing preventing you from creating your own disk paging based system or other optimizations necessary to handle seriously huge images. I wish Stan felt the need to comment-- he is one of the primary developers of Tiffany3... T3 has an amazing set of features-- way beyond photoshop-- with the ability to multithread image processing as well as distribute the image processing out across mutliple machines. As well, it supports air brushed undo-- i.e. apply an effect, then use an air brush to undo bits and pieces... The point is that all of these features require the exact set of features that you have asked for and all of the features in T3 are implemented through YB. It runs on NT/95/DR2/OpenStep and was developped by only one or two individuals in less than a third of the time of PhotoShop. And it is fast. In other words, T3 is *working* *running* *portable* *fast* *really cool* *yellow box* demo of everything you want from YB in terms of image manipulation. > Again that can lead back to another OO Holy War(tm), but it is a fact that a > lot of other programmers have come face to face with. A fact is an inarguable truth. An opinion is a statement of personal belief that may or may not be fact. Sounds like opinion, not fact.
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 27 May 1998 16:04:39 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6khdin$btt$2@ns3.vrx.net> References: <6khfnm$af3$1@news.cs.tu-berlin.de> <1998052719541700.PAA15634@ladder01.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: larrysb@aol.com In <1998052719541700.PAA15634@ladder01.news.aol.com> LarrySB claimed: > What I have seen of the NSImage family of classes is that they are extremely > easy to use for opening image files and displaying them and storing the data. Correct. > However, they are not well suited without a *lot* of intervening work in order > to get bitmap representations of image data in a well-conditioned form. Incorrect, as shown in any number of attempts. Mine was too simple and created two buffers, but since then people have posted quite short examples that do just that and use up maybe 20 lines of code. In the end the "best" solution was to convert to a known space, then hand you the pointer. This is effectively identical to converting it into a RGB format and placing it into a GWorld, with the exception that it took about 2 lines of code to do. > authors intended, the NS classes are awesome. But, if you need to do something > they didn't anticipate, the Framework can make life more difficult. You have failed to demonstrate this. How exactly does YB make this MORE difficult? Your other examples were just plain WRONG. People talking directly to serial ports? Changing colors and then asking for them in a different color space? Come on. Maury
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <1998052723270000.TAA11534@ladder01.news.aol.com> Date: 27 May 1998 23:27:00 GMT Organization: AOL http://www.aol.com References: <Pine.NXT.3.96.980527164001.19969A-100000@pathos> In a message dated 5/27/98 3:57:50 PM, Bill Bumgarner <bbum@codefab.com> wrote: >OK.. can't resist. > >On 27 May 1998, LarrySB wrote: > >> Ok, I've let my temper get the best of me in my previous posts. I will endeavor >> to de-escalate now. >> >[ ... lots of useful image processing theory deleted ... ] >> >> What I have seen of the NSImage family of classes is that they are extremely >> easy to use for opening image files and displaying them and storing the data. > >> However, they are not well suited without a *lot* of intervening work in order >> to get bitmap representations of image data in a well-conditioned form. If you >> have ever implemented image processing code, it is incredibly troublesome to >> try and special-case every possible image data format within the processing >> code. It's just about always easier to put the data into a known format and >> convert it back later if need be. > >Ask yourself this. Why is there so much intervening work? Is it because >NSImage makes it hard or because it supports so many image formats in such >a transparent manner for loading/saving/imaging that supporting a similar >set for processing is a lot of work? I think it is obvious. As I said, the framework designers had only considered what they felt was the most common case. They had failed to consider *every* possible way someone might want to use it. Hmm, sounds like an impossibility, considering every possible thing every possible user might want to do. That's precisely the point. >I.e. would you rather have a kit that can open/save/render a bazillion >different image formats or one that limits you to open/save/render only a >few? > >Better question; Would you want the NSImage classes to automatically >convert *any* image into a limited set of formats that make processing >easy even when 90% of the developers are just going to blast the bits on >teh screen? > >As you say-- it is just about always easier to toss the bits into a known >format and then convert back later if needed. > >Nothing about NSBitmapImageRep prevents you from doing that! Someone >previously posted a very useful code snipped that will do exactly that! > >So, convert to a useful format... munge, munge, munge... convert back. Ok fine. No big deal. Would be easier if: 1. Had better documentation. 2. Had some notion of the class internals. If number 2 is not a possiblity, then particular emphasis needs to be on number 1. >> In my mind, this is the fundmental problem with frameworks. For what the >> authors intended, the NS classes are awesome. But, if you need to do something >> they didn't anticipate, the Framework can make life more difficult. For YB, >> this is exacerbated by the fact that you only get "black-box" classes to work >> with. Years of experience tell me that "clear-box" classes are far more robust >> and useful in real life. > >So, just because the designers of the framework DIDN'T make a particular, >relatively esoteric [yes, image processing is common in your world-- but >it is not something that most developers will have regular need to do] >feature, an easy and highly supported thing to do makes the ENTIRE >framework useless? Well, it isn't all that uncommon. Maybe it's very uncommon in the niche *you* play in. It is a matter of perspective, wouldn't you agree? If you spend your life reading and writing database records and calculating npv, image processing would seem very esoteric indeed. Of course the converse is true as well. We all see the world from our own perspective. >I don't understand how the frameworks make it any more difficult-- for >that matter, they seem to make it a lot *less* difficult.... Getting back to perspective, you can look at it as a feature, difficulty, or bug depending on where you are coming from. >Regardless of the system, the bits still have to be twiddled into a format >that your image processing algorithms can deal with... if it happens >inside your code before being passed off to the frameworks, as an >extension to the frameworks, or external to the application before being >loaded by the app, it STILL HAS TO BE DONE. > >With the NS* frameworks, at least you do the conversion to a useful format >ONCE-- raising exceptions for any formats that you can't/don't deal with >(not can't as incapable, can't as in not enough return on investment to >spend the time doing so)-- the framework objects then make imaging, etc, >very easy.... > >Now, for particularly large images are particularly memory-hungry >processing, you mention the need to stripe or otherwise subdivide the >image processing mechanism... Fine-- go for it-- it can certainly can >be [and has been] done. There is nothing about NSBitmapImageRep that >makes this task any more/less difficult than it would be otherwise; >multiple NSBitmapImageRep's can be instantiated against a single NSData >object-- each pointing to a different hunk of memory... Alternatively, you >can create sub-data objects against a single data object... Considering >that a data object is just a pointer to a piece of memory with an >associated length, there is nothing preventing you from creating your own >disk paging based system or other optimizations necessary to handle >seriously huge images. > >I wish Stan felt the need to comment-- he is one of the primary developers >of Tiffany3... T3 has an amazing set of features-- way beyond photoshop-- >with the ability to multithread image processing as well as distribute the >image processing out across mutliple machines. As well, it supports air >brushed undo-- i.e. apply an effect, then use an air brush to undo bits >and pieces... > >The point is that all of these features require the exact set of features >that you have asked for and all of the features in T3 are implemented >through YB. It runs on NT/95/DR2/OpenStep and was developped by only one >or two individuals in less than a third of the time of PhotoShop. And it >is fast. > >In other words, T3 is *working* *running* *portable* *fast* *really cool* >*yellow box* demo of everything you want from YB in terms of image >manipulation. I've seen T3, and it is indeed a very impressive program. Indeed, it does a lot of the things I need to do. Seeing T3 left me with confidence that what I needed to do was indeed possible in YB. I ran into these bitmap problems probably 20 minutes after I started my own first YB project. Now, you may consider my problem domain very esoteric - but I damn sure consider it quite normal. It really left me wondering that if the YB frameware were so restrictive as to not leave any clear cut means of getting something so basic (for me) then what else had the framework designers left out? Certainly, the MacOS is very ameniable to my line of work. PowerPlant, TCL and MacApp don't pose any particular challenges either. Even Win32 and MFC can be reasonably adapted for image processing without a concerted effort. In all of the frameworks above, I can at least review the source code if the documentation is lacking. I can understand what some feature/difficulty/bug is doing and possibly how to exploit/workaround/fix said anomoly. >> Again that can lead back to another OO Holy War(tm), but it is a fact that a >> lot of other programmers have come face to face with. > >A fact is an inarguable truth. >An opinion is a statement of personal belief that may or may not be fact. > >Sounds like opinion, not fact. > I dunno, people argue over facts all the time. However, the software development world is repleat with examples of a feature/difficulty/bug in an application framework that was resolved by looking at the inner workings. The only other avenue would have been phoning up the responsible engineer and asking what was going on. To think that the same kind of problems *don't* exist in YB is pretty naive. What irks me is the way some people hype the YB - that it is perfectly suited for all aplications and it will take no time at all to produce really great applications with it. Then you run into reality sooner or later and find that it does indeed have plenty of shortcomings. Then the same people who hyped it as the SilverBullet(tm) back pedal by claiming that what I want to do is somehow esoteric and outside the mainstream so it really doesn't matter anyway. Yeah, right. So what do 90% of NS applications do anyway?
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.msx,comp.sys.ncr,comp.sys.newton,comp.sys.newton.announce,comp.sys.newton.misc,comp.sys.newton.programmer,comp.sys.next,comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.nsc,comp.sys.nsc.32k,comp.sys.palmtops,comp.sys.pen,comp.sys.pens,comp.sys.powerpc,comp.sys.powerpc.advocacy Subject: cmsg cancel <896304538.501894@nexus.polaris.net> Control: cancel <896304538.501894@nexus.polaris.net> Date: 27 May 1998 21:43:40 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.896304538.501894@nexus.polaris.net> Sender: "Shannonrx7" <shannonrx7@juno.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.msx,comp.sys.ncr,comp.sys.newton,comp.sys.newton.announce,comp.sys.newton.misc,comp.sys.newton.programmer,comp.sys.next,comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.nsc,comp.sys.nsc.32k,comp.sys.palmtops,comp.sys.pen,comp.sys.pens,comp.sys.powerpc,comp.sys.powerpc.advocacy Subject: cmsg cancel <896308819.951908@nexus.polaris.net> Control: cancel <896308819.951908@nexus.polaris.net> Date: 27 May 1998 23:17:19 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.896308819.951908@nexus.polaris.net> Sender: "Shannonrx7" <shannonrx7@juno.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: quinlan@intergate.bc.ca Newsgroups: comp.sys.next.programmer Subject: Re: Best development environment, a nitpick Date: Thu, 28 May 1998 02:45:11 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6kij3o$eoj$1@nnrp1.dejanews.com> References: <6kfcm3$gq$1@nnrp1.dejanews.com> <6kgbuc$8s6$1@nz12.rz.uni-karlsruhe.de> In article <6kgbuc$8s6$1@nz12.rz.uni-karlsruhe.de>, wori0011@rz03.FH-Karlsruhe.DE (Richard Woeber) wrote: > ever written a device-driver in smalltalk? No. You probably have a point but I don't see it unless you are just pointing out that smalltalk is not the right tool for all problems. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: Jared Brockway <brockway+@cs.cmu.edu.nospam> Newsgroups: comp.sys.next.programmer Subject: Re: Rhapsody and jdk classpaths... Date: 28 May 1998 03:20:29 GMT Organization: Carnegie Mellon Univ. -- Computer Science Dept. Message-ID: <6kil5t$ctc$1@goldenapple.srv.cs.cmu.edu> References: <6kcqo8$943@drn.newsguy.com> In-Reply-To: <6kcqo8$943@drn.newsguy.com> >Where do you set the classpath for java From the Rhapsody release notes at Apple's web site: CLASSPATH Sets the path to use for looking up class files. If the variable is not defined in the environment, you can set it in a shell script using "javaconfig DefaultClasspath". If javaconfig does not return a class path, the default /System/Library/Java:/System/Library/Frameworks/JavaVM.framework/Class es/classes.jar is used. (On Windows, substitute the value of NEXT_ROOT for /System). Specifying a class path on the command line overrides this variable. The default is not automatically searched and has to be added to the class path if one is specifed.
From: jesjones@halcyon.com (Jesse Jones) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 27 May 1998 20:33:12 -0700 Organization: Edmark Message-ID: <jesjones-ya02408000R2705982033120001@news.accessone.com> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <6kgtva$nc0$9@ns3.vrx.net>, maury@remove_this.istar.ca (Maury Markowitz) wrote: > In <jesjones-ya02408000R2605982143170001@news.accessone.com> Jesse Jones > claimed: > > MacApp has moved rather heavily into MI: > > Fair enough, haven't looked at it since 3.0. > > > TObject and added a bunch of mixin classes. Also Taligent's CommonPoint > > frameworks are built from the ground up around MI (Taligent published an > > excellent C++ style guide that includes a nice section on using MI in C++ > > without stumbling into the virtual base class nightmare). > > However I don't believe it ever released on the Mac, no? Well no, but it is an example of a relatively recent C++ framework that was designed by people who clearly knew a lot about OOD (eg Erich Gamma was at Taligent). -- Jesse
From: "Michelle L. Buck" <buck.erik@mcleod.net> Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 27 May 1998 22:45:31 -0500 Organization: McleodUSA - http://www.mcleodusa.net Message-ID: <6kimof$g1k1@odie.mcleod.net> References: <Pine.NXT.3.96.980527164001.19969A-100000@pathos> <1998052723270000.TAA11534@ladder01.news.aol.com> LarrySB wrote in message <1998052723270000.TAA11534@ladder01.news.aol.com>... >>So, convert to a useful format... munge, munge, munge... convert back. > >Ok fine. No big deal. > >Would be easier if: >1. Had better documentation. >2. Had some notion of the class internals. > >If number 2 is not a possiblity, then particular emphasis needs to be on number >1. > >>> In my mind, this is the fundmental problem with frameworks. For what the >>> authors intended, the NS classes are awesome. But, if you need to do >something >>> they didn't anticipate, the Framework can make life more difficult. For YB, >>> this is exacerbated by the fact that you only get "black-box" classes to >work >>> with. Years of experience tell me that "clear-box" classes are far more >robust >>> and useful in real life. >> >>So, just because the designers of the framework DIDN'T make a particular, >>relatively esoteric [yes, image processing is common in your world-- but >>it is not something that most developers will have regular need to do] >>feature, an easy and highly supported thing to do makes the ENTIRE >>framework useless? > >Well, it isn't all that uncommon. Maybe it's very uncommon in the niche *you* >play in. It is a matter of perspective, wouldn't you agree? If you spend your >life reading and writing database records and calculating npv, image processing >would seem very esoteric indeed. Of course the converse is true as well. We >all see the world from our own perspective. > >>I don't understand how the frameworks make it any more difficult-- for >>that matter, they seem to make it a lot *less* difficult.... > >Getting back to perspective, you can look at it as a feature, difficulty, or >bug depending on where you are coming from. > >>Regardless of the system, the bits still have to be twiddled into a format >>that your image processing algorithms can deal with... if it happens >>inside your code before being passed off to the frameworks, as an >>extension to the frameworks, or external to the application before being >>loaded by the app, it STILL HAS TO BE DONE. >> [Deleted] > >I've seen T3, and it is indeed a very impressive program. Indeed, it does a >lot of the things I need to do. Seeing T3 left me with confidence that what I >needed to do was indeed possible in YB. > >I ran into these bitmap problems probably 20 minutes after I started my own >first YB project. Now, you may consider my problem domain very esoteric - but >I damn sure consider it quite normal. It really left me wondering that if the >YB frameware were so restrictive as to not leave any clear cut means of getting >something so basic (for me) then what else had the framework designers left >out? > >Certainly, the MacOS is very ameniable to my line of work. PowerPlant, TCL and >MacApp don't pose any particular challenges either. Even Win32 and MFC can be >reasonably adapted for image processing without a concerted effort. > >In all of the frameworks above, I can at least review the source code if the >documentation is lacking. I can understand what some feature/difficulty/bug is >doing and possibly how to exploit/workaround/fix said anomoly. > Cool, where can I get the source code to Win32 ? Come on...I don't know anything about MacOS imaging APIs, but Win32 and MFC SUCK! They provide basically zero support or help. > >I dunno, people argue over facts all the time. However, the software >development world is repleat with examples of a feature/difficulty/bug in an >application framework that was resolved by looking at the inner workings. > >The only other avenue would have been phoning up the responsible engineer and >asking what was going on. To think that the same kind of problems *don't* >exist in YB is pretty naive. > Documentation can always be better. I don't understand what you want. Are you saying that you will not be happy with yellow box unless all images are automatically stored in exacly the format your application requires ? What about my application that wants a different format ? Why not allocate memory for your image via malloc or NSData ? Do whatever you want using C pointer de-refs or assembly language. When you are done and you want to display your image, create an NSImage and initialize it with your data. YB is not perfect, it is just as close and man has come to date. >What irks me is the way some people hype the YB - that it is perfectly suited >for all aplications and it will take no time at all to produce really great >applications with it. Then you run into reality sooner or later and find that >it does indeed have plenty of shortcomings. Then the same people who hyped it >as the SilverBullet(tm) back pedal by claiming that what I want to do is >somehow esoteric and outside the mainstream so it really doesn't matter anyway. > >Yeah, right. I program commercial avionics for embedded systems. YB does not help me with that directly either. What it does is let me build GUI tools that do help. Then I get so productive building avionics that I take the market by storm. YB does not make your imaging application any harder to write than any of the common alternatives. It will make all of the non imaging portions of your application very much easier. Heck, you may even discover that the imaging operations (things like built in transparency and automatic format conversion) are very powerful and convenient once you "get it". YB does thing differently than other systems (BETTER). > >So what do 90% of NS applications do anyway? > What exactly is your disappointment ? Did you think you weren't going to have to do anything ? How can you even begin to compare YB with MFC or Wind32.
Newsgroups: comp.sys.next.programmer From: Timothy J Luoma <nospam+yes-this-is-a-valid-address@luomat.peak.org> Sender: luomat@luomat Subject: apache 1.3b7 compile errors Message-ID: <Pine.NXT.3.96.980527213549.15207A-100000@luomat> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Date: Thu, 28 May 1998 01:39:59 GMT NNTP-Posting-Date: Wed, 27 May 1998 18:39:59 PDT Organization: @Home Network http_main.c: In function `process_child_status': http_main.c:3921: request for member `w_S' in something not a structure or union http_main.c:3921: request for member `w_T' in something not a structure or union http_main.c:3929: request for member `w_S' in something not a structure or union http_main.c:3929: request for member `w_T' in something not a structure or union Ugh ugh ugh....Searched Dejanews for a workaround and couldn't find any... TjL
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6ki02m$c9k$3376@dalen.get2net.dk> Control: cancel <6ki02m$c9k$3376@dalen.get2net.dk> Date: 28 May 1998 01:59:50 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6ki02m$c9k$3376@dalen.get2net.dk> Sender: KOYLA<koyla21@get2net.dk> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: jesjones@halcyon.com (Jesse Jones) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 27 May 1998 21:01:32 -0700 Organization: Edmark Message-ID: <jesjones-ya02408000R2705982101320001@news.accessone.com> References: <6kcc12$hb1$3@news.idiom.com> <1998052617163600.NAA23784@ladder03.news.aol.com> <6kej3r$366$1@crib.bevc.blacksburg.va.us> <jesjones-ya02408000R2605982218490001@news.accessone.com> <6kgahc$4lv$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <6kgahc$4lv$1@crib.bevc.blacksburg.va.us>, nurban@vt.edu wrote: > In article <jesjones-ya02408000R2605982218490001@news.accessone.com>, jesjones@halcyon.com (Jesse Jones) wrote: > > > Let me give you an example from my own > > framework Raven: the document class TDocument. This is the model in the MVC > > pattern. It descends from: > > > > 1) MCommander which enables it to respond to menu commands (chain of > > reponsibility pattern). > > 2) MStreamable so that the document can be streamed in and out. > > 3) MReferenceCounted so that the document can be deleted automatically when > > the last view goes away (typically when the user closes a window). > > 4) MBroadcaster<SDocumentMessage> so that the document can broadcast to > > interested objects as the model state changes (MVC and Observer patterns). > > > How would something like my TDocument class be written in Objective-C? > > I mentioned this elsewhere in the thread, but.. you would generally make > each of those mixin classes into regular Objective-C classes, and then > _compose_ them into the derived object rather than _inheriting_ them. > Composition is more flexible anyway. You also typically create protocols > (like Java interfaces, or akin to C++ pure virtual abstract interface > mixins) for the derived class to conform to (i.e., telling the compiler > and runtime system that the object implements that interface, without it > actually inheriting any implementations). So the TDocument class would > conform to those four protocols, meaning that all of the methods are > known to exist within the class, but it doesn't inherit any implementation > from the protocols; that is done separately by composition. I agree that composition is more flexible than inheritance, but often you don't need this extra flexibility and there are places where it just doesn't make sense (as in my TDocument example above). Below you mention that NSObject includes the functionality that I provide in MCommander, MStreamable, and MReferenceCounted. This is very similar to old C++ frameworks, and is IMO, less than ideal. I find small narrowly tailored mixin classes much easier to work with than obese base classes that include all sorts of cruft in case a subclass requires it. Careful use of MI allows classes to be designed this way rather easily. I think it's a pity that Java and (apparently) Objective-C don't accomadate this style of programming. -- Jesse
From: pemmerik@solair1.inter.NL.net (P.J.L.van Emmerik) Newsgroups: comp.sys.next.programmer Subject: Upgrading from NS3.2 to OpenStep 4.2, Pointers to info wanted. Date: Thu, 28 May 1998 06:54:45 GMT Organization: NLnet Message-ID: <6kj1ns$1ho$1@news.NL.net> I want to move some apps to OpenStep 4.2. In the past i have seen some info on this subject, like a porting guide, but i can not find this info any more. Pointers to the info an experiences with upgrading to OpenStep 4.2 are welcome. Please Email to: emmerik@qtecq.com P.J.L. van Emmerik Holec Projects B.V. Email: emmerik@qtecq.com PO.BOX 565, 7550 AN Hengelo pemmerik@solair1.inter.NL.net The Netherlands Phone: +31 74 2558 688 --
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <1998052807155500.DAA29116@ladder03.news.aol.com> Date: 28 May 1998 07:15:55 GMT Organization: AOL http://www.aol.com References: <6kimof$g1k1@odie.mcleod.net> "Michelle L. Buck" <buck.erik@mcleod.net> wrote: <<<< YB does thing differently than other systems (BETTER). >>> Not from my standpoint. In the case of yellowbox, *my* choice of programming technique and style are severely restricted. I'm also of the opinion that YB frameworks may actually create more work than the common C++ OS-specific class libraries. The framework leaves a lot to be desired in terms of documentation and flexibility. For a supposedly mature OS, it is surprisingly lacking in many areas. I have written and shipped commercial software packages on multiple platforms. I'm not a student somewhere, I'm a real life get paid for it programmer. As such, I have a right to criticise what I think are less productive ways to make a living. As much as some people may turn their noses up at it, a procedural API is not entirely a bad thing. Neither are multiple inheritance and ahem, C++. I would probably agree with most criticisms of C++. But you can't deny that fact that it *does* work. It's had years of evolution that some may reguard as cruft, but indeed C++ does the job. You might call it bloated, but I would call it well-worn and tested. When I first saw YB, I was enthusiastic. After I dug into, the euphoria of newness wore off. A little further into YB and I'm begining to realize a substantial number of shortcomings. YB is not the Panacea some people claim it is. It has some serious flaws from where I sit. Maybe, in the long run, it really is worth it. But before diving into it and commiting substantial development resources, you've got to hear both sides of it. A realistic perspective: 1. Must code in Obj-C or possibly Java. a. requires learning a new language with its own little peccadillos. The basics are easy, the intricate little details require a long time as with any other language. b. Some techniques, such as MI, are eliminated and this may pose problems for developers who favor them. c. Obj-C offers some constructs that may foster better coding styles than C or C++. d. Obj-C hasn't had the wide exposure of C++, which is used by thousands of developers around the world, it hasn't had the benefit of widespread use. e. Java is still a very immature programming language. 2. There is no "api" per-se. Indeed, the whole thing is a set of frameworks. a. Programming effort may be reduced in many cases. b. Unlike familiar class libraries (MFC, PowerPlant, etc) source is not available. c. Documentation is lacking in many areas. d. The framework may not offer methods or data unique to your situation, and potentially require difficult work-arounds, complicated by the lack of source and documentation. 3. Deployment of completed applications at risk. a. very limited number of OpenStep installations in the field. b. Unknown royalty conditions for Win32 runtime. c. No MacOS platform deployment for a minimum of a year (OS-X). d. Rhapsody on Apple RISC platforms is not likely to reach beyond a tiny fraction of the already small Mac marketshare. e. A very real possibility that YB will never ship, should a shift in Apple management or priorities occur. Now, if you are not bothered by anything in the list, then go for it! But the rest of us may find these conditions unacceptable, at least today.
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 28 May 1998 00:39:12 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6kippg$79h$1@crib.bevc.blacksburg.va.us> References: <6kcc12$hb1$3@news.idiom.com> <jesjones-ya02408000R2605982218490001@news.accessone.com> <6kgahc$4lv$1@crib.bevc.blacksburg.va.us> <jesjones-ya02408000R2705982101320001@news.accessone.com> In article <jesjones-ya02408000R2705982101320001@news.accessone.com>, jesjones@halcyon.com (Jesse Jones) wrote: > Below you mention > that NSObject includes the functionality that I provide in MCommander, > MStreamable, and MReferenceCounted. NSObject doesn't include the MCommander stuff, that's in NSResponder. > This is very similar to old C++ > frameworks, and is IMO, less than ideal. IMO, NSObject is exactly the right place for both object serialization and reference counting. This is behavior that really ought to be in all objects, or else container classes don't work well, you can't treat objects generically, etc. > I find small narrowly tailored > mixin classes much easier to work with than obese base classes that include > all sorts of cruft in case a subclass requires it. The stuff in NSObject is used more or less by _all_ subclasses, at least in the OpenStep framework.
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 28 May 1998 10:05:20 GMT Organization: Idiom Communications Message-ID: <6kjct0$d08$2@news.idiom.com> References: <6kcc12$hb1$3@news.idiom.com> <1998052617163600.NAA23784@ladder03.news.aol.com> <6kej3r$366$1@crib.bevc.blacksburg.va.us> <jesjones-ya02408000R2605982218490001@news.accessone.com> <6kgahc$4lv$1@crib.bevc.blacksburg.va.us> <jesjones-ya02408000R2705982101320001@news.accessone.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: jesjones@halcyon.com Jesse Jones may or may not have said: [snip] -> I agree that composition is more flexible than inheritance, but often you -> don't need this extra flexibility and there are places where it just -> doesn't make sense (as in my TDocument example above). Below you mention -> that NSObject includes the functionality that I provide in MCommander, -> MStreamable, and MReferenceCounted. Well, to be precise: categories of NSObject and NSResponder include that functionality. -> This is very similar to old C++ frameworks, and is IMO, less than ideal. I -> find small narrowly tailored mixin classes much easier to work with than -> obese base classes that include all sorts of cruft in case a subclass -> requires it. In Objective-C, classes are typically coded in several "categories" which provide logical grouping of methods with finer granularity than a class. In place of your "small narrowly tailored mixin classes", we use small narrowly tailored categories, and we place them at the appropriate point in the class tree. If I have a category for, say, text-to-speech output, I'd attach it to NSString, not to NSObject. If I had a category for describing an object's contents in ASCII form for debugging, I'd attach that category to NSObject. At runtime, a category is loaded only if any of the methods in that category are called. -> Careful use of MI allows classes to be designed this way rather easily. I -> think it's a pity that Java and (apparently) Objective-C don't accomadate -> this style of programming. I'm certainly not going to defend Java. Gosling knew better, and shipping a language whose syntax is so close to C++ is simply inexcusable. As for style, Objective-C lends itself to better code organization and legibility than does C++. -jcr
From: Amando Blasco <ablasco@mad.servicom.es> Newsgroups: comp.sys.next.programmer Subject: JAVA ON OPENSTEP Date: Thu, 28 May 1998 12:03:35 +0200 Organization: GamesReview Message-ID: <356D3676.F4215222@mad.servicom.es> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; x-mac-type="54455854"; x-mac-creator="4D4F5353" Content-Transfer-Encoding: 7bit Can I program in Java with Openstep 4.2 or Nextstep 3.3? I know Java is available for major platforms, but are these O.S included in? Regards, Amando Blasco
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 28 May 1998 10:24:37 GMT Organization: Technical University of Berlin, Germany Message-ID: <6kje15$9dn$1@news.cs.tu-berlin.de> References: <Pine.NXT.3.96.980527164001.19969A-100000@pathos> <1998052723270000.TAA11534@ladder01.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit larrysb@aol.com (LarrySB) writes: >>Nothing about NSBitmapImageRep prevents you from doing that! Someone >>previously posted a very useful code snipped that will do exactly that! >> >>So, convert to a useful format... munge, munge, munge... convert back. >Ok fine. No big deal. >Would be easier if: >1. Had better documentation. >2. Had some notion of the class internals. >If number 2 is not a possiblity, then particular emphasis needs to be on number >1. The documentation is all there: NSBitmapImageRep supports planar or packed image data with up to 5 components. So, -getBitmapDataPlanes: is the method to call, it will also work for the packed case, which will give you a single pointer. For 8 bit data, the packed routines can be identical to the planar ones by constructing planar pointers offset 1 byte from each other. >>I don't understand how the frameworks make it any more difficult-- for >>that matter, they seem to make it a lot *less* difficult.... >Getting back to perspective, you can look at it as a feature, difficulty, or >bug depending on where you are coming from. Not really. The AppKit is called that for a reason. It is a Kit containing *generic* functionality common to most/all Apps. It's not called the ImageKit, if yuou want that you'll have to look elsewhere. After all, Apple doesn't provide a SpreadsheetKit or a NaturalLanguageKit. [...] >I ran into these bitmap problems probably 20 minutes after I started my own >first YB project. Now, you may consider my problem domain very esoteric - but >I damn sure consider it quite normal. It really left me wondering that if the >YB frameware were so restrictive as to not leave any clear cut means of getting >something so basic (for me) then what else had the framework designers left >out? Again, what is cool about AppKit is that it allows you to concentrate on you application domain and not spend most of your time worrying about generic application issues (managing the event loop, windows, etc.). But you still have to code what is specific to your application. >Certainly, the MacOS is very ameniable to my line of work. PowerPlant, TCL and >MacApp don't pose any particular challenges either. Even Win32 and MFC can be >reasonably adapted for image processing without a concerted effort. The same is true for the AppKit, although it does make other things so much easier that it tends to be a surprise that you do have to write *some* code yourself. What else should the framework do? Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6kir9o$lca$13845@dalen.get2net.dk> Control: cancel <6kir9o$lca$13845@dalen.get2net.dk> Date: 28 May 1998 10:19:11 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6kir9o$lca$13845@dalen.get2net.dk> Sender: Dino<easymoney@get2net.dk> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 28 May 1998 10:35:34 GMT Organization: Idiom Communications Message-ID: <6kjelm$d08$3@news.idiom.com> References: <6khfnm$af3$1@news.cs.tu-berlin.de> <1998052719541700.PAA15634@ladder01.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: larrysb@aol.com LarrySB may or may not have said: [a dissertation on basic image processing] Larry, I'm an imaging expert myself (started at Scion Corporation in 1982, went to Media Cybernetics, went back to Scion, developed Scion's first frame grabber, worked on imaging projects ranging from text conversion to LANDSAT to video compression to radiology.) I'm going to give you a tip: The NSImage class is not what you want to use for serious image processing. It's a general-purpose, "draw an image from this file for me" class, and it was not designed to deal with large images or provide easy access to individual pixels, let alone 3x3 pixel neighborhoods. Also, if it follows the behaviour of NXImage, it will have a habit of trying to malloc all of the pixels in a single block, which loses terribly if the image is any large fraction of the size of your RAM. Now then, what's an imaging hacker to do? This strategy originated at TRW: (I think. One of those defense contractors, anyhow.) You keep your raw pixel storage in any format you like. Then, you write a "tile" class, whose job is to map your raw storage to the NSView that you're drawing in. What the PixelMaster developers did, (circa 1991) was just to let the View set up the postscript context, and then have the Tiles talk to the DPS interpreter. Each Tile was responsible for mapping about 4K (vm_pagesize) worth of pixels to the View. Now, it will probably make sense for your Tile class to be a subclass of NSImageRep, and conform to that API. That's your call. As for how you put the pixels in the view, in Rhapsody today, you use the postscript "image" operator. In MacOS X, you'll use whatever you like from Quickdraw, or you can still use the Postscript image operator (which is pretty much the same in PDF.) Do *not* try drawing the pixels one at a time. Do not stripe the image, segment it into squares. Do not try to alloc the entire pixel store in one block if your image can be bigger than your physical RAM. It doesn't hurt anything to have thousands of entries in your malloc table. -jcr
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: TIFF Error: Can only handle associated-alpha extra samples. Date: Thu, 28 May 1998 13:54:24 +0200 Organization: Square B.V. Message-ID: <356D5070.977F0F6B@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Under WinNT I sometimes get the message 'TIFF Error: Can only handle associated-alpha extra samples.' on the Project Builder console. The image works futher okay. The tiff was created using Corel Photopaint, the 'alpha' part is the selection in Corel, which seems to be interpreted as complete alpha. What does the warning mean? How can I prevent this? (besides not using alpha) Maurice le Rutte -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto:mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+ Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" and "to trick into believing or accepting as genuine something false and often preposterous"
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: access to [NSString cString] functionality? Date: 28 May 1998 12:59:10 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6mqnt1.6gj.rog@talisker.ohm.york.ac.uk> References: <slrn6mlrr8.4s8.rog@talisker.ohm.york.ac.uk> <6khd25$p6n$1@owl.inetnebr.com> On Wed, 27 May 1998 10:55:14 -0500, "Jeremy Bettis" <jeremy@hksys.com> wrote: > It is a simple concept, make a class that takes a void pointer in > init: then frees it on dealloc. But in your case, you are creating an > autoreleased NSData object, why not just use -[NSData bytes] to get an > autoreleased pointer to the bytes? because the NSString method dataUsingEncoding: method doesn't null terminate its string, and there's no way of making it return an NSMutableData object. your first suggestion is good, and i was stupid not to think of it! cheers, rog.
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 28 May 1998 12:30:06 GMT Organization: Technical University of Berlin, Germany Message-ID: <6kjlce$ch3$1@news.cs.tu-berlin.de> References: <6kimof$g1k1@odie.mcleod.net> <1998052807155500.DAA29116@ladder03.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit larrysb@aol.com (LarrySB) writes: >"Michelle L. Buck" <buck.erik@mcleod.net> wrote: ><<<< >YB does thing differently than other systems (BETTER). >>>> >Not from my standpoint. In the case of yellowbox, *my* choice of programming >technique and style are severely restricted. I'm also of the opinion that YB >frameworks may actually create more work than the common C++ OS-specific class >libraries. If you fight the framework and try to do things in a C++ style, you may actually achieve that. If you try and understand the framework and go with the flow, it will make life a lot easier for you. However, my experience in projects (echoed by many others) has been that programmers exposed to C++ have a *lot* more difficulty getting into Objective-C than for example people with *no* OO exposure. So you may actually have a somewhat difficult start ahead of you, but it gets a LOT better once you get into it. >The framework leaves a lot to be desired in terms of documentation and >flexibility. What exactly? So far, the only concrete example has been NSBitmapImgeRep, and that *is* documented. Well, it doesn't specifically say "you're on your own" for image processing, but again, what exactly did you expect? >For a supposedly mature OS, it is surprisingly lacking in many areas. I have Where? I am sure it is, but you haven't provided any examples yet. >written and shipped commercial software packages on multiple platforms. I'm not >a student somewhere, I'm a real life get paid for it programmer. As such, I >have a right to criticise what I think are less productive ways to make a >living. You can criticize all you like. But you will also have to accept that people who know the system you criticize, and have similar credentials, will take you to task for factual errors or substandard arguments. >As much as some people may turn their noses up at it, a procedural API is not >entirely a bad thing. Neither are multiple inheritance and ahem, C++. I would >probably agree with most criticisms of C++. But you can't deny that fact that >it *does* work. It's had years of evolution that some may reguard as cruft, >but indeed C++ does the job. You might call it bloated, but I would call it >well-worn and tested. Well, AppKit is also well-worn and tested. Without the bloat. It also works. Very well. Objective-C also does the job, usually much easier than C++, and it's about as old, but without much evolution. Why? Maybe because it didn't need to evolve that much because they got things right the first time. By copying from the right place. What exactly is your point here? This thread started by you claiming YB is *generally* inflexibile, inadequate and plain gets in your way, when so far you have offered *no* evidence of this. >When I first saw YB, I was enthusiastic. >After I dug into, the euphoria of newness wore off. >A little further into YB and I'm begining to realize a substantial number of >shortcomings. You realized that it didn't do *everything* for you? That you still had to write *some* code? >YB is not the Panacea some people claim it is. It has some serious flaws from Who, besides sales-droids, is making that claim? >where I sit. Maybe, in the long run, it really is worth it. But before diving >into it and commiting substantial development resources, you've got to hear >both sides of it. >A realistic perspective: >1. Must code in Obj-C or possibly Java. Can code core in *any* language you like. > a. requires learning a new language with its own little peccadillos. The >basics are easy, the intricate little details require a long time as with any >other language. The intricacies take longer than the basics, but still a lot less time than C++ mastery. And the intricacies are really not about Objective-C, but about well structured design. I doubt the same goes for being master of C++ constructors or other arcana. > b. Some techniques, such as MI, are eliminated and this may pose problems >for developers who favor them. > c. Obj-C offers some constructs that may foster better coding styles than >C or C++. Possible negates (b). You certainly can't have both... > d. Obj-C hasn't had the wide exposure of C++, which is used by thousands >of developers around the world, it hasn't had the benefit of widespread use. Do you mean benfits to the language itself? Or just the usual that you can more easily get staff trained in C++? There is an interesting section on the pitfalls posed by C++ in Bruce Webster's excellent "Pitfalls of Object Oriented Development" One pitfall is using C++, because your project will face numerous language induced difficulties and not really benefit from OO. The next pitfall is *not* using C++. Argh. >2. There is no "api" per-se. Indeed, the whole thing is a set of frameworks. > a. Programming effort may be reduced in many cases. > b. Unlike familiar class libraries (MFC, PowerPlant, etc) source is not >available. 1. Because the framework, in many respects, *is* the OS. 2. A couple of years ago, I would have agreed that this is a problem. I know very much think that it is a GOOD THING. 3. You don't really need source code that much. > c. Documentation is lacking in many areas. Where? Again, I am sure there are areas where it is lacking, but which exactly do you mean? > d. The framework may not offer methods or data unique to your situation, >and potentially require difficult work-arounds, complicated by the lack of >source and documentation. Unique methods are easily added using categories. I hardly every need to code workarounds. [political issues deleted] Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 28 May 1998 09:24:28 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6kjagc$q9p$5@ns3.vrx.net> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: jesjones@halcyon.com In <jesjones-ya02408000R2705982033120001@news.accessone.com> Jesse Jones claimed: > Well no, but it is an example of a relatively recent C++ framework that was > designed by people who clearly knew a lot about OOD (eg Erich Gamma was at > Taligent). I've heard it described as one of the worst frameworks though - remember I've never touched it. For instance someone posted an example of the hierarchy, "mouse" ended up being many levels deep and was a "pixel based random access point device controller" IIRC. Maury
From: John Hornkvist <sorry@no.more.spams> Newsgroups: comp.sys.next.programmer Subject: Optimized drawing with DPS Date: Tue, 26 May 1998 18:20:36 GMT Organization: Chalmers Tekniska Högskola Sender: john@haddock.cd.chalmers.se (John Hprnkvist) Distribution: World Message-ID: <EtKtMC.Mz3@haddock.cd.chalmers.se> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I have written an article on fast drawing for large views with a very large number of objects using DPS. You can find it at http://www.dtek.chalmers.se/~d95nhoj. If Scott wants to put it on StepWise, he is certainely welcome to do so. Regards, John Hornkvist Name: nhoj Address: cd.chalmers.se
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 28 May 1998 09:50:59 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6kjc23$q9p$6@ns3.vrx.net> References: <Pine.NXT.3.96.980527164001.19969A-100000@pathos> <1998052723270000.TAA11534@ladder01.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: larrysb@aol.com In <1998052723270000.TAA11534@ladder01.news.aol.com> LarrySB claimed: > I think it is obvious. As I said, the framework designers had only considered > what they felt was the most common case. They had failed to consider *every* > possible way someone might want to use it. Is dealing with an image that consists of files in any format, and can contain BOTH vector and bitmap representations "the most common case"? Hardly. In fact this is the ONLY framework I have seen that even considers more than one case. Now your claims are bordering on outlandish. YB indeed may have no built in functionality for dealing with pixels in a bitmap, but then again, neither does anyone else's basic framework. You've stated you had to write all of this yourself for PP, yet when a solution that let's you do the same thing for YB is presented you refuse to acknowlege it and continue to make the claim that YB is somehow getting in your way. > Hmm, sounds like an impossibility, considering every possible thing every > possible user might want to do. > > That's precisely the point. No it's not. You have stated in the past that YB makes your job *harder*. You have failed completely to provide examples of this, including this one. As far as I can tell, YB makes even this job much easier, because it at least does all of the file manipulation and loading for you. > >Nothing about NSBitmapImageRep prevents you from doing that! Someone > >previously posted a very useful code snipped that will do exactly that! > > > >So, convert to a useful format... munge, munge, munge... convert back. > > Ok fine. No big deal. Good. > Would be easier if: > 1. Had better documentation. > 2. Had some notion of the class internals. > > If number 2 is not a possiblity, then particular emphasis needs to be on number > 1. Maybe so, but again you're holding YB to a standard that you don't hold to PP (was it PP?). PP's docs are marginal at best, and the classes themselves basic to the extreme. For instance, nowhere in any of the documentation is it suggested that PP's App class knows how to deal with more than one Document subclass, nor how to use such functionality. YB's docs use such a case as one of the teaching examples. > >I don't understand how the frameworks make it any more difficult-- for > >that matter, they seem to make it a lot *less* difficult.... > > Getting back to perspective, you can look at it as a feature, difficulty, or > bug depending on where you are coming from. Look at WHAT as a feature or a bug? YB requires you to do the same thing as you do currently with the GWorld, yet does all the prior setup work for you. How can this be anything but good? > I ran into these bitmap problems probably 20 minutes after I started my own > first YB project. Now, you may consider my problem domain very esoteric - but > I damn sure consider it quite normal. It really left me wondering that if the > YB frameware were so restrictive as to not leave any clear cut means of getting > something so basic (for me) then what else had the framework designers left > out? So you're admitting you don't know? Then why all the fuss? > Certainly, the MacOS is very ameniable to my line of work. PowerPlant, TCL and > MacApp don't pose any particular challenges either. You mean aside from the fact that they have NO support at all for what you're doing and you had to write all the code yourself. > In all of the frameworks above, I can at least review the source code if the > documentation is lacking. Fair enough, point taken. > The only other avenue would have been phoning up the responsible engineer and > asking what was going on. To think that the same kind of problems *don't* > exist in YB is pretty naive. Perhaps so, but you haven't really demonstrated anything other than that you didn't find the solution in the docs. > What irks me is the way some people hype the YB - that it is perfectly suited > for all aplications and it will take no time at all to produce really great > applications with it. I get the feeling your app would likely have benefitted greatly, considering some of the code that's been posted. If you _remove_ the actual pointer math and bit manipulation, how many lines of code does your app have to get the image into RAM? That is, ask the user for a file, find and open the file, get the file into memory, and then provide you a handle to it. 100 lines? 500 lines? > as the SilverBullet(tm) back pedal by claiming that what I want to do is > somehow esoteric and outside the mainstream so it really doesn't matter anyway. I've seen no one make this claim. Maury
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 28 May 1998 09:57:58 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6kjcf6$q9p$7@ns3.vrx.net> References: <6kimof$g1k1@odie.mcleod.net> <1998052807155500.DAA29116@ladder03.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: larrysb@aol.com In <1998052807155500.DAA29116@ladder03.news.aol.com> LarrySB claimed: > Not from my standpoint. In the case of yellowbox, *my* choice of programming > technique and style are severely restricted. HOW?!?!? We;ve asked you numerous times to provide even a single example, and you have not. How does YB GET IN YOUR WAY??? > I'm also of the opinion that YB > frameworks may actually create more work than the common C++ OS-specific class > libraries. How? > The framework leaves a lot to be desired in terms of documentation and > flexibility. How? > For a supposedly mature OS, it is surprisingly lacking in many areas. Which exactly? > written and shipped commercial software packages on multiple platforms. I'm not > a student somewhere, I'm a real life get paid for it programmer. As such, I > have a right to criticise what I think are less productive ways to make a > living. So do I. What are these ways? > As much as some people may turn their noses up at it, a procedural API is not > entirely a bad thing. I know, and I use it in my YB programs in places. > When I first saw YB, I was enthusiastic. > After I dug into, the euphoria of newness wore off. > A little further into YB and I'm begining to realize a substantial number of > shortcomings. Which in the cases you presented were 100% avoidable. I've now seen no less than 4 complete solutions. > YB is not the Panacea some people claim it is. I don't think you know enough about it to comment. I base this statement on the fact that I knew a solution to your problem and I've never even looked at pixmap work under YB. > It has some serious flaws from where I sit. The documentation? > A realistic perspective: > > 1. Must code in Obj-C or possibly Java. Incorrect. My app has plenty of plain old C where I felt it was appropriate. > 2. There is no "api" per-se. Indeed, the whole thing is a set of frameworks. That is a meaningless statement. > d. The framework may not offer methods or data unique to your situation, > and potentially require difficult work-arounds, complicated by the lack of > source and documentation. NSData was created for specifically this purpose. You have failed to address this fact. > 3. Deployment of completed applications at risk. True, but this has little to do with the abilities of the framework itself. > Now, if you are not bothered by anything in the list, then go for it! But the > rest of us may find these conditions unacceptable, at least today. If they were true, sure. Maury
From: rex@smallandmighty.com (Eric King) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Thu, 28 May 1998 15:01:20 -0400 Organization: The Small & Mighty Group Message-ID: <rex-2805981501210001@192.168.0.3> References: <Pine.NXT.3.96.980527164001.19969A-100000@pathos> <1998052723270000.TAA11534@ladder01.news.aol.com> <6kjc23$q9p$6@ns3.vrx.net> In article <6kjc23$q9p$6@ns3.vrx.net>, maury@remove_this.istar.ca (Maury Markowitz) wrote: :In <1998052723270000.TAA11534@ladder01.news.aol.com> LarrySB claimed: :> I think it is obvious. As I said, the framework designers had only :considered :> what they felt was the most common case. They had failed to consider :*every* :> possible way someone might want to use it. : : Is dealing with an image that consists of files in any format, and can :contain BOTH vector and bitmap representations "the most common case"? :Hardly. In fact this is the ONLY framework I have seen that even considers :more than one case. Actually, the Quicktime image decompression manager does this. If you ask for a PICT file it will readily translate TIFF, JPEG, Photoshop, GIF, and a variety of other formats into PICT before passing it back to you, and it also handled vector representations via GX. -Eric
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <1998052819012200.PAA09263@ladder03.news.aol.com> Date: 28 May 1998 19:01:22 GMT Organization: AOL http://www.aol.com References: <6kjelm$d08$3@news.idiom.com> jcr.remove@this.phrase.idiom.com (John C. Randolph) wrote: <<<<<< I'm going to give you a tip: The NSImage class is not what you want to use for serious image processing. It's a general-purpose, "draw an image from this file for me" class, and it was not designed to deal with large images or provide easy access to individual pixels, let alone 3x3 pixel neighborhoods. >>>>>> Indeed, I was coming to the same conclusion on my own. <<<<<< Also, if it follows the behaviour of NXImage, it will have a habit of trying to malloc all of the pixels in a single block, which loses terribly if the image is any large fraction of the size of your RAM. Now then, what's an imaging hacker to do? >>>>>>>> I dunno, I am completely in the dark about how the internals work. <<<<<< This strategy originated at TRW: (I think. One of those defense contractors, anyhow.) You keep your raw pixel storage in any format you like. Then, you write a "tile" class, whose job is to map your raw storage to the NSView that you're drawing in. What the PixelMaster developers did, (circa 1991) was just to let the View set up the postscript context, and then have the Tiles talk to the DPS interpreter. Each Tile was responsible for mapping about 4K (vm_pagesize) worth of pixels to the View. >>>> Now, that sounds intriguing. <<<<< Now, it will probably make sense for your Tile class to be a subclass of NSImageRep, and conform to that API. That's your call. As for how you put the pixels in the view, in Rhapsody today, you use the postscript "image" operator. In MacOS X, you'll use whatever you like from Quickdraw, or you can still use the Postscript image operator (which is pretty much the same in PDF.) >>>>>> Ok, this is shaping up to be a worthwhile strategy. <<<<<< Do *not* try drawing the pixels one at a time. Do not stripe the image, segment it into squares. Do not try to alloc the entire pixel store in one block if your image can be bigger than your physical RAM. It doesn't hurt anything to have thousands of entries in your malloc table. -jcr >>>>>>> Ok, now I see how to leverage the VM scheme and possibly the NSImageRep inheritance. Thank you for a truly useful insight into the problem. So far there's been good input from several people, including Christian Brunschen, John Randolph, Roger Peppe and a few others. I thank you very much for your help and insight. Maury Markowitz and Nathan Urban have offered incomplete and over-simplistic solutions. These two also seem more bent on conducting gihad than on offering any concrete or substantive discussion. I will stop responding to either of them at this point. Frankly, their perspective appears clouded by a lack of real-world experience.
From: rex@smallandmighty.com (Eric King) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Thu, 28 May 1998 15:19:46 -0400 Organization: The Small & Mighty Group Message-ID: <rex-2805981519470001@192.168.0.3> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> In article <6kjagc$q9p$5@ns3.vrx.net>, maury@remove_this.istar.ca (Maury Markowitz) wrote: :In <jesjones-ya02408000R2705982033120001@news.accessone.com> Jesse Jones :claimed: :> Well no, but it is an example of a relatively recent C++ framework that was :> designed by people who clearly knew a lot about OOD (eg Erich Gamma was at :> Taligent). : : I've heard it described as one of the worst frameworks though - remember :I've never touched it. Go visit a Barnes & Noble and see if you can find a copy of Inside Taligent Technology. It contains an overview of CommonPoint's design. Also the latest versions of IBM's OpenClass framework for OS/2 and AIX are basically a 'best of Taligent.' There's a lot of nice design and functionality in the system. For instance, its graphics, text, and localization classes were very nicely done. Maybe as a whole the system was bloated, but I think the indidual components had a lot of merit. :For instance someone posted an example of the :hierarchy, "mouse" ended up being many levels deep and was a "pixel based :random access point device controller" IIRC. I've seen sample code from their OpenClass frameworks, and overall it doesn't look that bad. More complex than the BeOS, but a hell of a lot more functionality also. Personally, I think any system that tries to incorporate that level of functionality is going to run into serious problems. Even the Yellow Box. -Eric
From: rdieter@math.unl.edu (Rex Dieter) Newsgroups: comp.sys.next.programmer Subject: Re: apache 1.3b7 compile errors Date: 28 May 1998 16:03:13 GMT Organization: University of Nebraska-Lincoln Message-ID: <6kk1s1$su2$1@unlnews.unl.edu> References: <Pine.NXT.3.96.980527213549.15207A-100000@luomat> In article <Pine.NXT.3.96.980527213549.15207A-100000@luomat> Timothy J Luoma <nospam+yes-this-is-a-valid-address@luomat.peak.org> writes: > > http_main.c: In function `process_child_status': http_main.c:3921: > request for member `w_S' in something not a structure or union > http_main.c:3921: request for member `w_T' in something not a structure > or union http_main.c:3929: request for member `w_S' in something not a > structure or union http_main.c:3929: request for member `w_T' in > something not a structure or union > > > Ugh ugh ugh....Searched Dejanews for a workaround and couldn't find > any... I've got it working, and will post a patch here and to apache soon. It seems apache's authors are assuming POSIX wait (waitpid, etc...) functions. -- Rex A. Dieter rdieter@math.unl.edu (NeXT/MIME OK) Computer System Manager http://www.math.unl.edu/~rdieter/ Mathematics and Statistics University of Nebraska-Lincoln
From: rdieter@math.unl.edu (Rex Dieter) Newsgroups: comp.sys.next.programmer Subject: Re: apache 1.3b7 compile errors Date: 28 May 1998 17:22:41 GMT Organization: University of Nebraska-Lincoln Message-ID: <6kk6h1$r8$1@unlnews.unl.edu> References: <6kk1s1$su2$1@unlnews.unl.edu> In article <6kk1s1$su2$1@unlnews.unl.edu> rdieter@math.unl.edu (Rex Dieter) writes: > In article <Pine.NXT.3.96.980527213549.15207A-100000@luomat> Timothy J > Luoma <nospam+yes-this-is-a-valid-address@luomat.peak.org> writes: > > > > http_main.c: In function `process_child_status': http_main.c:3921: > > Ugh ugh ugh....Searched Dejanews for a workaround and couldn't find > > any... > I've got it working, and will post a patch here and to apache soon. It > seems apache's authors are assuming POSIX wait (waitpid, etc...) functions. OK, draft 1, ugly (but working) patch... attached to the end here (Newsgrazer attachment). -- Rex A. Dieter rdieter@math.unl.edu (NeXT/MIME OK) Computer System Manager http://www.math.unl.edu/~rdieter/ Mathematics and Statistics University of Nebraska-Lincoln -- NewsGrazer, a NeXTstep(tm) news reader, posting -- M>UQR=&8P7&%N<VE[7&9O;G1T8FQ<9C!<9FUO9&5R;B!/:&QF<SM]"EQM87)G M;#$R,`I<;6%R9W(Q,C`*7'!A<F1<='@Y-C!<='@Q.3(P7'1X,C@X,%QT>#,X M-#!<='@T.#`P7'1X-3<V,%QT>#8W,C!<='@W-C@P7'1X.#8T,%QT>#DV,#!< M9C!<8C!<:3!<=6QN;VYE7&9S,C!<9F,P7&-F,"!);B!A<G1I8VQE(#PV:VLQ M<S$D<W4R)#%`=6YL;F5W<RYU;FPN961U/B!R9&EE=&5R0&UA=&@N=6YL+F5D M=2`H4F5X($1I971E<BD@=W)I=&5S.EP*/B!);B!A<G1I8VQE(#Q0:6YE+DY8 M5"XS+CDV+CDX,#4R-S(Q,S4T.2XQ-3(P-T$M,3`P,#`P0&QU;VUA=#X@5&EM M;W1H>2!*("!<"CX@3'5O;6$@/&YO<W!A;2MY97,M=&AI<RUI<RUA+79A;&ED M+6%D9')E<W-`;'5O;6%T+G!E86LN;W)G/B!W<FET97,Z7`H^(#X@7`H^(#X@ M:'1T<%]M86EN+F,Z($EN(&9U;F-T:6]N(&!P<F]C97-S7V-H:6QD7W-T871U M<R<Z(&AT='!?;6%I;BYC.C,Y,C$Z7`I<"CX@/B!59V@@=6=H('5G:"XN+BY3 M96%R8VAE9"!$96IA;F5W<R!F;W(@82!W;W)K87)O=6YD(&%N9"!C;W5L9&XG M="!F:6YD7`H^(#X@86YY+BXN7`H@7`H^($DG=F4@9V]T(&ET('=O<FMI;F<L M(&%N9"!W:6QL('!O<W0@82!P871C:"!H97)E(&%N9"!T;R!A<&%C:&4@<V]O M;BX@($ET("!<"CX@<V5E;7,@87!A8VAE)W,@875T:&]R<R!A<F4@87-S=6UI M;F<@4$]325@@=V%I="`H=V%I='!I9"P@971C+BXN*2!F=6YC=&EO;G,N7`I< M"D]++"!D<F%F="`Q+"!U9VQY("AB=70@=V]R:VEN9RD@<&%T8V@N+BX@871T M86-H960@=&\@=&AE(&5N9"!H97)E("A.97=S9W)A>F5R(&%T=&%C:&UE;G0I M+EP*7`HM+5P*4F5X($$N($1I971E<@D)"7)D:65T97)`;6%T:"YU;FPN961U M("A.95A4+TU)344@3TLI(%P*0V]M<'5T97(@4WES=&5M($UA;F%G97()(`EH M='1P.B\O=W=W+FUA=&@N=6YL+F5D=2]^<F1I971E<B]<"DUA=&AE;6%T:6-S M(&%N9"!3=&%T:7-T:6-S("`)"2!<"E5N:79E<G-I='D@;V8@3F5B<F%S:V$M M3&EN8V]L;EP*7`H*>WM<3D=$;V-U;65N=#@Q-R!A<&%C:&5?,2XS8C<N;W!E M;G-T97`T+G!A=&-H"C<W,#@@32=95C`X/B(D)C@N1E,I/#@K1CHH3CXG0B,Y M4%([3CLP*C`N)T%`3B$]8$`B5BQ"44A</B\H)"XJ)R1'4@I-*%A",RU&0S!@ M8"=`*2A`."Q&,E`]+T4R.41..BM65V!(8"0C.3A4.3`F+RQ.(R<C44Q$.#(F M-RPZ)492"DU*/2Y'-"HM*D<T2E5*339+-BLM2E<\2#4J2414.4Q28#HB)4<[ M*591.$PY(F`Q)5`X+"8C(C$]7CXR+$<*35,Y1#<Y+5HL/CXO)D,E5E9@0$0Z M,2I#4"%%6ET_+U8^-#TB0BPV(E=`7B\^1B5/5SM-6ET_/E1&3B4J)@I-(V!` M0CD\.F!&(2=#*3!8/2PF(D!;041#05A4/")%.R@_0E`X.TXG(T8A-$=(*"M& M.RDL4#XT(RA0+R12"DU!0%@Z+$<S*"HM5%HA4%@E,"@B+$$\)D)96$@X*R@R M(D(A65LA8$!).3\B/$9@+$)=75Q9.3Y#-$`P+"(*32$Q4%E:/D0D.5PO5R-@ M3#`Y/2-<,B$D6TU(3#=9254Q(5=00"DL(24[8$147458.3].)S%04$TN020\ M)PI-)%9&*"@T/#@\3C-18$8]0C,E)B8G)SE<(BPJ(2I2,S@P1T)!33LA0$!0 M729"+#XD)35:.%@X.44]+T%!"DU")C8Q.B(I)#@X0R$Q041-32\D)B8W-UQ6 M)"PX/3TB(5%!/4%.+"8W0#!`5E%*+B\P8$DI22DE*"I/1"4*328V*E1@-C8I M-$M+,5A872$C145$)C-61%0]/#LB-U)92%E!-$5'1$5&02HE2#U8/#L^2#A9 M24$U1D9$1`I-/R<_-"TU)CXT.5@H*5E%-D=-1R,G8%Y12#XX.4$H*4DJ*5U2 M*D5),"8F,5Q"*CX[/EPZ.4HI)T%0.20F"DT_)T!*7"4U.#A9-S50141-22TW M(C9&5S4Q24$\/2U'55!!45)#+R<B)"A'.2TV8#E-*U152E8F6%I,24L*33E` M*R)164%#3#!(2R8V-E8E2TU+2UQ&-DPU3#U0+U@V)R100%`G(B,C1"M)-&`M M2RXS0#=65E,M,4<Z.PI-32P\2EU.34Y/."9`2S<C)"@W-3PC,TLL1"PM*UDS M6#=7-4!46",A(SU*-DTD+R-@+$\S)#<X)20B+2<D"DTE5%!$*#0P+RA(0D`P M*"1;6R=@43,R2E18*#HD)T1,2"0T-$I<+3TB8$1=+%0\+"DD-BDA43!$*$E+ M4E(*3214-E@O8#0U,31@(3Q812\P)"4E)$163%<L*#,L(3Q802$E+"$G)B12 M0BHP/"<S,C)8724A+%1.-EQ14`I-)#0Y7#PT-#`S12(A42$Q,B4G)213,U=$ M-#PX/$,B-$`A/3PQ,U)3(B<C73(A)"@M+E@C4$!!04I/1R)!"DTF6%HJ.#\G M*5=.5STM4$%<0#(B1#$B(E0S8#596BTE(4=)*V!4224F)EQ@)C9)7"L\6"Y@ M5T%#1%<Q)28*329',F!$+3\Y04T](21@04=/(R1$0$%')U%')R@K6SM`)"0P M0U!5)5)3(DH_7"PQ(R0T5R1'04544CY/-@I-4T!$+"P\,RHX6%$A/RQ<7R%7 M5R)4+#!?2S,S,S-"+%U//3$T-%8E44(]*$]?/#0T,BLP32TA15E"+44F"DTF M2BI5."@Z,RL^4S5!22%.)"Q<)"0]7B<K4R=24D(L6$]?-SY`4E]>7%XP4R=? M2"9-4DA7.2Q(3R=?+BP*35E>,T!!1#PQ0U!!.%112S=/8#A63#DG+3LF0R-& M+#E05$LE+$,T)58\4R,C6S!(1$Q1+6`\8$!`,BE$,0I--T%@(T8I52HK)C!1 M(E5+-"1"4554.DA235Q+-2)0*E(P4T%<.#!93CXH7CI:)F`F)5A`1C<[.3DR M8%E5"DU1/"Y@7R,B*#TW)",D.%982R$[-"PH1C-*/34]22TU)"=0*3%")2A, M6$%`,E0H,R)3,$Q)+3-%+B-),R,*34<G6CTZ4#4W)RTX,CY=0R,I6"E10D8O M1$,F/T8I."P[/29.+2TI0B%.)3LR)R==*4E0-%=$8"0K0BLN(@I-,28Y)RE` M3D-",2X^8"@H/CTN3%XV,D,B)2)+641(3U5@,#%"-"U6+T9/,B4L1E)0+E9@ M42@H+6`X*#!`"DTL)B0J,3M#.")@6D@I1B)57$DU)3HE+"(X+6`]0BE0*D4^ M3%LH8$)(*"@T2E$A*"1,5C$P*S@].5,E4E8*32HP.ETM(B$X.21'*T-2-DXB M1%,P63!&6&`H+D(I,2XA4"1)(E5;6T`A)E%&+%`D2#`L)#`P-6!2)B]@)PI- M1CTT)%5!-C)@+"1/0BDL*CLZ*B\F43M34$4S*4PE-%0Q+BI02"LU)2<Q.B0C M,%XQ4%%1/CQ@,4Q3-B0N"DTJ.$Y#.2XY6%0H,$Y-*B5'0%Y`-$I"0S%002(T M2$8U8$HJ.#A:12(M*49@*CU&*45#.#$S)D%"3"PS-R<*33HZ+EDY8"TG0#19 M-#9!-#TJ-#@_1D,I6%XP.%A=(DHN3F`P66!91#0D0THG120D+24N1R<E,D$U M1#Q/.0I-(T%@3D5(,$DY*EPG."0C(R92,D(I3F`I45`]6"HX+#TW3B(F*$LD M(2,A/#LP44,P8%@J1C4S-5LH0"(A"DU&0SM&)$$T2%M'.$<B6%4V5E=()RTM M02E+(5XB)S%4)#933#%5(E1<2"Y--CU1*D!54S]053I<,U`Y)4`*32(^3%M0 M7B@G(UA'15DP2#8N6#1;22XD04=#)B,G+5LP(4LZ8#`K,4E<(C0[0$4L)UQ0 M)4LC+T@D)4\C*@I-1$M4-UTJ4"@S0",A)2=31V!-/S$I8"<Y,EE$-4)"8"5" M1C552EDI.R100UA(*T<N0RPT+2@A0R\B55%`"DU54RI8,R="1UA0*"I-0"(C M)4XR8")26TA`0$I81&`K2D](8"LL-T`A)R188$%.5$5@(E!`5B0I.S(Q(SP* M32HP43<B1R5-4#%%(BU8+B1`*F`K3T<U*6!"(T=!0"-#62(S54D]0TP^1#XQ M6SXF)$PV46!#6"0O1R9@,0I-/RY1)4$Y55!@3U,R1U$I12M642PL*EDV3CI) M/%E**R<L45HK6T0J24I30TQ(1B4D5EY73CU<+%M73RI*"DU8*U-)-S9=6S`X M(D!;8#4A(D0M8&!%.%$Q(2)414I"45Q=5B-#6"8L02-9+$5`,2I+040U-$TX M)RHG7E@*32(D4T@P55`O0EU',#H_.4(B-D`]/TXX4$0H3CDB0E!0-2$V2S=@ M-DM0(T,K2%5*/$A924`V6#TL8#)`40I-*T="43,]/D504T932#Q914XG+3]< M3%,Q+SU<4#9'559@65=-1RE,7$1=)BPP)#5#(R9`0$M)/RI=(T14"DU;.D`R M23XU1S!@3EI")D,X0R<O65HL/#A*(C@F5UHJ(30X6UI+43LL2&`X5C@D)3,W M455+6S)33$9<3EP*328Q)R1>,#1#*2@D*4XJ-D0X3B9),R9/2EY`135()"I/ M/U%73#-!*#M'*TI%8$`R12@H-45@5TXR4$A'+0I-(E4P8&!/1F`]239-,S0O M02\I/3@\32]!(E<R2%985ELD7%PD1B%//2(P0U=.(UTG(59(-DY7*U0E030Y M"DTT(6`F)4A"(2$V+2A@,$,L,S@M5$%$)T`O.R%$(E1+4T!66B$P8#]-+C-` M*S0B(4]:*4H[1S)%8$`Q03(*36`G8#9%*2PI,RXC+R)'(4U3,D\T*V!%-$XE M1#!*)2(E*CHB0#-'/#(N)%<T*CPV+$!6(CXQ0T!94TA#4`I-/$-+.$HP65,D M)T,C-#U@/D9@+EU/)S]88$=04BQ",4`B,5-"(E$\65(^-BDL4D14.#!)*2XD M*R\_45E4"DT]8%LQ645.+TXP3C(O0219328D+C]1)5D_6TH]059>*SDY5UHQ M3SPF25!:(D5!/"$X1%U04$$F-%1@,$@*358A3SI$7U(B5SM>-UTR8"Q8-CU9 M+U,N2CA+0DHF.UDJ5%5')"@Z/#5?-CQ;2%@D+CTO(4$F4UM'04-1(PI--TQ: M-%U?,T4L7UXE4SI.*50H5E%(2B1#0U9!/3T]2U)&+3Q<65%474$]."HK4T)# M,C(M230K+D](5CE:"DU=.B$_/2HA)RM?3S-!54HO35$^+B=555HG(4XH2E59 M3CLO-D0A73LT6CQ-.DI!3T-=13\H,RHU3DA2)"T*35DC)2,F-U@N1U],8"<U M0"LO5E=08$`T6D!90B$C*U1=4CM)(21@,$@I+B0K+3@O(B1:,6`A,3,X6#!5 M20I-,&`X(EE4,U`\,#M`-5<Y4V`S15)<8#0D1#`Y3$PF75`E6&!,4"%`)R%` M*U\E6"--2#`J1BE4-%Q(1D%%"DTH)V!4)5-+4$PU.5P\,5PW6B%$2C!@.594 M1EU2*30V.EA7+%$Y6"A@0B@H2EHG+#,B)#PG/4@H.4XH)#,*32@R.#XB*"M+ M+2@I4S0H*#8N*"],-2@L3%XF6D$X5S$A*5@L.$8H*$LN*"0E2$1,)TU$6TPM M424_7"=820I-+"<_444D*2-<8#5!)%`Q,T0G/3]@)"1+05`Q36!`+"E,)E=% M7#=4+R)@*5XQ,5D_5#%'5T@U.4Q<(BU%"DU4,#E@)5-)5E)@.4!)1$PI4"18 M0%0E2"4^*&!"4$4U5U(D6$]()T%)741=3"%'1#\I,E0[7F`K)%U0-U0*33PP M.29=)T8R4B=&-"8G3C9>8#HX+F`Z)5E&.%91)T@E23).,S8H1C<Z)#(O(455 M1S4B7U@P0"\F-#4S7`I-+$4O2#!(23@M)#Y)6BA&2T8D/D,A13%=720]2D$A M2B0E)U8G,6`L5T1@+S<Y4R,I/B=4*RA43DDI)%(A"DTA5BDI)&`Q(F`F43\X M)UTP-&`S5%!0,2],(S`W4%@Z4TTT(CY8-"5,0"(E5BA&6S!--EA-8"<O2#18 M,#@*32HB2D!=.$5'-"=*/%58+2TL0#M@)$@H0"@H,21**"TP.B@L34XH,%`^ M)S`Z3%`K3BHA2%HJ*%X],#A?5@I-+B0P/4HH7TTD.#-#+5@Q3E@H,%0F*#0C M,59?2%U(*3Q-6#5&64DF,BA12#$N*%XW7B0L(CPP/S!9,5H_"DU5.2(W4B19 M4%!@.T]-8#TH+3$X550A2RL^*2\J6B0\*U10/$%2,4`I(CA>/E`Y(418*6!< M)BDG6S(I)$0*33XI)2(^.#`Z3C19+"HG*2%..6!--CDP1SXY,%XJ*2=<6BA= M+"(Y(2@^(5TJ*BDI)SU3)%1",2TD,CA<4`I-4C`Z/$$H,D=5)%XL358X(4$X M3",E-CX[1B,^-4XH)CDB)U%56C`Z/3(G25@J.$I)(5A%5$HT+EM:*3Y3"DU9 M6$U;(2A074@Q,DA<,E4Q63=.5R(A1EE.*3U=-C=2/2U4.EQ<-2\A+#%1(C4P M+S]82495.B1!0S`Q2%<*34PH4D1:.%$S73@I4C@C2D%<4#@J4B@A3C8I,U`H M,4E%,C$N3C8Q8"M6*&`O2B<R05`T6#=%1EE4/3$H,0I-,F`X.CA4-$Y(."Q( M2"@G5T(I,6`J)#0X4C1;5UPW-"<R,2\H2CE0,$8Y-#\N*2<N5CDP2F`I.#PR M,#Q`"DU",2T^3C@H.S$A+38J(D-.,5E2/DDC)S(I4"Q9-"E+-5E96U!"-U1< M0413,4XD2B%=4#$^,%51/R0F,E\*33%`,#<T5U]715=?6U577$<N*&!>-2A` M0TQ8(E=>,B906CA@,%HH*UHM2"1')%A`63HH,RE".3=91BE5.PI--CDG7SHH M,STN.2%"5C@P*#A*)T)".B<V4C@G.%8Y4$):*B4](CHK/B(S2#TF.34R,B@W M/%XY-$(^2#=)"DTN.2\A,C-"3THY3E)6.4Q*2"$]4E`I6B<I(3`B3%A1*#)' M5U1:.2]54C@Y4$PE-4A94B\A+#@K3%0R,%X*34))3"]>2"-73#<P(V!%5%U@ M-5PX/2%?22%572XD0TM,)BDQ*EHT+"U&24Y14CA<,RI**3$F,4U*6D@I-@I- M1D@A.B)*,5`J2"0[*CHL)%XY)416.BU-3DHU(2I!22,N0#E2/"%<2E@B,B@^ M1$I-34%024PX4C`R0"Y("DTJ-$Y>+2140EXX6392*2540#9:,S10.4]$(C`L M)BI(/#8J,4952%A02BI$3EXF6CTJ0T8T)%I)5%`H7$H*32PJ/"A&.DA$3CI* M/$Y((DXB2DDI0$`I02XZ3"4F-S(A.DE?3TU42U-98#$E73<_*U5@,2%`8#4[ M0%<I+PI-*&`Q*B@B7RTB0S$U7",C+TI&4S%*0#$E/"=`.U4V2UHH03)&1C`O M0E))+DE:.C)'+DDU)EXA65I$-28D"DU>1"\A3DH]1"0I5$XR*D]*)DA>0TDX M7"]=.EDL*BI<43Y&65\F1R\C,D$M3"Y@,E$J,TI)(3=<3D`C1U`*344U05,V M*%LX5$96+B4P.D=-."E*7C(F0T4Z04!11E578")14T@E-258045(55`X428Q M44)<(T5*(CI.70I-539(7#@P+2PA,"@F8%`L3C9:1U5:-EE.3BA<03`B2%XI M4#]'44--/TPT7$`X8"A&8&`H05PK)2E$03Q*"DU>,#1>7E)`(C59.$8J2BE@ M7D0O5&`Y4$<X65M5)DI0)UX[,4%2.#`[72LS(U(T1DQ*2DX_(DI,2#A@*3$* M33XT2"4P2R,\04`U3")7)"TF4TA>35`Z1$A474<H(BDJ0$HV14PS15PD-4LA M)%`Z*28J245@.C-0-#8A3@I-2C8Z+E56.C(N.3XR,5E"/"I@/48L8#HS*4$Q M62HY/%=1.T%96%M","I83UTA.4,L,$M%4$U+0DA6*T(O"DTF6T8[5EA+*250 M+U581U)0(4):-UU$.CHZ*T5<0CM&1%)82R@Q5DLB+B);*C`R6T941DM$3DA2 M73I:6RH*33I$53(X8#$C,"M)/U`J43@P8#!+)5`P3T0Q2EPU*5Q#/"E=0BP[ M4R5*6T4_5"-1)4`F4310(4%4,5M7)@I-0EM7.D908%-()4<],EM3/R)3,SE& M5E`I7"%3*#TJ148D*U,V4$,U.40H5"5.1$U+4B95(5A@.R%<,#E/"DTF*C1< M+E=@+DHI.5E(.$-04E1),EQ@,"\T4#4N3&`R,SU@-EE5-3I-33`Y+DA)*#8T M6T(N)BD^*5Y24B,*34(X25X\0SI76$<L)D$B)41<8"DV8$I=/"U;1#E3)CP\ M+R(\*S4R/$$K8#Q#-RPI+#(D-B%.-THQ,BE`/0I-5RPH0$LH*%`S+"DC.$PI M5U!(230D*R=31$-$44@Q4C1%0R)!6CM)6D(\+$Q?+"4M4$PL45\L+%-#*4,I M"DTE6%)&24%2(E$A+2LK)U184C1!*$@X4R0\-DY/(5)00D@H2D!&)CTV8%A7 M)2I!.SE06U=063`Q.%0I734*33`F5EA)-E$C0"PC5B%64S4A1BDE*"@Y)30D M1C8I*$Y<56`K,2I1.3]<8#1?64E<6#Q7054_)ST_*"Q")@I-.%`L6$Q@.V!5 M(5E2,RP]5B\D65E7)S\L54P].%U,/E`I+$8D*R<U0"@T44TP+$8V7R93*T%@ M,3I@8"A)"DTX.D0Q-4`X)5LL2RU25$]-0C`M*BY!4%0N4"A!*38G5B8W2CE@ M0"A<3"=$+5XC)TI!.%E-1"8\7C@J1S(*35TK6B9;*ULQ/2))5E!65SDP7",P M5E@T2E%7*3,\*U,M)D,D5%12.E%3)BT]-#Q;1C\L6TA'+%M*3RQ;3`I-5RQ; M3E\L63HH,%E(,&`C4#<L7U8_+%]81RQ?6D\L7UQ7+%]>7RQ<8"<]8"(O/6`D M-SU@)C\](UX[(C(W"DU$0D0J)3)%3S!"1C]`0CI!+$))1T10(BQ:)%XM(54B M+SE<+#LO(E<H7$).3SI24S1<0DY#1BPJ)SU(2#X*35LR6U`\,E<Y)$9$4E@R M5T`H1$-#3$PR7UE;1CM)1"]-6$),2T<Y13<W(E\F,#`O6$A"73T_*%$I0D0X M)@I-1%HQ*D\C8"E4(V`O*%-8-6!6.4)=0TI?,C5$1#A3*$A$4TXJ3%-1."LV M-%=$4R\K6%,M8"A3,21@5CL]"DTP4U5%0S-5,CE#-2LG-"4[5%,U/R13.4-` M4#5%/%,Y24Q3.2Y8+3%0-%`Y4C!3/59`4SY15E(Q75Q3/"D*320C0F`W,B8D M8%56)R132BI04T-)7#-`)50V5C\N1R)#331>.#10+RU1,T5=)D-*0S!165H^ M(TI)3%-(10I-5$-(6"HG7C---B)5/%-,7$133ELP4TY>2%=915PQ3"0D,U<G M/%`K*5Q=-RTH73LQ4%-=1DL]34@W,S0H"DTT33M57%-<320D(4$\(UM)3%-9 M3U0C6D0A,U]2,%-?7CQ37E1',UQ%3$-=,%1<5R\O,R1@-2-;)2\T(B8*33<T M(U%))&`X5$0C6UTD(4PO-"5/.S0E4D,T)3`W.5E5+3U024Q%-$)=)%DW3S,H M1T4D*3-,)"A-/2TE3PI-/CM"(B([2519)#$_)20U6"4M*U];-#4P+20Q0C4M M*E)(3"=>5S0U5BDN(4(Y*TE'42%3*3`P+4M=*4%."DTU+2Q&,S)06EDN8"%# M3B)47EA..TXW4F`I+2\E3BM.)UDD0BQ9)$,\)$1$3D`V5C8Q(51`1"1&.SDD M1S`*32%@*C\E)$M<+E=93C0D2D<M)RLR4E1*3B$D3"4_)$Y0221/1"I<56`G M1%)+,213)C4L(2Y;1%`I8"140PI-*5Q63#LD5#])5%<\6215/3XD7T0Y)%E- M1$1;2U4D6T\Q)S]3-21?5T4D7U5(/5XF2ETF*$Y>(C]65THT"DTF)2@V,34H M.24U)2U+(DQ9234A5E-%+#LU/"Q6.SA(0"DU*5Q/12Q&034J03A0,$LE-3<L M(54N4C(M+C$*32(^/%=+/CU42T4L154U+%Q3-#Q>+S4S63A%,R8N(3DO*3I% M*U$P*3%-,R%/*38U."TV*RE9-3128"4Z.0I-+$4Y13TX/4@U,5E+.3PX1#PN M23546"A=-D543D!5/BPQ,2(N*3`I75TU/C(M/39?-C5!/$5%1BM!-C,B"DTC M149@*3`^,ST^4C5%/E$^548D0SPE*CXA-4I'+35*13$[)DA)-4I,635+120U M3DY',T$R.45/.T8E3EL*33$Q)ETM/4A'8#51,BE%4R0I-BLH534V+CDV*RY% M-BLR438K-C$P-$XA1B\\63570"DU6T0Y-5M(235;3`I-635;4"DU72TT55]7 M135?6RTQ2UXA-B-.-T)@)"U&)%@T-B9&-S5$.EU,3E0W+R0Y)UY=+"A>5"U) M1B0\"DU-140^*48T0#%.7$1913=80T8K,$T^7EQ;3EY07S8K.B4_8%,Q3V!6 M048L*$]6+%TI1C8K*R8Q0#`F,4@*33@F,E]/13TM)%8T3B1?,2XM334V*4\I M-4E/,R1!)C4Z(4`K-$M%0CM)-"0W/%,Y15%&+#9'+28O4BTG7@I--SPZ/UE' M15E?0"E$5U9!2$@N(4\]32LA-5=&(4E/.#(G1D<L,#9$53@F1C`W12XA7T\Z M/BE//E4B6%)%"DT]1DI)349*/UM&2E`Y24HG.U9.23(W/EI11DY>/497(EU+ M4BPK5E<Z.E97+3U.+C,O5E<U74]#+2E`+SH*35%&5SY504=`540C4S5&6S]* M-EHK+B<K34E(,U$Q3T@V,59<)3LF6UI11E]>(4=@(C%78"9!5V`J45=@+@I- M(5<G7%%')#197THY0S<D.E%7)4-8)R=&*2<U/3HG*3A`)RL[5R<I03U7,%@D M."A605<L6E%85S$G5S!`"DTL)S$E)5I1*%%932PM7%$O+5$X44@O4U<]1RTU M/5<U.45?5$E8)S15,5<S3"]7.TA=0B5'15<X/3U#)D$*32M44R,J-SPE+R=# M/DHG/5Q17ET\+C9@,3Y@."]50T0B04`_42Y@)%8]7"16-DHB6S@Z62=&8%0_ M,U@R-PI-72Q3/D934R))4RA=22Q:4UU`(SU!2"Y@.B1<(CI,8"H^)2E,,U$F M.V`[2R<N*T(T3R0_0%(M63U&0U<G"DTQ7#!))U8F1#I)6R\T(BU?0#0Q*5DT M(T`I-UPP)4\_620]6"=?*5(V7#HL/E),7$`X0D]).5LY*$U/.C,*34$^(2]= M144D72P_7E@A7$Q()U]80U]>.3`P0BHM449`(3@V1R<I,CXF1%@T2D`I-2E0 M5U5`6S$S0E5)*@HJ4U)$0$PR(DA9+U5$)"%<3`I@"GT*K'U<<&%R9%QT>#DV M,%QT>#$Y,C!<='@R.#@P7'1X,S@T,%QT>#0X,#!<='@U-S8P7'1X-C<R,%QT M>#<V.#!<='@X-C0P7'1X.38P,%QF,%QB,%QI,%QU;&YO;F5<9G,R,%QF8S!< )8V8P(%P*"GT* `
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 28 May 1998 14:52:11 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6kjtmr$9da$2@ns3.vrx.net> References: <Pine.NXT.3.96.980527164001.19969A-100000@pathos> <1998052723270000.TAA11534@ladder01.news.aol.com> <6kjc23$q9p$6@ns3.vrx.net> <rex-2805981501210001@192.168.0.3> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: rex@smallandmighty.com In <rex-2805981501210001@192.168.0.3> Eric King claimed: > Actually, the Quicktime image decompression manager does this. Sure, and QT is a perfect example of something far and away better than what's available on other platforms. QT is not a "framework" in the sense of AppKit though - you wouldn't use it to build an app for instance. The specific argument here is that YB is _bad_ because it can't do _this_. However not only can YB do this in the AppKit, but in any number of other ways too. Even if you do it the hard way I'll bet it's still less work. Maury
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 28 May 1998 14:53:44 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6kjtpo$9da$3@ns3.vrx.net> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: rex@smallandmighty.com In <rex-2805981519470001@192.168.0.3> Eric King claimed: > Go visit a Barnes & Noble and see if you can find a copy of Inside > Taligent Technology. It contains an overview of CommonPoint's design. Also > the latest versions of IBM's OpenClass framework for OS/2 and AIX are > basically a 'best of Taligent.' I have to admit I'm a bit confused about the whole thing. Can you outline the current frameworks? What does VisualAge use? > I've seen sample code from their OpenClass frameworks, and overall it > doesn't look that bad. More complex than the BeOS, but a hell of a lot > more functionality also. Well.... >Personally, I think any system that tries to > incorporate that level of functionality is going to run into serious > problems. Even the Yellow Box. So far so good. Maury
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 28 May 1998 15:16:40 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6kjv4o$9da$4@ns3.vrx.net> References: <6kjelm$d08$3@news.idiom.com> <1998052819012200.PAA09263@ladder03.news.aol.com> <6kkeuj$8p2$1@crib.bevc.blacksburg.va.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: nurban@crib.bevc.blacksburg.va.us In <6kkeuj$8p2$1@crib.bevc.blacksburg.va.us> Nathan Urban claimed: > In article <1998052819012200.PAA09263@ladder03.news.aol.com>, larrysb@aol.com (LarrySB) wrote: > > > Maury Markowitz and Nathan Urban have offered incomplete and over-simplistic > > solutions. These two also seem more bent on conducting gihad than on offering > > any concrete or substantive discussion. I will stop responding to either of > > them at this point. > > How mature. Actually, I don't think he replied to any of my posts in the first place. From his description of the solution under MacOS, I still think the solution offered up to use the ImageRep as a storage for pointer access is essentially identical to his use of GWorlds, although it of course gets to leverage the YB's file manipulation system - even a wrapper if you're so inclined. John Randolph's solution is indeed interesting, but nothing at all like the problem space. A solution using a GWorld has to go something like this... a) find the file b) read the file and figure out what's inside c) build a GWorld of the required size and depth d) setup your pointer math based on the geometry e) read the file into the GWorld, converting as you go f) do pixel manipulation based on pointer refs into the GWorld g) repeat while on (g) h) at some point map the GWorld either to the screen - or - h) decompose the GWorld back into another format Here's the YB solution that most closely matches it... a) find the file b) let the imageRep class read it in for you c) setup your pointers based on information returned to you from the imageRep d) do work on the pixels (in raw C if you wish) based off of bitmapData e) repeat while on (d) f) send that to an NSImage, which takes care of the rest So not only does YB not interfere with the process in any way, it does indeed make it easier. Of course he has been offered a solution like this on no less that three occasions (now four) but has failed to even mention any of them. Instead he is very quick to mention other people's examples wherever they say "it's not that easy". In addition repeated calls for specific examples of why the solution above does not solve the problem (as he descrived it at least) have gone unanswered, and in addition the other examples he gave were sophmoric - "direct access to hardware" and "color manipulation". In both cases YB has some form of solution that does _something_ at least, when the normal MacOS has none in that space. Maury
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re-creating DPS. Date: 28 May 1998 21:52:04 GMT Organization: Idiom Communications Message-ID: <6kkma4$p4o$2@news.idiom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Guys, Let's consider for a moment the possiblity of re-creating a DPS interpreter, given whatever Apple's likely to ship in Mac OS X: First, the hardest part of writing postscript is not the interpreter, its the raster operations, font cache machinery, etc. Given NSScanner, NSDictionary, NSArray, and a PDF back-end to do the actual rendering, how much work would this really be? Let's further assume that the first cut wouldn't take binary object sequences. Comments? -jcr
Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <dPye7vuaJS1o@cc.usu.edu> From: root@127.0.0.1 Date: 28 May 98 14:35:09 MDT References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kcc12$hb1$3@news.idiom.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit In <6kcc12$hb1$3@news.idiom.com> John C. Randolph wrote: > > Keep in mind, that the Apple you know to be an unreliable business partner > has been taken over by NeXT. > Get me the Pepto-Bismol... Come to think of it, if you're relying on the old NeXT to be a reliable partner, you'd better just get me a couple of cyanide tabs and I can get it over with quickly. *Thunk*
Subject: Re: future of mac programming From: "Baskaran Subramaniam" <baskar@RemoveMe.snet.net> Newsgroups: comp.sys.next.programmer References: <6keipe$35e$1@crib.bevc.blacksburg.va.us> MIME-Version: 1.0 Message-ID: <B19366A7-26AA48@192.168.128.1> Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Date: Thu, 28 May 1998 23:13:52 GMT NNTP-Posting-Date: Thu, 28 May 1998 19:13:52 EDT On Tue, May 26, 1998 10:15 AM, Nathan Urban <mailto:nurban@crib.bevc.blacksburg.va.us> wrote: >In article <1998052616533300.MAA26118@ladder01.news.aol.com>, larrysb@aol.com (LarrySB) >wrote: [ stuff deleted ] >> YB may be the greatest thing since sliced bread. But until it's wrapped up >> and on the shelves and the end-user's computers it has no value or relevance. > >Wait two years to adopt it and that's just two more years of lost time >for you. Well, I am no expert on YB. I have spent quite a few hours learning it. I think it is a very nice framework. That is not the same thing as saying that YB is the best tool for every developer or every application. Apple supporting Carbon API in MacOS 10 (Sorry guys, I don't care for Roman numerals, and writing MacOS 10 as MacOS X is plain misleading at best.) is not an admission that Carbon API is any superior to the YB API. It is purely a pragmatic decision to keep the current MacOS developers from abandoning Apple. Without their support Apple WILL not survive. Sure it is easy for Nathan to tell the current MacOS developers to forget about their cash cows (currently shipping MacOS software) and start rewriting their code using the YB API and in the process abandon the current MacOS market. But that is complete nonsense. Why would anyone whose economic survival depends on this market do that? Do you follow? When MacOS 10 ships, it won't run on anything other than Apple Powermac G3 or newer machines. What percentage of the installed base of Macs do you think that is going to be when MacOS 10 ships? So by rewriting your apps to YB you are abandoning most of the Mac users. What are you getting in return? An entry into the Windows market provided that you had not already entered this market by writing the same software in Win32/MFC. Most commercial software companies already have a Windows version in addition to a Mac version of their software. So the supposed benefit of using YB is really not there. In summary, if I already have a Mac and a Windows version of my software shipping and making money for me, I don't care about rewriting to YB since by doing that I am abandoning most the installed base of Mac users. On the other hand, by staying carbon compliant I can sell to most of the Mac customers who cannot run MacOS 10. No company that has a large investment in Mac Toolbox code is going to be in a hurry to rewrite it in YB. For them, having Carbon API support on YB/Windows will make more sense. Baskaran
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 28 May 1998 22:53:17 GMT Organization: Idiom Communications Message-ID: <6kkpst$p4o$4@news.idiom.com> References: <6kjelm$d08$3@news.idiom.com> <1998052819012200.PAA09263@ladder03.news.aol.com> <6kkeuj$8p2$1@crib.bevc.blacksburg.va.us> <6kjv4o$9da$4@ns3.vrx.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: maury@remove_this.istar.ca Maury Markowitz may or may not have said: [snip] -> From his description of the solution under MacOS, I still think the -> solution offered up to use the ImageRep as a storage for pointer access is -> essentially identical to his use of GWorlds, although it of course gets to -> leverage the YB's file manipulation system - even a wrapper if you're so -> inclined. John Randolph's solution is indeed interesting, but nothing at all -> like the problem space. Correction: It's not *my* solution, I learned about it from Mike Barthelemy, and he learned about it from someone at TRW. The point of it is, it's a lot of work to use NSImages as the raw storage for major image-bashing problems, just like it's a mistake to use a single NSAttributedString to store a novel. These classes are the right thing to use in *most* cases, and they can be extended for use in specific cases. Now, If I were writing a medical imaging app, I'd still use the Yellow Box classes to deal with everything in my application that was common to other apps. I'd still use the Interface builder to set up my tool pallettes, menus, etc. I don't expect NSImage to include a -convolveWithKernal: (float**) kernal or a - (NSArray) histogram method. I also don't expect it to behave gracefully when I hand it a 200 megabyte image to munge. That's where *my* bag of tricks has to come in. -jcr
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 28 May 1998 15:46:27 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6kkeuj$8p2$1@crib.bevc.blacksburg.va.us> References: <6kjelm$d08$3@news.idiom.com> <1998052819012200.PAA09263@ladder03.news.aol.com> In article <1998052819012200.PAA09263@ladder03.news.aol.com>, larrysb@aol.com (LarrySB) wrote: > Maury Markowitz and Nathan Urban have offered incomplete and over-simplistic > solutions. These two also seem more bent on conducting gihad than on offering > any concrete or substantive discussion. I will stop responding to either of > them at this point. How mature.
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 28 May 1998 22:00:10 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6kl4ra$9c0$1@crib.bevc.blacksburg.va.us> References: <6kks1f$95j$1@crib.bevc.blacksburg.va.us> <B19382E0-2D4CAC@192.168.128.1> In article <B19382E0-2D4CAC@192.168.128.1>, "Baskaran Subramaniam" <baskar@RemoveMe.snet.net> wrote: > On Thu, May 28, 1998 7:29 PM, Nathan Urban > <mailto:nurban@crib.bevc.blacksburg.va.us> wrote: > >In article <B19366A7-26AA48@192.168.128.1>, "Baskaran Subramaniam" <baskar@RemoveMe.snet.net> wrote: > >> Sure it is easy for Nathan to tell the current MacOS developers to > >> forget about their cash cows (currently shipping MacOS software) and start > >> rewriting their code using the YB API and in the process abandon the > >> current MacOS market. But that is complete nonsense. > >You're right. Can we say "strawman"? I knew we could. > I am not sure, what you mean here. Sorry :-( A "strawman" argument is one where someone attacks a misrepresentation of a person's position rather than their actual position. > Well, as a long time Mac developer, I don't like the idea of abandoning > the Mac installed base and start developing for Windows. Applications > written using YB API will run on all Intel machines than can run > Windows, but only on a puny fraction of the Mac hardware. It's only partly a question of the relative importance of targeting Windows vs. Mac markets. It's also a question of: what fraction of the _buyers of your software_ (as opposed to the total Mac installed base) will be running MacOS X by the time you ship your product, and how will this ratio change with time over the expected lifespan of the product? What is the likely adoption rate of OS X (and Rhapsody, I suppose), and of the non-migrating users, how many would buy your product (or have hardware capable of running it adequately)? As others have pointed out, those who are unlikely to upgrade to OS X are also those who are unlikely to buy much in the way of new software. On the other hand, adopting does take time, so your choice depends on when you intend to release your product, how many of your customers are expected to adopt, how quickly they do so, whether you have other products to sustain you, etc. Personally, I would expect the adoption rate of OS X to be higher than that of 8.x, assuming the major apps get ported to Carbon, purely for the sake of major performance and stability gains. For some developers, the benefits of YB development and deployment will outweigh the disadvantages of putting "MacOS X, Windows 95/NT4 or higher" on the box.
From: myst <myst@hkstar.com> Newsgroups: comp.sys.mac.programmer.misc,comp.sys.mac.programmer.tools,comp.sys.mac.programmers.misc,comp.sys.newton.programmer,comp.sys.next.programmer,comp.sys.psion.programmer,comp.unix.programmer,comp.unix.sco.programmer,comp.windows.ms.programmer,de.comp.os.ms-windows.programmer,de.comp.os.os2.programmer,dk.edb.programmering,fj.os.ms-windows.programming Subject: Help Me PLease !!! Date: Fri, 29 May 1998 10:56:27 +0800 Organization: Netvigator Message-ID: <356E23DB.1F04832B@hkstar.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Dear Sir, I'm a HK student, i'm very like to made program... I have a big problem... How to captrue DirectX Game Screen !? May you teach me ? Thankyou very Much !!! From Myst Ho.(myst@hkstar.com)
Subject: Re: future of mac programming From: "Baskaran Subramaniam" <baskar@RemoveMe.snet.net> Newsgroups: comp.sys.next.programmer References: <6kks1f$95j$1@crib.bevc.blacksburg.va.us> MIME-Version: 1.0 Message-ID: <B19382E0-2D4CAC@192.168.128.1> Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Date: Fri, 29 May 1998 01:14:17 GMT NNTP-Posting-Date: Thu, 28 May 1998 21:14:17 EDT On Thu, May 28, 1998 7:29 PM, Nathan Urban <mailto:nurban@crib.bevc.blacksburg.va.us> wrote: >In article <B19366A7-26AA48@192.168.128.1>, "Baskaran Subramaniam" < >baskar@RemoveMe.snet.net> wrote: >> Sure it is easy for Nathan to tell the current MacOS developers to forget >> about their cash cows (currently shipping MacOS software) and start >> rewriting their code using the YB API and in the process abandon the >> current MacOS market. But that is complete nonsense. >You're right. Can we say "strawman"? I knew we could. I am not sure, what you mean here. Sorry :-( >I never suggested anything of the sort. The issue in this thread has >always been: which is better for _new_ development? For someone starting >from scratch with no MacOS code to leverage, YB is more compelling. Well, as a long time Mac developer, I don't like the idea of abandoning the Mac installed base and start developing for Windows. Applications written using YB API will run on all Intel machines than can run Windows, but only on a puny fraction of the Mac hardware. To even suggest that I do this, Apple (oops... I mean NeXT) must be smoking something. Baskaran
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 28 May 1998 19:29:51 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6kks1f$95j$1@crib.bevc.blacksburg.va.us> References: <6keipe$35e$1@crib.bevc.blacksburg.va.us> <B19366A7-26AA48@192.168.128.1> In article <B19366A7-26AA48@192.168.128.1>, "Baskaran Subramaniam" <baskar@RemoveMe.snet.net> wrote: > Sure it is easy for Nathan to tell the current MacOS developers to forget > about their cash cows (currently shipping MacOS software) and start > rewriting their code using the YB API and in the process abandon the > current MacOS market. But that is complete nonsense. You're right. Can we say "strawman"? I knew we could. I never suggested anything of the sort. The issue in this thread has always been: which is better for _new_ development? For someone starting from scratch with no MacOS code to leverage, YB is more compelling.
From: rex@smallandmighty.com (Eric King) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Thu, 28 May 1998 19:24:04 -0400 Organization: The Small & Mighty Group Message-ID: <rex-2805981924060001@192.168.0.3> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> In article <6kjtpo$9da$3@ns3.vrx.net>, maury@remove_this.istar.ca (Maury Markowitz) wrote: :In <rex-2805981519470001@192.168.0.3> Eric King claimed: :> Go visit a Barnes & Noble and see if you can find a copy of Inside :> Taligent Technology. It contains an overview of CommonPoint's design. Also :> the latest versions of IBM's OpenClass framework for OS/2 and AIX are :> basically a 'best of Taligent.' : : I have to admit I'm a bit confused about the whole thing. Well, I'm by no means an expert, but my understanding is that when IBM absorbed Taligent, one of their main goals was to incorporate technology from CommonPoint into their VisualAge C++ and Java products. They've been doing that gradually over the past couple of years.I have an OpenClass preview CD, and from the docs on it, it looks like they have moved quite a bit of code over. : Can you outline the current frameworks? What does VisualAge use? OpenClass is the name of the frameworks included with VisualAge. It is currently being infused with Taligent technology. :> I've seen sample code from their OpenClass frameworks, and overall it :> doesn't look that bad. More complex than the BeOS, but a hell of a lot :> more functionality also. : : Well.... Well, you truly underestimate just how much functionality that was put into those frameworks. If you want something added to the YellowBox, chances are you'll find something similar already in the Taligent frameworks. I'm not saying the Taligent stuff would work as well, but chances are they had already thought to add it. :>Personally, I think any system that tries to :> incorporate that level of functionality is going to run into serious :> problems. Even the Yellow Box. : : So far so good. Exactly, my point. So far. The YellowBox works very nicely as is, but it doesn't cover anywhere near as much ground. There were frameworks for all sorts of things in CommonPoint. Probably too many frameworks. Not too many in the sense that it was hopelessly confusing for developers, but too many in that it was probably *extremely* difficult to implement. I doubt they ever had time to think about optimization. Wonky management probably didn't help matters out either. -Eric
Subject: Re: future of mac programming From: "Baskaran Subramaniam" <baskar@RemoveMe.snet.net> Newsgroups: comp.sys.next.programmer References: <6kgu3m$nc0$10@ns3.vrx.net> MIME-Version: 1.0 Message-ID: <B193753E-2A1857@192.168.128.1> Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Date: Fri, 29 May 1998 00:16:07 GMT NNTP-Posting-Date: Thu, 28 May 1998 20:16:07 EDT On Wed, May 27, 1998 7:40 AM, Maury Markowitz <mailto:maury@remove_this.istar.ca> wrote: >> This is no good if you want to target MacOS 8.x derivative or MacOS 7.5.x. >> If you stick with Carbon > ... this is no good if you wish to target Win32. Hmmm, Win32 vs. MacOS... I guess you don't realize that most commercial software houses already ship software for both of these platforms. In order for them to continue to do so even after MacOS 10 ships, they will have to write their Mac code using Carbon and not YB. Because YB is not going to run on most of the installed base of the Mac hardware out there. On the Windows side, nothing changes, write to Win32 API or use MFC. Even for new projects, writing to the YB API prevents you from being a player in the Mac software arena. The percentage of Macs that can run your Apps are going to be even smaller than the current 4% mac market share :-( >> What we really need is support for Carbon on the Windows version of the >> Yellow Box. > *coff* Ummm, you don't do Carbon "in" or "on" anything - it talks to the >kernel directly. What you suggest is likely impossible (or close enough) >MacOS -> YB -> Win32, and seems to stem from a potential misunderstanding of >how the current solution is to be implemented. Well, those are your inferences...based on your limited understanding of what is out there. Actually expanding the number of Powermac models MacOS 10 can run would be a much better option. If that is not done, then expand the work Apple has already done in bringing QTML to Win32 and support the entire Carbon API right on top of Win32. >Maury Baskaran
Subject: Re: future of mac programming From: "Baskaran Subramaniam" <baskar@RemoveMe.snet.net> Newsgroups: comp.sys.next.programmer References: <1998052807155500.DAA29116@ladder03.news.aol.com> MIME-Version: 1.0 Message-ID: <B1937ACF-2B6758@192.168.128.1> Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Date: Fri, 29 May 1998 00:39:52 GMT NNTP-Posting-Date: Thu, 28 May 1998 20:39:52 EDT On Thu, May 28, 1998 3:15 AM, LarrySB <mailto:larrysb@aol.com> wrote: >A realistic perspective: >1. Must code in Obj-C or possibly Java. > a. requires learning a new language with its own little peccadillos. The >basics are easy, the intricate little details require a long time as with any >other language. > b. Some techniques, such as MI, are eliminated and this may pose problems >for developers who favor them. > c. Obj-C offers some constructs that may foster better coding styles than >C or C++. > d. Obj-C hasn't had the wide exposure of C++, which is used by thousands >of developers around the world, it hasn't had the benefit of widespread use. > e. Java is still a very immature programming language. >2. There is no "api" per-se. Indeed, the whole thing is a set of frameworks. > a. Programming effort may be reduced in many cases. > b. Unlike familiar class libraries (MFC, PowerPlant, etc) source is not >available. > c. Documentation is lacking in many areas. > d. The framework may not offer methods or data unique to your situation, >and potentially require difficult work-arounds, complicated by the lack of >source and documentation. >3. Deployment of completed applications at risk. > a. very limited number of OpenStep installations in the field. > b. Unknown royalty conditions for Win32 runtime. > c. No MacOS platform deployment for a minimum of a year (OS-X). > d. Rhapsody on Apple RISC platforms is not likely to reach beyond a tiny >fraction of the already small Mac marketshare. > e. A very real possibility that YB will never ship, should a shift in Apple >management or priorities occur. >Now, if you are not bothered by anything in the list, then go for it! But the >rest of us may find these conditions unacceptable, at least today. That is a very good summary of the benefits and costs involved in moving development to YB APIs. These are real issues that must be resolved by Apple. These issues make it very difficult to convince project managers to invest resources in starting development using YB APIs. However great Obj-C and YB may be, these issues come in the way of us adopting them. Remember, the current OpenStep developer community needs no convincing here. Also, they have nothing to lose, except may be their Intel hardware :-) [I wonder how they are going to feel when they have to pay Billy boy big bucks to continue to use their favorite development tools on their Intel hardware]. Baskaran. [Hi Larry! We met a while ago... remember?]
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 28 May 1998 22:06:11 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6kl56j$9d8$1@crib.bevc.blacksburg.va.us> References: <1998052807155500.DAA29116@ladder03.news.aol.com> <B1937ACF-2B6758@192.168.128.1> In article <B1937ACF-2B6758@192.168.128.1>, "Baskaran Subramaniam" <baskar@RemoveMe.snet.net> wrote: > On Thu, May 28, 1998 3:15 AM, LarrySB <mailto:larrysb@aol.com> wrote: > > A realistic perspective: > That is a very good summary of the benefits and costs involved in moving > development to YB APIs. These are real issues that must be resolved > by Apple. Some of his points were quite valid, but others were either misconceptions or misrepresentations -- some absurdly so -- as others have pointed out. ("Documentation is lacking in many areas."? "There is no `api' per-se"??? Give me a break!) > These issues make it very difficult to convince project > managers to invest resources in starting development using YB APIs. I know I could convince _my_ management to do it, if we ever started a new project (which we probably won't, being a one-product shop). OTOH, we're Windows developers, so the issues are different.
Subject: Re: future of mac programming From: "Baskaran Subramaniam" <baskar@RemoveMe.snet.net> Newsgroups: comp.sys.next.programmer References: <6kl4ra$9c0$1@crib.bevc.blacksburg.va.us> MIME-Version: 1.0 Message-ID: <B1939271-30F58E@192.168.128.1> Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Date: Fri, 29 May 1998 02:20:40 GMT NNTP-Posting-Date: Thu, 28 May 1998 22:20:40 EDT On Thu, May 28, 1998 10:00 PM, Nathan Urban <mailto:nurban@crib.bevc.blacksburg.va.us> wrote: > Personally, I would expect the adoption rate of >OS X to be higher than that of 8.x, assuming the major apps get ported >to Carbon, purely for the sake of major performance and stability gains. What? Ported to Carbon? You didn't mean ported to YB, didn't you? Adapting existing Mac Toolbox code to Carbon is not what I would call porting. Anyway, performance and stability of OS 10 are good and I am sure everyone will agree with that. But the catch is, OS 10 won't run on anything prior to the PowerMac G3 series and that makes the installed base of OS 10 capable hardware insignificantly small. So, writing in Carbon is the only way to reach the bulk of the Mac installed base, unless OS 10 ends up supporting more Powermac models. Baskaran
From: nurban@crib.bevc.blacksburg.va.us (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 28 May 1998 22:48:29 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6kl7lt$9hj$1@crib.bevc.blacksburg.va.us> References: <6kl4ra$9c0$1@crib.bevc.blacksburg.va.us> <B1939271-30F58E@192.168.128.1> In article <B1939271-30F58E@192.168.128.1>, "Baskaran Subramaniam" <baskar@RemoveMe.snet.net> wrote: > On Thu, May 28, 1998 10:00 PM, Nathan Urban > <mailto:nurban@crib.bevc.blacksburg.va.us> wrote: > >Personally, I would expect the adoption rate of > >OS X to be higher than that of 8.x, assuming the major apps get ported > >to Carbon, purely for the sake of major performance and stability gains. > What? Ported to Carbon? You didn't mean ported to YB, didn't you? No, I meant ported to Carbon. The availability of Carbon apps is what will determine the near-term adoption of MacOS X. (I can't say about the long-term factors.) > Adapting existing Mac Toolbox code to Carbon is not what I > would call porting. Whatever -- a matter of semantics. Say, rather, "assuming the major apps get adapted and recompiled for Carbon". Which I think is quite likely. > Anyway, performance and stability of OS 10 are good and I am sure > everyone will agree with that. But the catch is, OS 10 won't > run on anything prior to the PowerMac G3 series Is that certain? I haven't been paying attention to that, but I remember there was some controversy over whether that was true. > which makes the installed base of OS 10 capable hardware insignificantly > small. But as I mentioned before, and you apparently ignored completely, the size of the installed base is not the important factor -- it is the number of software consumers. And the buyers of new software tend to correlate strongly with the owners of newer systems, and the fraction of OS X-capable buyers will only increase.
From: phy070@spo109 (H.-R. Oberhage) Newsgroups: comp.sys.next.programmer Subject: Re: Re-creating DPS. Date: 29 May 1998 08:18:16 GMT Organization: Universitaet Essen GH, Germany Message-ID: <6klr08$epa1@mx2.hrz.uni-essen.de> References: <6kkma4$p4o$2@news.idiom.com> John C. Randolph (jcr.remove@this.phrase.idiom.com) wrote: : Guys, : : Let's consider for a moment the possiblity of re-creating a DPS interpreter, : given whatever Apple's likely to ship in Mac OS X: : : First, the hardest part of writing postscript is not the interpreter, its the : raster operations, font cache machinery, etc. : : Given NSScanner, NSDictionary, NSArray, and a PDF back-end to do the actual : rendering, how much work would this really be? Let's further assume that the : first cut wouldn't take binary object sequences. : : Comments? Interesting idea ..., one question is if the rendered PDF is 'internally' available or has to go over the file-system or worse display. And, pixel rendering is, of course, resolution dependant, which resolutions will be available to us? In other words: can we 'pipe' meaningful intermediate results? Greetings, Ruediger Oberhage -- H.-R. Oberhage Mail: Univ.-GH Essen E-Mail: phy070@sp2.power.Uni-Essen.DE Fachbereich 7 (Physik) ruediger@Theo-Phys.Uni-Essen.DE S05 V07 E88 Universitaetsstrasse 5 Phone: (+49) 201 / 183-2493 D-45117 Essen, Germany FAX: (+49) 201 / 183-2120
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer Subject: Re: Re-creating DPS. Date: 29 May 1998 11:40:13 GMT Organization: Technical University of Berlin, Germany Message-ID: <6km6qt$i6f$1@news.cs.tu-berlin.de> References: <6kkma4$p4o$2@news.idiom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit jcr.remove@this.phrase.idiom.com (John C. Randolph) writes: >Guys, >Let's consider for a moment the possiblity of re-creating a DPS interpreter, >given whatever Apple's likely to ship in Mac OS X: >First, the hardest part of writing postscript is not the interpreter, its the >raster operations, font cache machinery, etc. >Given NSScanner, NSDictionary, NSArray, and a PDF back-end to do the actual >rendering, how much work would this really be? Let's further assume that the >first cut wouldn't take binary object sequences. You wouldn't want to uses the base classes because (a) they don't have adequate performance, and more importantly (b) they aren't semantically compatible with the comparable PS-objects. For example, you need to track some attribute flags for each object, there need to be references to objs (but with their own flags) and sub-arrays/sub-strings. In addition, objects need to obey save/restore, which would be very difficult to do with the Foundation classes, and it also seems difficult to optimize object allocation. At least my little interpreter (started sunday, spare time only) is now using custom objects and has been executing procedures since wednesday. Next up is the save/restore logic and optimized name lookup, both of which aren't trivial but I *think* I have reasonable solutions. After that, the scanner (not too hard, even with BOS and token support) and then primarily grunt work implementing the operators and some more datatypes as well as hooking everything up to the YB graphic routines. The implementation as Objective-C objects means that it will be possible to have full language integration like Obj-Tcl or the Java/Objective-C bridge. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 29 May 1998 10:24:22 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6km2cm$s3m$4@ns3.vrx.net> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: rex@smallandmighty.com In <rex-2805981924060001@192.168.0.3> Eric King claimed: > OpenClass is the name of the frameworks included with VisualAge. It is > currently being infused with Taligent technology. Ok, that makes much more sense now. > Well, you truly underestimate just how much functionality that was put > into those frameworks. If you want something added to the YellowBox, > chances are you'll find something similar already in the Taligent > frameworks. No, I was poking fun at Be. From what I've looked at (old BTW) the Be libs were in many was similar to PP or TCL. Solid, but rather "thin". I heard all the stories about thin being good, then I used YB. I have no doubt that VA is similar to YB in most ways in terms of the frameworks, but I do find it difficult to believe they have some of the "gimme" classes in YB, like NSColor. I'll bet for "information processing" it's likely very similar though. So specifically though, does IBM's Java stuff also have a OpenClass framework set? > Exactly, my point. So far. The YellowBox works very nicely as is, but > it doesn't cover anywhere near as much ground. As Taligent? > There were frameworks for > all sorts of things in CommonPoint. Examples? Did they have anything like EO that was as good? Maury
From: rpk@lotatg.lotus.com (Robert P. Krajewski) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Fri, 29 May 1998 11:24:23 -0400 Organization: Lotus Development Corporation Message-ID: <rpk-2905981124240001@iriri.lotus.com> References: <6kl4ra$9c0$1@crib.bevc.blacksburg.va.us> <B1939271-30F58E@192.168.128.1> <6kl7lt$9hj$1@crib.bevc.blacksburg.va.us> ~lr4R6:/]e<sbpfH<oCq8[1|GuQigM!M1#b.ln@O- In article <6kl7lt$9hj$1@crib.bevc.blacksburg.va.us>, nurban@vt.edu wrote: >But the catch is, OS 10 won't >> run on anything prior to the PowerMac G3 series > >Is that certain? I haven't been paying attention to that, but I >remember there was some controversy over whether that was true. Well, it looks like OS 10 will only work on G3s -- although I really hope "optimized for" and "required" are getting confused by folks at Apple. But that's another story. Anyway, there will be a Carbon library for MacOS back to 8.x (8.1 or 8.5, I dunno). So you can still deliver Carbon apps to non-G3 users. Remember that older MacOSes will be maintained, just as they were in the coexistence-with-Rhapsody scenario.
From: "kjc" <casac@inetarena.com> Newsgroups: comp.sys.next.programmer Subject: Re: Best development environment, a nitpick Date: Fri, 29 May 1998 09:27:13 -0700 Organization: Internet Arena Message-ID: <6kmnle$c3o$1@play.inetarena.com> References: <6kfcm3$gq$1@nnrp1.dejanews.com> <6khbfj$loc2@onews.collins.rockwell.com> Erik M. Buck wrote in message <6khbfj$loc2@onews.collins.rockwell.com>... >In <6kfcm3$gq$1@nnrp1.dejanews.com> quinlan@intergate.bc.ca wrote: >> Just a comment for those of you who are claiming the YB is the best >> development environment on the market: you should really take a look at a >> good smalltalk implementation and see if you still feel that way. I'm not >> saying the YB is bad, just that working in a smalltalk environment is >better. >> >I am a big fan of Smalltalk also. I think you will find that many >Objective-C users are. > >> Why? My top three reasons are: >> >> 1) All types are objects. >> 2) No primitive source/include file model >> 3) Code modifications can be made and take effect while the code is running >> > >1) This is possible in Objective-C if you are disciplined and are willing to >give up some performance and have appropriate frameworks available. With the >introduction of NSString etc. in NeXT Objective-C, the use of non-scalar >non-object types is rare. > >2) Actually, many people think the "artificial" separation of interface from >implementation is a GOOD thing. The "primitive" include model is a file >system issue. It is really not all that primitive and class browsers etc. >are available. > >3) This would be nice to have > >There are however some very large problems that have traditionally plagued >the "superior" smalltalk environments. The biggest issue has always been >supporting more than one developer per project. If you have 20 programmers >on a team, they will be stepping all over each other in smalltalk. As a Smalltalker I must strongly disagree with you. IBM VisualAge for Smalltalk has the best team development environment I have ever experienced. And this comes right out of the box. The environment also supports version control and software ICs (import/export-able objects) that can be loaded at runtime. If your interested, the team environment is called ENVY. The other major Smalltalk vendor, Object Share offers the same team development environment for their VisualWorks Smalltalk product line. For more information you can go to: http://www.software.ibm.com/ad/smalltalk for Visual Age and: http://www.objectshare.com The >"shoot from the hip" tradition in smalltalk can be a disaster in some cases. >How can anyone figure out where and when the new bug was introduced ? >Distribution of smalltalk apps has been a problem too. Is there a runtime Yes I will agree that deployment CAN BE rather involved. But to date there are a number of very good tools that facilitate the effort. >available ? Do I have to distribute source ? The program just produced an Nope you do not have to distribute the source. >error, how do I continue ? Exceptions >
From: "A. J. LaSalle" <alasalle@ctron.com> Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Fri, 29 May 1998 12:52:16 -0400 Organization: Cabletron Systems Inc. Message-ID: <356EE7C0.2F1CF0FB@ctron.com> References: <6kgu3m$nc0$10@ns3.vrx.net> <B193753E-2A1857@192.168.128.1> <6km2m1$s3m$5@ns3.vrx.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Maury Markowitz wrote: > > In <B193753E-2A1857@192.168.128.1> "Baskaran Subramaniam" claimed: SNIP > > Apple has already done in bringing QTML to Win32 and support the entire > > Carbon API right on top of Win32. > > Why? So you can use those outdated Mac API's to port your app, rather than > the considerably fresher YB ones? That's lossage for you! > > Maury But for the MacOS developer, moving to Carbon is fairly painless (or so I've heard), moving to YB isn't. So if Apple supports Carbon on Wintel, MacOS developers get relatively "painless" access to a huge market, and they don't have to learn anything new (well, new marketing and support issues!). Who cares if the the API is "outdated" if you've got a million lines of it to sell! I think Apple would be better off, and more successful, encouraging current MacOS developers with Carbon/Wintel rather than trying to enlarge the developer pool with YB. I think that YB will fail to interest Wintel developers. If you want to be on windows, just learn MFC and the Win APIs (or VB, which a looks like a nasty language, but has a huge following). Why waste your time learning YB so that you can pick up X thousands of Mac sales? Heck, with all the hype, if you really want cross platform (including unix), you'd probably be better off learning Java. Besides, learning MFC or VB makes a programmer instantly marketable to thousands of software companies and IT users, not so with YB. Apple needs to think again. The strength of its platform is the billions of lines of code written to the Mac API. They need to give developers reasons to stay with that API - period. AJ LaSalle --
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 29 May 1998 10:29:21 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6km2m1$s3m$5@ns3.vrx.net> References: <6kgu3m$nc0$10@ns3.vrx.net> <B193753E-2A1857@192.168.128.1> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: baskar@RemoveMe.snet.net In <B193753E-2A1857@192.168.128.1> "Baskaran Subramaniam" claimed: > I guess you don't realize that most commercial software houses > already ship software for both of these platforms. Not at all true, most software houses ship software only for PC's, and this is true for the vast majority of vertical app shops - like mine for instance. You refer to most _Mac_ shops. However if those people target YB, they get (say) 10% of the Mac market, and some 50% of the Win market. If they target Carbon they get some (say) 50% of the Mac market. I know which one I'd choose! > Even for new projects, writing to the YB API prevents you from > being a player in the Mac software arena. _currently_ yes, but that isn't stopping us. Any lossage there is easily made up by on the Win side. > that can run your Apps are going to be even smaller than the current > 4% mac market share :-( But the percentage of PC's that can run your app is MANY MANY MANY times larger than the percentage of Macs that can't. > Well, those are your inferences...based on your limited understanding of > what is out there. My limited understanding as a professional YB developer and 8 year Mac developer? Please! > Actually expanding the number of Powermac models MacOS 10 can run > would be a much better option. For everyone, yes. > Apple has already done in bringing QTML to Win32 and support the entire > Carbon API right on top of Win32. Why? So you can use those outdated Mac API's to port your app, rather than the considerably fresher YB ones? That's lossage for you! Maury
From: pete@ohm.york.ac.uk (-bat.) Newsgroups: comp.sys.next.programmer Subject: Unpacking raw data froma Tiff Date: 29 May 1998 15:33:14 GMT Organization: The University of York, UK Sender: pcf1@york.ac.uk Message-ID: <6kmkfq$j03$1@pump1.york.ac.uk> Hmmm, I go away for a bit, come back to find a raging thread on one of my favourite subjects : accessing raw data in a BitmapImageRep. This is actually a remarkably useful thing to do and it's annoying that NeXT provide no method of getting the RGB pixel value at a point. Last year I needed to unpack the data from a Tiff which originated on the screen to use it as a texture for a surface under OpenGL. So I wrote a bit of code which will take a BitmpImageRep (NeXTStep or OpenStep) and unpack it into a block of memory as either 8 bit RGB or 8 bit greyscale. It works on almost all the Tiffs I have tried it on, with the exception of one obscure Tiff I found that seems to use an indexed palette (any ideas?) For those interested grab makekong.ohm.york.ac.uk:pub/pete/unpacktiff.tar.gz Yeah, it's not a very memory efficient way of doing things, but it does the job, which is the important thing in this case, and handles all the differing formats and depths. Meanwhile the advent of map data in GEOTiff format leaves me still longing for a getPixelColour method to get data from an arbitrary geographical location. -bat.
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 29 May 1998 10:43:35 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6km3gn$s3m$6@ns3.vrx.net> References: <6kjelm$d08$3@news.idiom.com> <1998052819012200.PAA09263@ladder03.news.aol.com> <6kkeuj$8p2$1@crib.bevc.blacksburg.va.us> <6kjv4o$9da$4@ns3.vrx.net> <6kkpst$p4o$4@news.idiom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: jcr.remove@this.phrase.idiom.com In <6kkpst$p4o$4@news.idiom.com> John C. Randolph claimed: > Correction: It's not *my* solution, I learned about it from Mike Barthelemy, > and he learned about it from someone at TRW. My appologies. > The point of it is, it's a lot of work to use NSImages as the raw storage for > major image-bashing problems Oh, for sure. But the solution in use on the Mac side is rather similar - a GWorld is being used as a pointer-accessable pixel store. THAT you can do with the NSData or NSBitmapImageRep classes (not NSImage) just as easily, staying inside the framework at all times, and leveraging YB's data manipulation and filtering at the same time. > NSAttributedString to store a novel. These classes are the right thing to > use in *most* cases, and they can be extended for use in specific cases. Sure, but any problem set covered by a GWorld is certainly covered by the far more studly NSBitmapImageRep and other related classes. The solution you outlined was something radically different than the problem required. If the image fits into a GWorld, it fits into a NSBMIR. Your solution is likely needed for large images and effecient memory use, but by using a GWorld you're admitting you don't need this anyway. > I don't expect NSImage to include a -convolveWithKernal: (float**) kernal or > a - (NSArray) histogram method. I also don't expect it to behave gracefully > when I hand it a 200 megabyte image to munge. That's where *my* bag of > tricks has to come in. Absolutely, but again that has not been demonstrated as needed in _this_ case. If you have a 200Mb image to munge, the GWorld ain't gonna do it either! The original complaint was that you can't get access to pixel data easily from NSImageRep's, this is thus a demonstration of why YB makes things harder. However you CAN get access to pixel data easily from a NSImageRep, and work on it with the same tools. Thus the complaint (unless the issue was misrepresented) is false. Then he stated that because it was hard to do, YB is thus inflexible. In that he ignored the fact that you can do EXACTLY the same thing simply by using raw C code and a malloc. YB in NO WAY ties your hands. Finally he claimed this is an example of the poor documentation. I pulled up PB, went to the AppKit reference for NSBitmapImageRep, and under the "Instance Methods" title (indexed earlier in the document as well) this description... bitmapData - (unsigned char *)bitmapData Returns a pointer to the bitmap data. If the data is planar, returns a pointer to the first plane. ... in addition I found... getBitmapDataPlanes: - (void)getBitmapDataPlanes:(unsigned char **)data Provides access to bitmap data for the image separated into planes. data should be an array of five character pointers. If the bitmap data is in planar configuration, each pointer will be initialized to point to one of the data planes. If there are less than five planes, the remaining pointers will be set to NULL. If the bitmap data is in meshed configuration, only the first pointer will be initialized; the others will be NULL. Color components in planar configuration are arranged in the expected orderÐf or example, red before green before blue for RGB color. All color planes precede the coverage plane. Which does even more of the work for you. Conclusion - if you are using a GWorld to do bitmap manipulation, then a NSBitmapImageRep will make it easier to do. Goes without saying - this solution may not work in all cases, but it DOES work if you have a GWorld doing it now. Maury
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: Re: Unpacking raw data froma Tiff Message-ID: <1998052918174800.OAA15769@ladder01.news.aol.com> Date: 29 May 1998 18:17:48 GMT Organization: AOL http://www.aol.com References: <6kmkfq$j03$1@pump1.york.ac.uk> pete@ohm.york.ac.uk (-bat.): <<<< Hmmm, I go away for a bit, come back to find a raging thread on one of my favourite subjects : accessing raw data in a BitmapImageRep. This is actually a remarkably useful thing to do and it's annoying that NeXT provide no method of getting the RGB pixel value at a point. >>>>> Ahh, Pete, this is just some esoteric operation that 90% of programmers don't need to do. <g> Just kidding. I'll grab your package and have a look at it. I'm very curious indeed.
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <1998052918104300.OAA15192@ladder01.news.aol.com> Date: 29 May 1998 18:10:43 GMT Organization: AOL http://www.aol.com References: <6kkpst$p4o$4@news.idiom.com> jcr.remove@this.phrase.idiom.com (John C. Randolph) wrote: :Now, If I were writing a medical imaging app, I'd still use the Yellow Box :classes to deal with everything in my application that was common to other :apps. I'd still use the Interface builder to set up my tool pallettes, :menus, etc. Indeed a perfectly sensible way to go. Just like using Powerplant, TCL, MacApp, MFC or whatever to handle the nuts-n-bolts stuff in an application based on those. :I don't expect NSImage to include a -convolveWithKernal: (float**) kernal or :a - (NSArray) histogram method. I also don't expect it to behave gracefully :when I hand it a 200 megabyte image to munge. That's where *my* bag of :tricks has to come in. I don't expect that either. However, without one of these: 1. Extensive experience. 2. Extensive and well prepared documentation (vendor or third-party). 3. Access to the source, at least as a last resort. A programmer new to the YB (negates item 1) has to consider things like this as part of the learning curve and price of admission. I can walk into my local Bookstop and pickup probably 3-4 worthwhile books on image processing in MFC/Win32. I can get 1-2 good ones on image processing on the MacOS. There is precious little in the way of third-party literature on OpenStep. On Win32, I can purchase a fairly cheap library of IP routines from a number of vendors. I can even get such items or build them into OCX's for a VB (it sucks, but again, it's perfectly suited for many tasks) application. Unlike the other C++ application frameworks on the market, I can't look at the source code when in doubt and say,"Hey, this isn't at all suitable for what I want to do." Or, even better, "Maybe I can take a little of this and build something I can use out of it." If you talk to any programmer who's delivered major app's using PP, TCL, MacApp, MFC, etc, I think you'll find a fellow who has done both. So, if you were starting an all new project, any kind of responsible management would have to consider the above. Add that to fact that there you'll have to license and distribute the YB runtime to target the Win32 market, and only a small fraction of MacOS users may be able to run it, and an even smaller minority of OpenStep OS installed base, and there are some pretty considerable doubts. Not the least of which is we don't know exactly when the YB is going to actually ship for Apple hardware, and we also *don't know* what the licensing issues are going to be for the Win32 runtime, and there goes your project into the unacceptable business-risk category.
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 29 May 1998 20:55:46 GMT Organization: Digital Fix Development Message-ID: <6kn7ci$q5r$1@news.digifix.com> References: <6kgu3m$nc0$10@ns3.vrx.net> <B193753E-2A1857@192.168.128.1> <6km2m1$s3m$5@ns3.vrx.net> <356EE7C0.2F1CF0FB@ctron.com> In-Reply-To: <356EE7C0.2F1CF0FB@ctron.com> On 05/29/98, "A. J. LaSalle" wrote: >Maury Markowitz wrote: >> >> In <B193753E-2A1857@192.168.128.1> "Baskaran Subramaniam" claimed: > >SNIP > >> > Apple has already done in bringing QTML to Win32 and support the entire >> > Carbon API right on top of Win32. >> >> Why? So you can use those outdated Mac API's to port your app, rather than >> the considerably fresher YB ones? That's lossage for you! >> >> Maury > >But for the MacOS developer, moving to Carbon is fairly painless (or so >I've heard), moving to YB isn't. So if Apple supports Carbon on Wintel, >MacOS developers get relatively "painless" access to a huge market, and >they don't have to learn anything new (well, new marketing and support >issues!). Apple will NEVER DO THIS. There is no benefit to them to do this at all. It only dilutes the Mac only software base when it comes to existing software. Yellow Box works as an attractant for new development because it removes the need to choose an either MS or Mac business decision (which the Mac will often loose on, or be delayed). -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: embuck@palmer.cca.rockwell.com (Erik M. Buck) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 29 May 1998 21:20:13 GMT Organization: Rockwell International Message-ID: <6kn8qd$loc3@onews.collins.rockwell.com> References: <6kkpst$p4o$4@news.idiom.com> <1998052918104300.OAA15192@ladder01.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: larrysb@aol.com, buck.erik@mcleod.net In <1998052918104300.OAA15192@ladder01.news.aol.com> LarrySB wrote: > However, without one of these: > > 1. Extensive experience. > 2. Extensive and well prepared documentation (vendor or third-party). > 3. Access to the source, at least as a last resort. > > A programmer new to the YB (negates item 1) has to consider things like this as > part of the learning curve and price of admission. > Of course... This is true of any new library or framework. > I can walk into my local Bookstop and pickup probably 3-4 worthwhile books on > image processing in MFC/Win32. I can get 1-2 good ones on image processing on > the MacOS. There is precious little in the way of third-party literature on > OpenStep. > Image processing books add value to Openstep also since they are usually targerted at C or C++. There are 4-5 good Openstep books also although some are out of print or somewhat obsolete. I am sure you can expect more books soon. > On Win32, I can purchase a fairly cheap library of IP routines from a number of > vendors. I can even get such items or build them into OCX's for a VB (it > sucks, but again, it's perfectly suited for many tasks) application. > You can get "Objectware" for Openstep also. Soon, you will be able to use those Active-X controls in Interface Builder. > Unlike the other C++ application frameworks on the market, I can't look at the > source code when in doubt and say,"Hey, this isn't at all suitable for what I > want to do." Or, even better, "Maybe I can take a little of this and build > something I can use out of it." If you talk to any programmer who's delivered > major app's using PP, TCL, MacApp, MFC, etc, I think you'll find a fellow who > has done both. > If you must have source code you probably don't want to use Yellow Box. > So, if you were starting an all new project, any kind of responsible management > would have to consider the above. > Of course! > Add that to fact that there you'll have to license and distribute the YB > runtime to target the Win32 market, and only a small fraction of MacOS users > may be able to run it, and an even smaller minority of OpenStep OS installed > base, and there are some pretty considerable doubts. Not the least of which is > we don't know exactly when the YB is going to actually ship for Apple hardware, > and we also *don't know* what the licensing issues are going to be for the > Win32 runtime, and there goes your project into the unacceptable business-risk > category. > On the other hand, if there is no reasonable chance of the application every being built in another environment, Yellow Box is the lowest risk (but still risky) way to go. You might ask, "What applications would only be built in Yellow Box ?" Well, Any large commercial application being built by 3 or fewer people pretty much needs Yellow Box if it is ever going to ship. I guess you just need to experience the difference if power and productivity between Yellow Box and MFC or Yellow Box and Power Plant in order to understand. In the past, I have reported my personal experiences. One company I contracted with built a complex graphical application in Windows 95, X/Motif, and Openstep simultaneously. The Windows team was largely composed of "experts" and had ~50 people. The X/Motif team had all the newest greatest tools and consisted of ~10 people. The Openstep team consisted of me and 5 other people (none of whom had ever done any Objective-C work before - Three fresh out of college) The Openstep team delivered a full featured application exceeding specifications in 18 months with ~40,000 lines of code. The Windows team delivered a less featured application in 30 Months with ~200,000 lines of code. The X/Motif team never delivered anything after 30 Months and ~100,000 lines of code. Even on Yellow Box for Windows, the Openstep application performs most operations 1.5 to 2x faster than the "native" Windows code. Very little code was shared between the projects largely beacuse the Openstep version was done so soon and the other teams were prevented from "stealing" code for contract reasons. (The company had not paid us for our work yet). By the way, the Windows users to whom the application was deployed almost universally prefered the Openstep version. THis is a MAJOR commercial application. Look at the Lotus teams experience with Improv for NeXTstep vs. Improv for Windows etc. At WWDC, a company the makes an Openstep newspaper layout application was sited as having forced the competition into abandoning its own Windows application reselling the Openstep application. This goes on and on. If you don't beleive it, what can I say. It works for us.
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: Re-creating DPS. Date: 28 May 98 21:01:27 Organization: Is a sign of weakness Message-ID: <SCOTT.98May28210127@slave.doubleu.com> References: <6kkma4$p4o$2@news.idiom.com> In-reply-to: jcr.remove@this.phrase.idiom.com's message of 28 May 1998 21:52:04 GMT In article <6kkma4$p4o$2@news.idiom.com>, jcr.remove@this.phrase.idiom.com (John C. Randolph) writes: Let's consider for a moment the possiblity of re-creating a DPS interpreter, given whatever Apple's likely to ship in Mac OS X: First, the hardest part of writing postscript is not the interpreter, its the raster operations, font cache machinery, etc. Given NSScanner, NSDictionary, NSArray, and a PDF back-end to do the actual rendering, how much work would this really be? Let's further assume that the first cut wouldn't take binary object sequences. Well, if they're using PDF as the basic drawing mechanism, that's easy. I'm not sure what the point of NSScanner would be in this case (I mean, we're talking about an RPN language, take maybe an hour to write a parser for it :-). I mean, what does pswrap do, other than parse your PS code and wrote a token encoded form of it into a C procedure... The real question is where this would fit in the grand scheme of things. If they really use a DPS-like underpinning to the window system, be it PDF, bravo, PGML, or something homegrown, the step to a PS mode is almost so trivial as to make me wonder what's up if it _isn't_ shipped with the OS. Barring legal concerns, of course, in which case they're probably everyone's legal concerns. That said, DPS.framework would be an interesting hack, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 29 May 1998 23:11:08 GMT Organization: Idiom Communications Message-ID: <6knfac$jkv$4@news.idiom.com> References: <6kgu3m$nc0$10@ns3.vrx.net> <B193753E-2A1857@192.168.128.1> <6km2m1$s3m$5@ns3.vrx.net> <356EE7C0.2F1CF0FB@ctron.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: alasalle@ctron.com "A. J. LaSalle" may or may not have said: -[snp] -> I think Apple would be better off, and more successful, encouraging -> current MacOS developers with Carbon/Wintel rather than trying to enlarge -> the developer pool with YB. Nah. The old MacOS toobox is merely *better* than the Win32 API. It's not *vastly* better, it doesn't offer great gains in productivity. ->I think that YB will fail to interest Wintel developers. Probably. If they're willing to put up with windows, then they're rather sadly deprived for a start. However, there are those who will look for decent tools, no matter what appaling excuse for a deployment platform they're forced to use, and that's where NeXT made their sales of OpenStep for WIndoze. -> If you want to be on windows, just learn MFC and the Win APIs (or VB, -> which a looks like a nasty language, but has a huge following). Why waste -> your time learning YB so that you can pick up X thousands of Mac sales? That's not why you learn it. You learn YB to gain the productivity of YB. If you haven't experienced this, let me tell you: it's shocking. -> Heck, with all the hype, if you really want cross platform (including -> unix), you'd probably be better off learning Java. Except that Java doesn't deliver on its cross-platform promises. Does the phrase "Which Java?" ring a bell? -> Besides, learning MFC or VB makes a programmer instantly marketable to -> thousands of software companies and IT users, not so with YB. Yeah, and by that token, just learning to type makes someone instantly valuable to thousands of temp agencies. There are only a few companies using YB, but consulting rates are typically over $100/hr. If you're just coding MFC like everyone else, you're lucky to find better than $50K/yr. -> Apple needs to think again. The strength of its platform is the billions -> of lines of code written to the Mac API. They need to give developers -> reasons to stay with that API - period. That sounds like an argument against innovation, and I can't agree with you at all. Move the same argument to 1984, and you get: "The strength of it's platform is the megabytes of 6502 machine code that runs on the Apple II. Apple needs to give developers a reason to stay with that API - period." Apple got into its present trouble by resting on their laurels. Apple's survival depends on innovation. MicroSquish isn't going to innovate, they're only going to copy Apple and NetScape (and Intuit and anyone else who tops $100mil in revenues.) -jcr
From: "Lawson English" <english@primenet.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: 29 May 1998 16:49:02 -0700 Organization: Primenet Services for the Internet Message-ID: <B19497A2-B5EEA@206.165.43.148> References: <SCOTT.98May28082746@slave.doubleu.com> nntp://news.primenet.com/comp.sys.mac.programmer.codewarrior, nntp://news.primenet.com/comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Scott Hess <scott@doubleu.com> said: >I don't understand, then, what the "I don't think so" is relative to. >Obviously, the backing for the Carbon APIs is going to be an improved >BlueBox, unless Apple were to reimplement the entire API set for >Carbon (unlikely in the _extreme_). > >["Improved", here, means "modified to allow for reentrancy and other >goodies promised by Apple". I take "BlueBox" to mean "Mac OS code >adapted to run over the new kernel".] Nope. The Carbon API lives along-side the YB and can be used to produce "native" MacOS X (Rahasody 2.0) applications that are just as buzz-word-compliant as YB apps are. It is a way of extending the life of Top 100 MacOS Applications without requiring a YB rewrite. Carbon API MacOS apps can run unmodified (recompiled?) under MacOS 8.5. Non-Carbon API MacOS apps can run unmodified on Rhaposdy 1.0 and MacOS X within the Blue Box MacOS emulator. ---------------------------------------------------------------------- Want Apple to license Cyberdog for third-party development? Go to: <http://www.pcsincnet.com/petition.html> ----------------------------------------------------------------------
From: arti@lava.DOTnet (Art Isbell - remove "DOT") Newsgroups: comp.sys.next.programmer Subject: Re: Best development environment, a nitpick Date: 31 May 1998 04:48:31 GMT Organization: LavaNet, Inc. Distribution: world Message-ID: <6kqnev$c6v@mochi.lava.net> References: <6kfcm3$gq$1@nnrp1.dejanews.com> <6kgbuc$8s6$1@nz12.rz.uni-karlsruhe.de> <SCOTT.98May28083942@slave.doubleu.com> scott@doubleu.com (Scott Hess) wrote: > Everyone who has ever written a device driver in any language, please > raise your hands. OK, now, everyone who writes programs (any > language) but hasn't written a device driver, please raise _your_ > hand. And everyone who writes programs (any language) but hasn't written a device driver, has never wanted to write a device driver, has no plans to ever write a device driver, and who would not consider having never written a device driver tantamount to a tragically incomplete life, please raise _your_ hand. > Hmm, interesting distribution, there. It seems that someone could > write a language which _couldn't_ be used for device drivers, yet > still manage to target 95% of everything anyone ever does in computer > languages. Amazing, that, Amen! -- Art Isbell NeXT/MIME Mail: arti@lavaDOTnet IDX Systems Corporation Voice/Fax: +1 808 394 0511 (for whom I don't speak) Voice Mail: +1 808 394 0495 Healthcare Info Technology US Mail: Honolulu, HI 96825-2638
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: You don't measure a language/framework by the number of books on it.. Re: future of mac programming Date: 31 May 1998 04:41:49 GMT Organization: Digital Fix Development Message-ID: <6kqn2d$3v9$1@news.digifix.com> References: <6kn8qd$loc3@onews.collins.rockwell.com> <1998053003404000.XAA17471@ladder01.news.aol.com> In-Reply-To: <1998053003404000.XAA17471@ladder01.news.aol.com> On 05/29/98, LarrySB wrote: >Four or five good books, even if they were all up-to-date, pale in comparison >to the hundreds of books devoted to MFC and solving domain specific code >problems with it. This is such a bone-headed metric that I feel compelled to comment on it. The market for C++/Java books is huge because every dimwit wannabe programmer will by them. More importantly, any publisher will persue these books when trying to find stuff to publish. If you approach a publisher with a C++ or Java manuscript, they know that they can sell copies of this to every dimwit wannabe programmer, and that generates more books.. Its true that there are few books on Objective-C, but the fact is that the few that they are are EXCELLENT, and cover the subject quite well. Your basic image processing capabilities can be done on OpenStep using C++ code if thats your bent. The language issues are not a major issue. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: arti@lava.DOTnet (Art Isbell - remove "DOT") Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 31 May 1998 05:27:31 GMT Organization: LavaNet, Inc. Distribution: world Message-ID: <6kqpo3$dvt@mochi.lava.net> References: <6kkpst$p4o$4@news.idiom.com> <1998052918104300.OAA15192@ladder01.news.aol.com> larrysb@aol.com (LarrySB) wrote: > On Win32, I can purchase a fairly cheap library of IP routines from a number of > vendors. And you'll likely get what you pay for. I can even get such items or build them into OCX's for a VB (it > sucks, but again, it's perfectly suited for many tasks) application. Maybe this would allow you to get an app out the door fast, but with VB, do you really have an app that is easily maintainable and enhanceable, or is it mostly a throwaway and start over situation? > Unlike the other C++ application frameworks on the market, I can't look at the > source code when in doubt and say,"Hey, this isn't at all suitable for what I > want to do." Or, even better, "Maybe I can take a little of this and build > something I can use out of it." If you talk to any programmer who's delivered > major app's using PP, TCL, MacApp, MFC, etc, I think you'll find a fellow who > has done both. The advantage of source availability is mostly overblown. The amount of time spent figuring out what the source is really doing and the temptation to then modify it without really understanding what else might break just doesn't make this development regimen that we consider robust enough for commercial development. > So, if you were starting an all new project, any kind of responsible management > would have to consider the above. We considered it and quickly discounted it in favor of adopting a development environment that someone might have been fired for adopting :-) but which promised to give us a significant advantage over our competitors who were all slogging it out with Win32 development environments, none of which gave them an advantage over one another. They developed with armies of programmers and produced bland applications of limited functionality. We built our app with a tiny number of highly skilled people and produced an app that potential customers universally found vastly superior to the competition. Unfortunately, that was back before OPENSTEP/Windows, so customers had to buy an operating system as well, NEXTSTEP, which they resisted. The net result - the market was frozen because customers didn't want to buy the inferior Windows apps after they learned about what we offered, but their IS departments for the most part refused to support an unknown operating system. Of course, those IS departments who weren't conservative and who decided to take a risk have been rewarded with an app suite that has recovered hundreds of thousands of dollars for their hospitals. Then along came OPENSTEP/Windows removing the objection, and our company was purchased by a large established healthcare software vendor *because of our app* which could now be deployed under Windows. > Add that to fact that there you'll have to license and distribute the YB > runtime to target the Win32 market, and only a small fraction of MacOS users > may be able to run it, and an even smaller minority of OpenStep OS installed > base, and there are some pretty considerable doubts. Not the least of which is > we don't know exactly when the YB is going to actually ship for Apple hardware, > and we also *don't know* what the licensing issues are going to be for the > Win32 runtime, and there goes your project into the unacceptable business-risk > category. Sure, sure, we've all heard those arguments time and time again. You'll sleep much better at night taking the conservative approach like most do, and the rest of us will stand a good chance of eating your lunch by adopting technology that will allow us to produce a superior product more quickly, with fewer programmers, and, thus, for less money. Sure, we could fail as well if the sky falls as you are wringing your sweaty hands about, but that's a risk some of us really enjoy taking. We'll have a lot more fun in our development environment, YB, while those who take the conservative approach will view programming as little more than just a job. I don't think we should waste our time trying to convince those who sincerely believe that YB development is a serious mistake. Those of us who have been listening to similar arguments for almost 10 years now know what YB can offer. Those either curious enough or not so risk-averse will discover YB in due time and most will be thrilled that they did. But YB isn't for everyone, so let's just get on with building our apps our way and see how they compare with apps built their way. I certainly have no fear of such a comparison. -- Art Isbell NeXT/MIME Mail: arti@lavaDOTnet IDX Systems Corporation Voice/Fax: +1 808 394 0511 (for whom I don't speak) Voice Mail: +1 808 394 0495 Healthcare Info Technology US Mail: Honolulu, HI 96825-2638
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <8873895982424@digifix.com> Date: 31 May 1998 03:49:46 GMT Organization: Digital Fix Development Message-ID: <3654896587219@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1993. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: ppo@fisk.com Newsgroups: comp.sys.next.programmer Subject: testing password post 13139 Date: Sunday, 31 May 1998 04:22:18 -0600 Organization: eh254 Message-ID: <31059804.2218@fisk.com> The password details for http://www.firewalluk.com/secure/systemdoor.htm ps make sure your follow the log on procedures carefully or you will get in a loop. If user1 is logged on try user2 or user3 etc with same password. Name : user1 Password : kitten `k3
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <31059804.2218@fisk.com> Control: cancel <31059804.2218@fisk.com> Date: 31 May 1998 14:55:31 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.31059804.2218@fisk.com> Sender: ppo@fisk.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: "Todd Heberlein" <todd@NetSQ.com> Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Mixing Carbon and OpenStep calls?? Date: 31 May 1998 18:42:54 GMT Organization: Net Squared, Inc. Message-ID: <01bd8cc3$c8ccfab0$04387880@test1> I seem to recall hearing/reading that applications can use both Carbon and OpenStep APIs. Questions: o Is this correct? (if not, ignore the rest) o Since an application can use both "carbon" and "yellow" system calls, should the "box" architecture description go away? o Can this be used as a bridge to get MacOS developers to use OpenStep? (That is, new additions or enhancements to existing Carbonized applications could be done in OpenStep; thus allowing MacOS developers to leverage the power of OpenStep while preserving their existing work.) Todd
From: "Todd Heberlein" <todd@NetSQ.com> Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: How mature is Carbon? Date: 31 May 1998 18:52:59 GMT Organization: Net Squared, Inc. Message-ID: <01bd8cc5$31a42990$04387880@test1> On one side, Carbon was just barely announced and Apple has not set in stone the APIs in Carbon. On the other side, Carbon appeared mature enough to port a rich application like Photoshop 5 (in two weeks!) and demonstrate it in front of a live audience. So, how mature is Carbon? Was that really Carbon (with Photoshop) being demonstrated on Rhapsody at WWDC 98 or was there some smoke ane mirrors? Any clue as to when the first developers release of Carbon will be shipped? That is, how soon can the top 100 MacOS developers move from Carbon dating their applications to actually testing their applications on Carbon? Just curious, Todd
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <1998053118593100.OAA15861@ladder01.news.aol.com> Date: 31 May 1998 18:59:31 GMT Organization: AOL http://www.aol.com References: <6kp9sv$cdl$1@crib.corepower.com> Can anyone name 3 sku's of shrink-wrap software, on the shelves and available at your local Fry's, CompUSA, Computer City, MicroCenter, etc, which was written in OpenStep? You guys can sit here and derride obviously clunky Win32/MFC all you want. You can even make it scary with statements like "eat your lunch". Woopety doo. So far, no one has bothered to "eat my lunch." If it were so good, I would have thought someone would have bothered by now. Either in the petro-engineering software I used to write or the medical image processing I work in now. I can't think of a single title written in OpenStep that was deployed commercially on Windows, at least in a shink-wrap form. Even Nathan, a clearly vociferous supporter of AppKit can't convince his own employer of it.
From: "Michelle L. Buck" <buck.erik@mcleod.net> Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Mixing Carbon and OpenStep calls?? Date: Sun, 31 May 1998 14:14:17 -0500 Organization: McleodUSA - http://www.mcleodusa.net Message-ID: <6ksa9p$ils1@odie.mcleod.net> References: <01bd8cc3$c8ccfab0$04387880@test1> Todd Heberlein wrote in message <01bd8cc3$c8ccfab0$04387880@test1>... >I seem to recall hearing/reading that applications can use both >Carbon and OpenStep APIs. > >Questions: >o Is this correct? (if not, ignore the rest) > Propably YES >o Since an application can use both "carbon" and "yellow" system >calls, should the "box" architecture description go away? > Apple has already stated that the "box" description is obsolete. Apple is not about boxes. >o Can this be used as a bridge to get MacOS developers to use >OpenStep? (That is, new additions or enhancements to existing >Carbonized applications could be done in OpenStep; thus allowing >MacOS developers to leverage the power of OpenStep while preserving >their existing work.) > Probably not. Mixing these APIs will result in a worst of both rather than a best of both situation in most cases. The sum will definitly be less than the parts. Which event loop will be used ? How will events and actions be translated/converted and when ? These are the least of the problems. This is a path to disaster with one exception. IF the logic/model of your existing application is seperate from the interface, there is no problem reusing the model within a YellowBox environment. Many of the most successful commercial Openstep applications have a core that was written in C or C++. All those years of tweeking and patching your ToolBox based GUI must be abandoned to use YellowBox effectively. The good news is that you can scrap 2/3 of your support burdon and bick up a better GUI for a small fraction of the effort and code required to build the old one. > >Todd >
From: Ari <ari@loa.com> Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: How mature is Carbon? Date: Sun, 31 May 1998 16:01:31 -0400 Organization: Log On America, Inc. Message-ID: <3571B71B.64D6AC86@loa.com> References: <01bd8cc5$31a42990$04387880@test1> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Its very mature. Its basically the old MacOS API w/the individual parts that restrict PMT hacked/removed/replaced. ari ari@loa.com Todd Heberlein wrote: > On one side, Carbon was just barely announced and Apple has not set > in stone the APIs in Carbon. > > On the other side, Carbon appeared mature enough to port a rich > application like Photoshop 5 (in two weeks!) and demonstrate it in > front of a live audience. > > So, how mature is Carbon? Was that really Carbon (with Photoshop) > being demonstrated on Rhapsody at WWDC 98 or was there some smoke ane > mirrors? > > Any clue as to when the first developers release of Carbon will be > shipped? That is, how soon can the top 100 MacOS developers move > from Carbon dating their applications to actually testing their > applications on Carbon? > > Just curious, > > Todd
From: nurban@crib.corepower.com (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 31 May 1998 16:16:31 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6ksdqv$j1j$1@crib.corepower.com> References: <6kp9sv$cdl$1@crib.corepower.com> <1998053118593100.OAA15861@ladder01.news.aol.com> In article <1998053118593100.OAA15861@ladder01.news.aol.com>, larrysb@aol.com (LarrySB) wrote: > Can anyone name 3 sku's of shrink-wrap software, on the shelves and available > at your local Fry's, CompUSA, Computer City, MicroCenter, etc, which was > written in OpenStep? Would you care to inquire as to the licensing costs of the OPENSTEP for Windows runtime? Even if you sold your app at cost, it would still be well out of the range of shrink-wrap software. A $20 runtime will change things for the shrink-wrap market, and a free one will change things more. Furthermore, your argument seems rather pointless; the number of shrink-wrap OpenStep apps in existence says little about the suitability of OpenStep apps for shrink-wrap development. Surely you're not naive enough to fall prey to "most popular = best to develop with" syndrome. Ergo, VC++/MFC must be infinitely superior as a development environment to anything else out there. Look at all that shrink-wrapped MFC-based software on the shelves! Being a professional MFC developer, I can tell you that VC++/MFC frankly sucks, and certainly doesn't compare to YB. > Even Nathan, a clearly vociferous supporter of AppKit can't convince his own > employer of it. Yeah, way to distort my statements. Is your argument so weak that you have to resort to such puerile behavior? Never mind, rhetorical question.. As I said quite clearly, I very likely _could_ convince my own employer of it, if we didn't already have 50k lines of code and two years of development effort invested in MFC. Management isn't going to let us put the app on freeze for the time it would take to rewrite in Yellow, no matter what the longer-term gains are. We have a constant stream of new feature requests and it's all we can do to get management to back off on some of the less important ones, and being a small company we don't exactly have the manpower to spare to go off and pursue a skunkworks project (as much as we grumble about wanting to). This is exactly the same reason why I'm _not_ advocating that MacOS developers port their apps to Yellow. It doesn't make sense in the majority of cases. But it does make a _lot_ of sense for developing _new_ apps that don't leverage existing code bases.
From: forrestDELETE-THIS!@west.net (Forrest Cameranesi) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.mac.programmer Subject: Re: How mature is Carbon? Message-ID: <forrestDELETE-THIS!-3105981358070001@term1-10.vta.west.net> References: <01bd8cc5$31a42990$04387880@test1> Date: Sun, 31 May 1998 20:55:05 GMT NNTP-Posting-Date: Sun, 31 May 1998 13:55:05 PDT (cross-posted to c.s.m.advocacy and c.s.m.programmer) In article <01bd8cc5$31a42990$04387880@test1>, "Todd Heberlein" <todd@NetSQ.com> wrote: > On one side, Carbon was just barely announced and Apple has not set > in stone the APIs in Carbon. > > On the other side, Carbon appeared mature enough to port a rich > application like Photoshop 5 (in two weeks!) and demonstrate it in > front of a live audience. Carbon is merely a subset of the Mac Toolbox, around 6000 calls (2000 less than the current Toolbox). While it's true that Apple has not set in stone exactly which of the Toolbox calls will be in Carbon, they have a general idea, and so Adobe could simply tune Photoshop not to use the calls that are certainly not going to be in Carbon, or to only use that calls that are certainly gonna be in Carbon, and viola - Carbon app. I'm not sure which OS it was running on (I wasn't at WWDC), but I'd bet it was on the current MacOS 8.1 (or possibly an 8.5 beta). That would not preclude it from being a Carbon app, because Carbon apps run unmodified on the current MacOS, just without all the modern buzzwords. Chances are that some apps are already 100% Carbon-ready, although those chances are rather slim (MacOS Rumors has some Carbon-dating specs on a few major apps). However, I've heard that it was running on Rhapsody, which could either be in the Blue Box or they might have already got a true Carbon "box" running (I know, old nomenclature, but I can't think of any other way of stating it). -Forrest Cameranesi Owner of The Universe, seeking employees for "Ruler" and "Master" positions. Apply at front desk. All prices subject to change. Universe, the Universe logo, and all objects contained within are (TM) and (C) Everything Unlimited, a wholly owned subsidary of Cameranesi Enterprises. Remember: "We know what you want. We know what you need. We know where you live."
From: Arnaud <debayeux@easynet.fr> Newsgroups: comp.sys.next.programmer Subject: Driver programming under OS 4.2 ? Date: Sun, 31 May 1998 23:11:31 +0200 Organization: [posted via] Easynet France Message-ID: <3571C783.FF925C5F@easynet.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello, I need to access the video registers (in.out ports like 0x3D0) and have a mapping to a physical memory address (0xA0000). I identified some kernel functions (IOMapPhysicalIntoIOTask, mapFrameBufferAtPhysicalAddress...) in headers or in the kernel symbols, but nothing in the libraries. Is it required to develop a driver ? Is it possible under Openstep 4.2 ? What's missing in OpenStep 4.2 to allow drivers development ? Simply a question of adding some NextStep 3.3 files ? Is there any documentation to develop drivers ? I believe I'll have to go back to Windows... Thanks for any support.
From: ggerard@flash.net (Greg Gerard) Newsgroups: comp.sys.next.programmer Subject: What ABI does carbon follow? Date: Sun, 31 May 1998 16:47:57 -0700 Organization: Flashnet Communications, http://www.flash.net Message-ID: <ggerard-3105981647580001@paltc4-194.flash.net> Do I get the power of the PowerPC/Mac runtime architecture (CFM and pals) or do I get the NeXT/Yellow runtime? thanks, greg
From: "Michael Brüggemann" <(no spam)mikebrug@vossnet.de> Newsgroups: comp.sys.mac.programmer.misc,comp.sys.mac.programmer.tools,comp.sys.mac.programmers.misc,comp.sys.newton.programmer,comp.sys.next.programmer,comp.sys.psion.programmer,comp.unix.programmer,comp.unix.sco.programmer,comp.windows.ms.programmer,de.comp.os.ms-w Subject: Re: Help Me PLease !!! Date: Mon, 1 Jun 1998 02:36:01 +0200 Organization: Voss Net Communications GmbH Message-ID: <6kssrl$tct$1@news.vossnet.de> References: <356E23DB.1F04832B@hkstar.com> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit NNTP-Posting-Date: 1 Jun 1998 00:32:53 GMT Capture what ? Direct X Screens ? 3D ? Be more specific. Mike myst schrieb in Nachricht <356E23DB.1F04832B@hkstar.com>... >Dear Sir, > > I'm a HK student, i'm very like to made > program... > > I have a big problem... > How to captrue DirectX Game Screen !? > > May you teach me ? > > Thankyou very Much !!! > >From Myst Ho.(myst@hkstar.com) > >
From: quinlan@intergate.bc.ca Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 01 Jun 1998 02:19:57 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6kt34d$h4l$1@nnrp1.dejanews.com> References: <6kgu3m$nc0$10@ns3.vrx.net> <B193753E-2A1857@192.168.128.1> <6km2m1$s3m$5@ns3.vrx.net> <356EE7C0.2F1CF0FB@ctron.com> <6kn7ci$q5r$1@news.digifix.com> In article <6kn7ci$q5r$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: > Apple will NEVER DO THIS. > > There is no benefit to them to do this at all. It only > dilutes the Mac only software base when it comes to existing software. But it may make some Mac developers who are thinking about focusing exclusively on Windows reconsider. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: *johnnyc*@or.psychology.dal.ca (John Christie) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: future of mac programming Date: Sun, 31 May 1998 23:44:39 -0300 Organization: ISINet, Nova Scotia Message-ID: <*johnnyc*-3105982344390001@jchristie.halifaxcable.dal.ca> References: <SCOTT.98May11153549@slave.doubleu.com> <B18C6E34-2E193D@192.168.128.1> <SCOTT.98May28082746@slave.doubleu.com> In article <SCOTT.98May28082746@slave.doubleu.com>, scott@doubleu.com (Scott Hess) wrote: > Obviously, the backing for the Carbon APIs is going to be an improved > BlueBox, unless Apple were to reimplement the entire API set for > Carbon (unlikely in the _extreme_). > > ["Improved", here, means "modified to allow for reentrancy and other > goodies promised by Apple". I take "BlueBox" to mean "Mac OS code > adapted to run over the new kernel".] Adapted to run over the new kernal and adapted to take advantage are not the same thing. Carbon takes advantage and Blue Box does not. the BlueBox is a virtual machine (or emulator to some, let's not get into that now) to run OS8 binaries. To extrapolate that the Carbon API is an improved Blue Box because it has a similar interface to the BlueBox does not do it justice. Apple needn't reimplement the entire API set. One could just as easily say that they are going to REINTERFACE the current YB and Mach API sets. They are doing some of both for many reasons. Carbon is an API similar to the current MacOS that makes it easy to move to a better OS (MacOSX). However, the BlueBox is a virtual machine that allows MacOS <X :) apps to run on X. To implement some of Carbon through the Blue Box would be bad news because the very nature of the Blue Box REQUIRES that it not have all of the modern (60's?) OS features touted for X. -- You aren't free if you CAN choose - only if you DO choose. All you are is the decisions you make. Remove "*" and ohnny (i.e. jc@) to reply via email
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 1 Jun 1998 03:32:23 GMT Organization: Digital Fix Development Message-ID: <6kt7c7$13g$1@news.digifix.com> References: <6kgu3m$nc0$10@ns3.vrx.net> <B193753E-2A1857@192.168.128.1> <6km2m1$s3m$5@ns3.vrx.net> <356EE7C0.2F1CF0FB@ctron.com> <6kn7ci$q5r$1@news.digifix.com> <6kt34d$h4l$1@nnrp1.dejanews.com> In-Reply-To: <6kt34d$h4l$1@nnrp1.dejanews.com> On 05/31/98, quinlan@intergate.bc.ca wrote: >In article <6kn7ci$q5r$1@news.digifix.com>, > sanguish@digifix.com (Scott Anguish) wrote: > >> Apple will NEVER DO THIS. >> >> There is no benefit to them to do this at all. It only >> dilutes the Mac only software base when it comes to existing software. > >But it may make some Mac developers who are thinking about focusing >exclusively on Windows reconsider. Thats what Yellow Box does. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: DTS sample: Son of SillyBalls Message-ID: <1998060103434600.XAA11810@ladder03.news.aol.com> Date: 01 Jun 1998 03:43:46 GMT Organization: AOL http://www.aol.com Ok, I'm looking at the latest DTS samples. Let's start with an easy one, Son of SillyBalls. Silly Balls was an age-old DTS sample program that drew painted circles randomly throughout a window, with a line of text inside each circle. It was a pretty dumb program, but it illustrated the basics of drawing in quickdraw. So, it got an update for YB. The program does basically the same thing, but is tremendously slower than the MacOS version. (to be fair, the MacOS version was a CPU hog and never called WaitNextEvent.) On examination, the S.o.S.B is an NSView which is overriden to draw the balloons. Simple enough. However, instead of a simple loop that continously draws the balls, it users an NSTimer, which calls the method in the NSView to draw a ball. OK, simple enough, but it seems a little bizare to me. Lets say that you want it to do the same thing, except draw the balls at the fastest possible speed. A NSThread seems out, because as far as I can tell, you can't use threads to update the UI of a YB program, because the Appkit isn't thread-safe. Is the NSTimer the right way to go, or is there a better way? Any of the NS experienced have an insight on a cooler, faster way? (Yes, despite my other comments in another message thread, I am asking in earnest.)
From: tbrown@netset.com (Ted Brown) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.mac.programmer Subject: Re: How mature is Carbon? Date: Mon, 01 Jun 1998 01:11:42 -0400 Organization: All USENET -- http://www.Supernews.com Message-ID: <tbrown-0106980111430001@mv131.axom.com> References: <01bd8cc5$31a42990$04387880@test1> <forrestDELETE-THIS!-3105981358070001@term1-10.vta.west.net> In article <forrestDELETE-THIS!-3105981358070001@term1-10.vta.west.net>, forrestDELETE-THIS!@west.net (Forrest Cameranesi) wrote: >Carbon is merely a subset of the Mac Toolbox, around 6000 calls (2000 less >than the current Toolbox). While it's true that Apple has not set in stone >exactly which of the Toolbox calls will be in Carbon, they have a general >idea, and so Adobe could simply tune Photoshop not to use the calls that >are certainly not going to be in Carbon, or to only use that calls that >are certainly gonna be in Carbon, and viola - Carbon app. I'm not sure >which OS it was running on (I wasn't at WWDC), but I'd bet it was on the >current MacOS 8.1 (or possibly an 8.5 beta). That would not preclude it >from being a Carbon app, because Carbon apps run unmodified on the current >MacOS, just without all the modern buzzwords. Chances are that some apps >are already 100% Carbon-ready, although those chances are rather slim >(MacOS Rumors has some Carbon-dating specs on a few major apps). However, >I've heard that it was running on Rhapsody, which could either be in the >Blue Box or they might have already got a true Carbon "box" running (I >know, old nomenclature, but I can't think of any other way of stating it). I wasn't at the WWDC either, but from what I remember reading, Carbon Photoshop was demoed on Rhapsody, not OS 8. -- Ted Brown tbrown@netset.com
From: rex@smallandmighty.com (Eric King) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 01 Jun 1998 03:20:45 -0400 Organization: The Small & Mighty Group Message-ID: <rex-0106980320490001@192.168.0.3> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> In article <6km2cm$s3m$4@ns3.vrx.net>, maury@remove_this.istar.ca (Maury Markowitz) wrote: : From what I've looked at (old BTW) the Be libs :were in many was similar to PP or TCL. Actually, with Replicants and a few other juicy tidbits, I think Be is progressing quite nicely beyond those other toolkits. They are having some pains supporting Intel, but then again who isn't. I'd like to see dynamic language for Be. Dave Stes' Portable Object Compiler is available, but it's kind of a large (and boring...) task to wrap up all of the Be APIs in Obj-C classes. There are also questions about multithreaded runtime performance. Any Obj-C app that used the Be API would need to have a *very* well-tuned multithreaded runtime. : Solid, but rather "thin". I heard :all the stories about thin being good, then I used YB. Give Be a few years. The Yellow Box didn't become that rich overnight. Personally, I wouldn't be surprised if Be's new Media Kit is actually better than what Apple provides on top of QTML. Adamation is also offering some pretty compelling technology for other Be developers to license. :but I do find it difficult to believe they have some of the :"gimme" classes in YB, like NSColor. Actually, it's highly likely that they do. The 2D graphics frameworks were among the oldest and most well-developed parts of Taligent. They were also influenced by GX, whose ink objects encapsulate quite a bit of color information. Taligent certainly had some juicy features like NURBS for both 2D & 3D graphics, and a time-based media framework that are far more interesting to me. : I'll bet for "information processing" :it's likely very similar though. You know, coupled with IBM's new incremental, headerless, C++ development system, I think OpenClass could kick some serious ass. : So specifically though, does IBM's Java stuff also have a OpenClass :framework set? Some of the Taligent stuff is being ported to Java for IBMs stuff, and Sun has licensed some other bits for inclusion into the Java spec. itself. The new 'area' primitive comes from Taligent. But AFAIK, there is no OpenClass for Java. (yet) :> There were frameworks for :> all sorts of things in CommonPoint. : : Examples? Time-based media, licensing, graphics editing, scanning, remote object calls, printing, etc. All sorts of stuff. I just don't think you could add all of that stuff to the Yellow Box without going through a major redesign of the Yellow Box. IMO, Taligent's troubles were mainly due to management. They gathered a whole bunch of really smart people to design the ultimate OS, then midstream, management decided that the world didn't need another OS, just some portable frameworks. (This is particularly ironic since Apple really could have used another OS...) The design decisions they had made for an OS, didn't quite make sense for a set of application frameworks. Everything was too big but they were still forced to ship *something* Also somewhere along the line, I believe their CEO died, which couldn't have helped matters. I think some interesting parallels could be drawn between Taligent and Java. I think perhaps too much functionality is being put into it. It was designed for applets, now it's being morphed into a pseudo-OS, and it's having every feature and capability under the sun built into it. I just can't see the end results turning out well. :Did they have anything like EO that was as good? You are asking two different things. Yes, they did have database access frameworks. (Remember it was aimed at the enterprise market...) However, I don't think anyone ever had the time to refine and test them to see if they were actually good in real world. Unless those classes end up in OpenClass and someone is inclined to do a comparison, I doubt we'll ever know. -Eric
From: quinlan@intergate.bc.ca Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 01 Jun 1998 09:38:28 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6ktsqk$dv7$1@nnrp1.dejanews.com> References: <6kgu3m$nc0$10@ns3.vrx.net> <B193753E-2A1857@192.168.128.1> <6km2m1$s3m$5@ns3.vrx.net> <356EE7C0.2F1CF0FB@ctron.com> <6kn7ci$q5r$1@news.digifix.com> <6kt34d$h4l$1@nnrp1.dejanews.com> <6kt7c7$13g$1@news.digifix.com> In article <6kt7c7$13g$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: > >But it may make some Mac developers who are thinking about focusing > >exclusively on Windows reconsider. > > Thats what Yellow Box does. I doubt that a Mac developer who is thinking of transitioning to Windows would be terribly enthused by the prospect of porting to YB. Maybe that's not rational put I believe it to be the case. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: pete@ohm.york.ac.uk (-bat.) Newsgroups: comp.sys.next.programmer Subject: Re: Unpacking raw data froma Tiff Date: 1 Jun 1998 10:34:58 GMT Organization: The University of York, UK Sender: pcf1@york.ac.uk Message-ID: <6ku04i$9td$1@pump1.york.ac.uk> References: <6kmkfq$j03$1@pump1.york.ac.uk> pete@ohm.york.ac.uk (-bat.) writes: > For those interested grab makekong.ohm.york.ac.uk:pub/pete/unpacktiff.tar.gz Oh dear - typo there I'm afraid ! That should read: maekong.ohm.york.ac.uk:pub/pete/unpacktiff.tar.gz sorry.
From: PC4LessOnline Subject: 200 MHZ PC's for less than $800 Message-ID: <QbM6UdUj9GA.112@cidintnews.infosel.com.mx> Newsgroups: comp.sys.next.programmer Date: Mon, 01 Jun 1998 05:23:34 -0500 That's right 200MHZ PC's with Monitor for less than $800. Check it out at http://www.pc4lessonline.com Thanks
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <QbM6UdUj9GA.112@cidintnews.infosel.com.mx> Control: cancel <QbM6UdUj9GA.112@cidintnews.infosel.com.mx> Date: 01 Jun 1998 05:56:25 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.QbM6UdUj9GA.112@cidintnews.infosel.com.mx> Sender: PC4LessOnline Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 1 Jun 1998 13:15:34 GMT Organization: Technical University of Berlin, Germany Message-ID: <6ku9hm$6ta$1@news.cs.tu-berlin.de> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit rex@smallandmighty.com (Eric King) writes: [...] > Time-based media, licensing, graphics editing, scanning, remote object >calls, printing, etc. All sorts of stuff. I just don't think you could add >all of that stuff to the Yellow Box without going through a major redesign >of the Yellow Box. I do. You underestimate the reflective abilities of Objective-C as well as the power of clean interfaces and categories. Could it be that your background includes heavy C++ programming and not really all that much exposure to truly dynamic languages, Objective-C, Smalltalk, CLOS? What makes these languages so different is exactly that, the ability to *integrate* entirely new concepts fully into existing, evolving systems. The C++ philosophy, OTOH, is that you must have a complete design to implement, and the language, AFAICT, doesn't really allow for anything else. [...] > I think some interesting parallels could be drawn between Taligent and >Java. I think perhaps too much functionality is being put into it. It was >designed for applets, now it's being morphed into a pseudo-OS, and it's >having every feature and capability under the sun built into it. I just ^^^ nice touch :-)) >can't see the end results turning out well. Yup. Especially since its foundations aren't really built for it. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: hhoff@ragnarok.en.REMOVETHIS.uunet.de (Holger Hoffstaette) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 1 Jun 1998 13:46:17 GMT Organization: the unstoppable code machine Message-ID: <6kubb9$54h@ragnarok.en.uunet.de> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Eric King wrote: > better than what Apple provides on top of QTML. Adamation is also offering > some pretty compelling technology for other Be developers to license. Adamation seem to have a knack for The NeXT Big Thing. Old time NeXTers for sure know that name! Adamation was also one of the first companies to develop for CommonPoint. Look where it got them. > :but I do find it difficult to believe they have some of the > :"gimme" classes in YB, like NSColor. > > Actually, it's highly likely that they do. The 2D graphics frameworks > were among the oldest and most well-developed parts of Taligent. They were > also influenced by GX, whose ink objects encapsulate quite a bit of color > information. Taligent certainly had some juicy features like NURBS for > both 2D & 3D graphics, and a time-based media framework that are far more > interesting to me. The Graphics and Text stuff in CommonPoint is indeed nifty, and I expect to see some of the features return in eQD and YB. > Time-based media, licensing, graphics editing, scanning, remote object > calls, printing, etc. All sorts of stuff. I just don't think you could add > all of that stuff to the Yellow Box without going through a major redesign > of the Yellow Box. The good news is that most of these features already exist in either Foundation or the AppKit. That leaves time based media, which is nicely handled by QT or NEXTIME or whatever Mike and the gang whip up. > IMO, Taligent's troubles were mainly due to management. They gathered a The incredible code bloat in CommonPoint cannot be blamed on management. It *can* be blamed on C++. > I think some interesting parallels could be drawn between Taligent and > Java. I think perhaps too much functionality is being put into it. It was > designed for applets, now it's being morphed into a pseudo-OS, and it's > having every feature and capability under the sun built into it. I just > can't see the end results turning out well. BINGO! When it comes to software, less is sometimes more. I'm not entirely sure why apparently not too many people are able to understand that. Java is a nice idea (one whose time had come - again) badly executed by the people who brought you vi and advocated C++. Why am I not surprised? They seem hell bent on repeating all the mistakes that Smalltalk, C++, Taligent and NeXT/OPENSTEP had to live through, while trying to please everyone at the same time. > :Did they have anything like EO that was as good? No. It's less than EOF *1* could do. In fact, it's about on par with the free SQLKit that NeXT shipped with WebObjects Pro (a free set of three or four SQL ODBC wrapper classes for people who couldn't afford or didn't need full EOF). > You are asking two different things. Yes, they did have database access > frameworks. (Remember it was aimed at the enterprise market...) However, I *cough* The whole description of the 'Enterprise services' fits on <20 pages in the 'Inside Taligent Technology' book, with the data access service covered on *2*. It leaves the same impression as the rest of the book: bla, bla, screenshots that look like NeXTSTEP, bla bla, not yet defined or implemented, more bla. The word 'framework' is used very loosely throughout the book, too. > don't think anyone ever had the time to refine and test them to see if > they were actually good in real world. Unless those classes end up in > OpenClass and someone is inclined to do a comparison, I doubt we'll ever > know. Why not stroll over to: http://hpsalo.cern.ch/TaligentDocs/TaligentOnline/DocumentRoot/1.0/Docs/ (I found this via AltaVista) Instead of just ranting and raving about the supposedly heavenly richness of CommonPoint, I heartily recommend checking this out, because it certainly makes YB look better than some people try to portrait it here. It also puts the 'Inside Taligent' book's fluff into perspective. Many people were happily coding away with NS/OS in '95, solving real problems and making money instead of just waving their hands about how kewl the future will be, Real Soon Now. It's one thing to make up a whole lot of great features, relegating the nitty gritty implementation details to some hazy future milestone, and getting something out the door that does the job 90% of the time while giving you the freedom to do the rest on your own if you are inclined to do so. Holger -- Object Architect, Object Factory GmbH | It's like a jungle sometimes, @work: holger"at"object-factory.com | it makes me wonder @home: hhoff"at"ragnarok.en.uunet.de | how I keep from going under
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Mixing Carbon and OpenStep calls?? Date: 1 Jun 1998 10:20:59 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6ktvab$a6t$7@ns3.vrx.net> References: <01bd8cc3$c8ccfab0$04387880@test1> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: todd@NetSQ.com In <01bd8cc3$c8ccfab0$04387880@test1> "Todd Heberlein" claimed: > I seem to recall hearing/reading that applications can use both > Carbon and OpenStep APIs. As far as I know, no. Maury sets run directly to the Mach/OSF code. Maury
From: nurban@crib.corepower.com (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 1 Jun 1998 13:38:04 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6kuots$kiv$1@crib.corepower.com> References: <6kgu3m$nc0$10@ns3.vrx.net> <6kt34d$h4l$1@nnrp1.dejanews.com> <6kt7c7$13g$1@news.digifix.com> <6ktsqk$dv7$1@nnrp1.dejanews.com> In article <6ktsqk$dv7$1@nnrp1.dejanews.com>, quinlan@intergate.bc.ca wrote: > I doubt that a Mac developer who is thinking of transitioning to Windows > would be terribly enthused by the prospect of porting to YB. Do you think they would be more enthused by the prospect of porting to MFC? I hope not!
From: rex@smallandmighty.com (Eric King) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 01 Jun 1998 14:23:36 -0400 Organization: The Small & Mighty Group Message-ID: <rex-0106981423360001@192.168.0.3> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6kubb9$54h@ragnarok.en.uunet.de> In article <6kubb9$54h@ragnarok.en.uunet.de>, hhoff@ragnarok.en.REMOVETHIS.uunet.de (Holger Hoffstaette) wrote: : Eric King wrote: :> better than what Apple provides on top of QTML. Adamation is also offering :> some pretty compelling technology for other Be developers to license. : :Adamation seem to have a knack for The NeXT Big Thing. Old time NeXTers :for sure know that name! Adamation was also one of the first companies :to develop for CommonPoint. Look where it got them. Yep, though, Studio/A does look like a pretty hot product. I haven't heard of any Yellow video editing suites in the works. My guess is that most will stay Carbonized. :The Graphics and Text stuff in CommonPoint is indeed nifty, and I expect :to see some of the features return in eQD and YB. Some of the features were already there. My main beef is that Apple's taking the features and not the design. GX and Taligent's graphics & text frameworks were very nicely designed. Scattering the features through ATSUI, eQD, and the Appkit is just going to remove a lot of the technology's power, IMO. When everything stops shifting around, I just hope that Apple does a good job of documenting the resulting system, cause it's not going to be pretty. :That leaves time based media, which is nicely :handled by QT or NEXTIME or whatever Mike and the gang whip up. Actually it's Mark and the gang (different division), and he's already said there's not going to be a high-level multimedia framework for the YellowBox. You'll be able to playback and record movies, but don't expect much beyond that. Though, maybe Apple will do something interesting with the Final Cut technology they bought. :The incredible code bloat in CommonPoint cannot be blamed on management. :It *can* be blamed on C++. I think featuritis is more to blame. Fewer features = less code bloat. Some ex-Taligent folks have blamed C++ others have blamed management. It's probably a mixture. But then again, no one (save perhaps Sun with Java) has really attempted such an ambitious OS project. Even more dynamic language environments like Java, aren't holding up well to feature bloat and constantly shifting strategic directions. :the people who brought you vi and advocated C++. Why am I not surprised? :They seem hell bent on repeating all the mistakes that Smalltalk, C++, :Taligent and NeXT/OPENSTEP had to live through, while trying to please :everyone at the same time. Yep, they really just need to *stop* for a couple of years and just make what they have work well. :Instead of just ranting and raving about the supposedly heavenly richness :of CommonPoint, I'm not raving about the heavenly richness. Really, I'm only interested in their graphics & text frameworks. I think they tried to put way too much in, and I don't think any environment, even a dynamic one like OpenStep, would have held up to the featuritis and mismanagement that Taligent underwent. :Many people were happily coding away with NS/OS in '95, solving real :problems and making money instead of just waving their hands about how :kewl the future will be, Real Soon Now Yup, though that's also a counterargument. Many more people were solving real problems and making more money using the Mac OS and Windows. Furthermore many more folks will be solving real problems with Carbon as opposed to the YellowBox. I think it will be interesting to see what directions MetroWerks takes PowerPlant, now that they have the YellowBox as a competitor. :. It's one thing to make up a whole :lot of great features, relegating the nitty gritty implementation details :to some hazy future milestone, and getting something out the door :that does the job 90% of the time while giving you the freedom to do the :rest on your own if you are inclined to do so. Well, that seems to be more of the approach IBM is taking with their OpenClass frameworks. Trim away the cruft and nonessential features, improve the development environment, etc. As much as I don't like C++, the next version of VisualAge C++ does look pretty intriguing. -Eric
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 1 Jun 1998 12:41:16 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6ku7hc$g29$2@ns3.vrx.net> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: rex@smallandmighty.com In <rex-0106980320490001@192.168.0.3> Eric King claimed: > Actually, with Replicants and a few other juicy tidbits, Uhhh, isn't this the part that's conceptually similar to forwarding under Obj-C? Or am I thinking of something else? > Give Be a few years. The Yellow Box didn't become that rich overnight. Sure, but the issue as I see it is whether or not it will make a difference by that time. Although the OSF projects are sticking with a "2nd gen" uKernel, they are making some rather interesting advnaces in it that I think will make it rather competitive in terms of throughput performance (that remains to be seen, but then so does Be). > Personally, I wouldn't be surprised if Be's new Media Kit is actually > better than what Apple provides on top of QTML. I would. It took years to build QT, that's not something anyone can quickly duplicate. > Some of the Taligent stuff is being ported to Java for IBMs stuff, and > Sun has licensed some other bits for inclusion into the Java spec. itself. > The new 'area' primitive comes from Taligent. But AFAIK, there is no > OpenClass for Java. (yet) Ok. > Time-based media, licensing, graphics editing, scanning, remote object > calls, printing, etc. All sorts of stuff. I just don't think you could add > all of that stuff to the Yellow Box without going through a major redesign > of the Yellow Box. Well, asside from the fact that they all DID exist in the past I assume! (or still do) > :Did they have anything like EO that was as good? > > You are asking two different things. Yes, they did have database access > frameworks. But _were_ they good? > don't think anyone ever had the time to refine and test them to see if > they were actually good in real world. Ahhh. Maury
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: What ABI does carbon follow? Date: 1 Jun 1998 12:43:45 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6ku7m1$g29$3@ns3.vrx.net> References: <ggerard-3105981647580001@paltc4-194.flash.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: ggerard@flash.net In <ggerard-3105981647580001@paltc4-194.flash.net> Greg Gerard claimed: > Do I get the power of the PowerPC/Mac runtime architecture (CFM and pals) > or do I get the NeXT/Yellow runtime? Mac. That's the point basically. It's more specifically Mac PPC, so no CFM-68k and such. Maury
From: rex@smallandmighty.com (Eric King) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 01 Jun 1998 15:38:56 -0400 Organization: The Small & Mighty Group Message-ID: <rex-0106981538570001@192.168.0.3> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku9hm$6ta$1@news.cs.tu-berlin.de> In article <6ku9hm$6ta$1@news.cs.tu-berlin.de>, marcel@cs.tu-berlin.de (Marcel Weiher) wrote: : Could it be that your background includes heavy C++ programming No, not really. I've always thought of C++ as an ugly language. Most of the low-level graphics I write code is in C, but I'm quite willing (actually I very much prefer...) to use other languages and environments for my high-level code. The best thing that can be hoped for with this Java nonsense is that it could eventually replace C++ in a lot of areas, and that would be a good thing. : and not really all that much exposure to truly dynamic languages, Objective-C, Smalltalk, CLOS? I've done quite a bit of work in Prograph, which is a very dynamic environment, more so than Obj-C. Prograph really spoiled me, I just hate working with textual languages; code is just so one-dimensional I've also been fiddling with Smalltalk on and off for the last three years or so. Right now, I'm becoming quite enamored with Squeak. Though even with all of the browsing and inspecting facilities of Smalltalk, I find myself really missing the ability to look at my code from a distance and instantly knowing what's going where. :What makes these languages so different is exactly that, the ability :to *integrate* entirely new concepts fully into existing, evolving :systems. Quite true, though at some point one needs to stop integrating new concepts and just do a redesign otherwise you're just going to end up with something that works, but is non-optimal. A while back, some EO developers were complaining in this group about the problems they were having scaling EO up to handle really large systems. (Something about wasteful resource duplication and such.) While they loved the interface and functionality, some of the assumptions it made just weren't appropriate for larger systems. Of course, the counter argument is that you can't please everyone, and that it just works for most people. But still, the incremental design approach can have its share of problems. :The C++ philosophy, OTOH, is that you must have a complete :design to implement, and the language, AFAICT, doesn't really allow for :anything else. No disagreement here. My argument isn't that C++ was the right language for Taligent, but that if you try a project that large with that kind of management in any language you're going to run into problems. Still, that fiasco did produce a few bits of very nice code. Enough so, that IBM's still sticking with it. :Yup. Especially since its foundations aren't really built for it. Without good management and clear strategic directions, it doesn't matter how good its foundations are. You're still going to get an icky gooey mess. -Eric
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 1 Jun 1998 13:03:07 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6ku8qb$g29$4@ns3.vrx.net> References: <6kp9sv$cdl$1@crib.corepower.com> <1998053118593100.OAA15861@ladder01.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: larrysb@aol.com In <1998053118593100.OAA15861@ladder01.news.aol.com> LarrySB claimed: > Can anyone name 3 sku's of shrink-wrap software, on the shelves and available > at your local Fry's, CompUSA, Computer City, MicroCenter, etc, which was > written in OpenStep? Hmmm, that's like asking if anyone can find 3 PowerBuilder apps on the store shelves. It's a meaningless question. On the other hand I can find quite a few WO apps on the web, which is a more useful comparison. > So far, no one has bothered to "eat my lunch." We still don't know what you do. > I can't think of a single title written in OpenStep that was deployed > commercially on Windows, at least in a shink-wrap form. We release soon. Maury
From: rex@smallandmighty.com (Eric King) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 01 Jun 1998 18:00:51 -0400 Organization: The Small & Mighty Group Message-ID: <rex-0106981800560001@192.168.0.3> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku7hc$g29$2@ns3.vrx.net> In article <6ku7hc$g29$2@ns3.vrx.net>, maury@remove_this.istar.ca (Maury Markowitz) wrote: :In <rex-0106980320490001@192.168.0.3> Eric King claimed: :> Actually, with Replicants and a few other juicy tidbits, : : Uhhh, isn't this the part that's conceptually similar to forwarding under :Obj-C? Or am I thinking of something else? No, it's closer to OpenDOC. Replicants are a way of instantiating flattened objects on the fly in the BeOS. The flattened objects are stored in BMessages and can be read from files, sent from app to app, over networks, etc. Most replicants are descended from BView, but anything that descends from BArchivable can become a replicant. :> Personally, I wouldn't be surprised if Be's new Media Kit is actually :> better than what Apple provides on top of QTML. : : I would. It took years to build QT, that's not something anyone can :quickly duplicate. It's also not something that anyone needs to fully duplicate. Does Be need to reproduce the portable versions of Quickdraw, the Resource, Memory, and other crufty Mac OS Managers, that are all built into QTML? No. Also a lot of QTML's features just aren't being used by developers. All Be needs to do is support the popular codecs (many of which Apple has licensed, like Indeo, Cinepak, and the new Sorenson codec) and provide good synchronization, playback, recording, and other services. Be is pretty focused in this area, and I think they'll be able to pull it off. The recent changes they've made to their windowing system are a strong indication that they're moving in the right directions. Plus there's that new Adamation stuff to play with. Incidentally, it would be nice to know what Apple's planning on doing with Final Cut. :> don't think anyone ever had the time to refine and test them to see if :> they were actually good in real world. : : Ahhh There graphics and text classes still look excellent though. It's truly a shame Apple's not using any of them. Licensing some of that technology for Java was one of Sun's smarter moves. Much more so than that non-existent Bravo cruft. -Eric
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 1 Jun 1998 16:59:14 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6kuml2$rhg$1@ns3.vrx.net> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku7hc$g29$2@ns3.vrx.net> <rex-0106981800560001@192.168.0.3> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: rex@smallandmighty.com In <rex-0106981800560001@192.168.0.3> Eric King claimed: > No, it's closer to OpenDOC. Ahhh, I knew I recognized the term. > Replicants are a way of instantiating > flattened objects on the fly in the BeOS. The flattened objects are stored > in BMessages and can be read from files, sent from app to app, over > networks, etc. Most replicants are descended from BView, but anything that > descends from BArchivable can become a replicant. How is it different from objects supporting the NSCoding protocol? I understand there's some support for embedding of some form, how do they do this? > It's also not something that anyone needs to fully duplicate. Does Be > need to reproduce the portable versions of Quickdraw, the Resource, > Memory, and other crufty Mac OS Managers, that are all built into QTML? No, but neither did Apple when they created it. Even 2.5 is the most capable media wrapper layer available - many others have tried, including MS, but all have failed to match it. Sorry, but I don't thing Be has much of a chance to win here where others have failed. > All Be needs to do is support the popular codecs (many of which Apple > has licensed, like Indeo, Cinepak, and the new Sorenson codec) Ohhhh, that's what you mean. Well, sure, I suppose. > There graphics and text classes still look excellent though. I've been looking at them. The color class is indeed a good subset of the YB version, and it even includes support for color matching in all classes, something that hasn't been added yet to YB (but is on the way). The text handling is similar, as is the string and date handling. In fact overall it looks like a direct rip-off of YB (and looking at the timeline I don't _think_ it could be the other way around). With the exception of the drawing stuff of course, which I actually quite like. Looking over it though it seems to be designed by some other team. Be that as it may, I think the drawing stuff would be great in YB. Maury
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 1 Jun 1998 20:24:12 GMT Organization: Digital Fix Development Message-ID: <6kv2lc$klv$1@news.digifix.com> References: <6kgu3m$nc0$10@ns3.vrx.net> <B193753E-2A1857@192.168.128.1> <6km2m1$s3m$5@ns3.vrx.net> <356EE7C0.2F1CF0FB@ctron.com> <6kn7ci$q5r$1@news.digifix.com> <6kt34d$h4l$1@nnrp1.dejanews.com> <6kt7c7$13g$1@news.digifix.com> <6ktsqk$dv7$1@nnrp1.dejanews.com> In-Reply-To: <6ktsqk$dv7$1@nnrp1.dejanews.com> On 06/01/98, quinlan@intergate.bc.ca wrote: >In article <6kt7c7$13g$1@news.digifix.com>, > sanguish@digifix.com (Scott Anguish) wrote: > >> >But it may make some Mac developers who are thinking about focusing >> >exclusively on Windows reconsider. >> >> Thats what Yellow Box does. > >I doubt that a Mac developer who is thinking of transitioning to Windows >would be terribly enthused by the prospect of porting to YB. Maybe that's not >rational put I believe it to be the case. Its not rational. If they're going to learn a whole new framework, why not learn one that will allow them to support both Windows and MacOS X? -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: quinlan@intergate.bc.ca Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 01 Jun 1998 21:58:28 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6kv863$5nb$1@nnrp1.dejanews.com> References: <6kgu3m$nc0$10@ns3.vrx.net> <6kt34d$h4l$1@nnrp1.dejanews.com> <6kt7c7$13g$1@news.digifix.com> <6ktsqk$dv7$1@nnrp1.dejanews.com> <6kuots$kiv$1@crib.corepower.com> In article <6kuots$kiv$1@crib.corepower.com>, nurban@vt.edu wrote: > Do you think they would be more enthused by the prospect of porting > to MFC? I hope not! Yes, I believe that they would be. I think that we are confusing the decision that is most rational and the decision that most people will make :) Not that I necessarily agree that a YB port is the most rational decision. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: rex@smallandmighty.com (Eric King) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 01 Jun 1998 19:24:54 -0400 Organization: The Small & Mighty Group Message-ID: <rex-0106981924540001@192.168.0.3> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku7hc$g29$2@ns3.vrx.net> <rex-0106981800560001@192.168.0.3> <6kuml2$rhg$1@ns3.vrx.net> In article <6kuml2$rhg$1@ns3.vrx.net>, maury@remove_this.istar.ca (Maury Markowitz) wrote: : How is it different from objects supporting the NSCoding protocol? It's probably rather similar. :understand there's some support for embedding of some form, how do they do :this? The easiest way is through BShelf objects. Put one in a view and you're all set to have embedded replicants. Programmatically, you can just instantiate the appropriate class from the replicant's BMessage. It's not as flexible as Obj-C or other dynamic languages, but it's still pretty cool, especially when coupled with Be's scripting system. In many instances, BMessages can take the place of true dynamic dispatching in the BeOS. :> There graphics and text classes still look excellent though. : : I've been looking at them. The color class is indeed a good subset of the :YB version, and it even includes support for color matching in all classes, :something that hasn't been added yet to YB (but is on the way). The text :handling is similar, as is the string and date handling. In fact overall it :looks like a direct rip-off of YB (and looking at the timeline I don't :_think_ it could be the other way around). Oh, they certainly snagged tons of ideas from NextStep, which in turn drew a lot from the various Smalltalk environments. There's no denying the heritage. For instance, the user interface shots are *very* NeXT-like. However, much of the graphics and text frameworks were inspired by new work Apple was doing with Quicktime, QD GX, and QD 3D. : With the exception of the drawing stuff of course, which I actually quite :like. Looking over it though it seems to be designed by some other team. From what I've been told, the 2D classes, go all the way back to when it was still called Pink. :that as it may, I think the drawing stuff would be great in YB. It'd be awesome. Ported to Obj-C or even just provided as a static C++ library. Full integration could take years to do right, but the functionality could be provided sooner. Apple's not going to do it, though. They've thoroughly divorced themselves from CommonPoint. IBM's still fiddling around with it though. -Eric
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 1 Jun 1998 19:11:12 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6kuucg$2uf$2@ns3.vrx.net> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku7hc$g29$2@ns3.vrx.net> <rex-0106981800560001@192.168.0.3> <6kuml2$rhg$1@ns3.vrx.net> <rex-0106981924540001@192.168.0.3> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: rex@smallandmighty.com In <rex-0106981924540001@192.168.0.3> Eric King claimed: > It's probably rather similar. Ok. > The easiest way is through BShelf objects. Ahhh, that's the term I was looking for. > Put one in a view and you're > all set to have embedded replicants. Programmatically, you can just > instantiate the appropriate class from the replicant's BMessage. So it encapsulates both the binary and the flattened instance variables? Or do you mean when you unpack it that you then create an instance from the data inside? If you mean the later then it's definitely similar to Coders. But what exactly does the Shelf do? I was under the impression that you could drag a "clock part" out to the "Finder" and it would just run. The problem is that I can't understand the mechanism used to support this. > as flexible as Obj-C or other dynamic languages, but it's still pretty > cool, especially when coupled with Be's scripting system. In many > instances, BMessages can take the place of true dynamic dispatching in the > BeOS. Here I certainly prefer to the Obj-C solution - which is "none, because you don't need it". > Oh, they certainly snagged tons of ideas from NextStep, which in turn > drew a lot from the various Smalltalk environments. There's no denying the > heritage. For instance, the user interface shots are *very* NeXT-like. Yes, this is something the boss remarked on. > However, much of the graphics and text frameworks were inspired by new > work Apple was doing with Quicktime, QD GX, and QD 3D. Lawson looks at this and says "look, GX", I look at it as say "look, PS". I get the feeling that there's as much new as old. > From what I've been told, the 2D classes, go all the way back to when > it was still called Pink. Wow. > It'd be awesome. Ported to Obj-C or even just provided as a static C++ > library. Full integration could take years to do right, but the > functionality could be provided sooner. Apple's not going to do it, > though. I'm not convinced something like it won't appear though. Heck, it is by bits and pieces now, the (poorly named) NSBezierPath is one small step down that road. Maury
From: nurban@crib.corepower.com (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 1 Jun 1998 18:28:14 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6kv9tu$lgr$1@crib.corepower.com> References: <6kgu3m$nc0$10@ns3.vrx.net> <6ktsqk$dv7$1@nnrp1.dejanews.com> <6kuots$kiv$1@crib.corepower.com> <6kv863$5nb$1@nnrp1.dejanews.com> In article <6kv863$5nb$1@nnrp1.dejanews.com>, quinlan@intergate.bc.ca wrote: > In article <6kuots$kiv$1@crib.corepower.com>, > nurban@vt.edu wrote: > > Do you think they would be more enthused by the prospect of porting > > to MFC? I hope not! > Yes, I believe that they would be. I think that we are confusing the decision > that is most rational and the decision that most people will make :) Well, I'm not aruging whether or not they're more likely to choose MFC over YB, or which the rational choice is for overall development purposes (though I'm sure you can guess my opinion from other threads) -- but though I can see people choosing MFC over YB for various reasons, I still don't see why they would be more _enthusiastic_ about using MFC, given that they know something about MFC and YB. Maybe they just don't have enough MFC experience for their skin to crawl.
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: DTS sample: Son of SillyBalls Date: 1 Jun 1998 22:17:57 GMT Organization: Idiom Communications Message-ID: <6kv9al$ku0$1@news.idiom.com> References: <1998060103434600.XAA11810@ladder03.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: larrysb@aol.com LarrySB may or may not have said: -> Ok, I'm looking at the latest DTS samples. -> -> Let's start with an easy one, Son of SillyBalls. -> -> Silly Balls was an age-old DTS sample program that drew painted circles -> randomly throughout a window, with a line of text inside each circle. It was a -> pretty dumb program, but it illustrated the basics of drawing in quickdraw. -> -> So, it got an update for YB. -> -> The program does basically the same thing, but is tremendously slower than the -> MacOS version. (to be fair, the MacOS version was a CPU hog and never called -> WaitNextEvent.) It's going at whatever rate the NSTimer is telling it to go. -> On examination, the S.o.S.B is an NSView which is overriden to draw the -> balloons. Simple enough. However, instead of a simple loop that continously -> draws the balls, it users an NSTimer, which calls the method in the NSView to -> draw a ball. -> -> OK, simple enough, but it seems a little bizare to me. Lets say that you want -> it to do the same thing, except draw the balls at the fastest possible speed. -> A NSThread seems out, because as far as I can tell, you can't use threads to -> update the UI of a YB program, because the Appkit isn't thread-safe. -> -> Is the NSTimer the right way to go, or is there a better way? Depends on what your goal is. Using an NSTimer, you can make the balls appear at whatever rate you want, basically. Just set the desired interval on the NSTimer. BTW, you *really*don't want the -drawRect: method of an NSView to loop indefinitely. -> Any of the NS experienced have an insight on a cooler, faster way? If you just want to see those balls getting drawn the fastest way possible, then you use a non-retained window, and draw the balls in a loop that quits when it sees a mousedown event. -jcr
From: quinlan@intergate.bc.ca Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Tue, 02 Jun 1998 01:48:50 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6kvlm2$pup$1@nnrp1.dejanews.com> References: <6kgu3m$nc0$10@ns3.vrx.net> <6ktsqk$dv7$1@nnrp1.dejanews.com> <6kuots$kiv$1@crib.corepower.com> <6kv863$5nb$1@nnrp1.dejanews.com> <6kv9tu$lgr$1@crib.corepower.com> In article <6kv9tu$lgr$1@crib.corepower.com>, nurban@vt.edu wrote: > Well, I'm not aruging whether or not they're more likely to choose > MFC over YB, or which the rational choice is for overall development > purposes (though I'm sure you can guess my opinion from other threads) > -- but though I can see people choosing MFC over YB for various reasons, > I still don't see why they would be more _enthusiastic_ about using MFC, > given that they know something about MFC and YB. Maybe they just don't > have enough MFC experience for their skin to crawl. Sorry, I misunderstood you. I don't disagree with your assement of the enthusiasm involved in porting to YB versus MFC. I am one of the few people that I know of using a framework so bad that MFC seems like paradise :) -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: quinlan@intergate.bc.ca Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Tue, 02 Jun 1998 01:56:33 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6kvm4h$qms$1@nnrp1.dejanews.com> References: <6kgu3m$nc0$10@ns3.vrx.net> <B193753E-2A1857@192.168.128.1> <6km2m1$s3m$5@ns3.vrx.net> <356EE7C0.2F1CF0FB@ctron.com> <6kn7ci$q5r$1@news.digifix.com> <6kt34d$h4l$1@nnrp1.dejanews.com> <6kt7c7$13g$1@news.digifix.com> <6ktsqk$dv7$1@nnrp1.dejanews.com> <6kv2lc$klv$1@news.digifix.com> In article <6kv2lc$klv$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: > Its not rational. > > If they're going to learn a whole new framework, why not learn > one that will allow them to support both Windows and MacOS X? Scott, you're missing the different between "should" and "will". People don't always make the rational choice. Also, I believe, people perceive Apple to not have sufficiently commitment to YB to make porting to YB a safe bet. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: quinlan@intergate.bc.ca Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Tue, 02 Jun 1998 01:57:04 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6kvm5g$qn7$1@nnrp1.dejanews.com> References: <6kgu3m$nc0$10@ns3.vrx.net> <B193753E-2A1857@192.168.128.1> <6km2m1$s3m$5@ns3.vrx.net> <356EE7C0.2F1CF0FB@ctron.com> <6kn7ci$q5r$1@news.digifix.com> <6kt34d$h4l$1@nnrp1.dejanews.com> <6kt7c7$13g$1@news.digifix.com> <6ktsqk$dv7$1@nnrp1.dejanews.com> <6kv2lc$klv$1@news.digifix.com> In article <6kv2lc$klv$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: > Its not rational. > > If they're going to learn a whole new framework, why not learn > one that will allow them to support both Windows and MacOS X? Scott, you're missing the different between "should" and "will". People don't always make the rational choice. Also, I believe, people perceive Apple to not have sufficiently committed to YB to make porting to YB a safe bet. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: rex@smallandmighty.com (Eric King) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Mon, 01 Jun 1998 22:29:37 -0400 Organization: The Small & Mighty Group Message-ID: <rex-0106982229380001@192.168.0.3> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku7hc$g29$2@ns3.vrx.net> <rex-0106981800560001@192.168.0.3> <6kuml2$rhg$1@ns3.vrx.net> <rex-0106981924540001@192.168.0.3> <6kuucg$2uf$2@ns3.vrx.net> In article <6kuucg$2uf$2@ns3.vrx.net>, maury@remove_this.istar.ca (Maury Markowitz) wrote: :In <rex-0106981924540001@192.168.0.3> Eric King claimed: :> Put one in a view and you're :> all set to have embedded replicants. Programmatically, you can just :> instantiate the appropriate class from the replicant's BMessage. : : So it encapsulates both the binary and the flattened instance variables? Not quite, the OS finds the binary through a mime type embedded in the BMessage. Though, I suppose it wouldn't be hard to cram the binary into the message also, which might make sense if you were sending one over a network. But there would be security issues to deal with. :Or do you mean when you unpack it that you then create an instance from the :data inside? If you mean the later then it's definitely similar to Coders. Yep, that's pretty much the way it works. Some clever folks are using it to patch the parts of the OS, since replicants run in their hosts address space. :But what exactly does the Shelf do? Provide a convenient user interface item. All you do is add a BShelf as a child of your view and you're ready to recieve replicants. It handles all of the instantiation work for you. : I was under the impression that you could drag a "clock part" out to the :"Finder" and it would just run. The problem is that I can't understand the :mechanism used to support this. The background window of the Tracker is a big BShelf. Drop a replicant on their and it starts to run. It's all done through shared libraries, mime types, and BMessages. It's simple and works remarkably well. : Here I certainly prefer to the Obj-C solution - which is "none, because you :don't need it". But I can see instances where Be's solution makes more sense. Since their OS is so heavily threaded, it makes a sense to directly use thinly-wrapped kernel messages to communicate. This is probably what PDO is doing behind the scenes. Currently, it's a minor inconvenience, but it is very efficient. :> However, much of the graphics and text frameworks were inspired by new :> work Apple was doing with Quicktime, QD GX, and QD 3D. : : Lawson looks at this and says "look, GX", I look at it as say "look, PS". :I get the feeling that there's as much new as old. Actually, I looked and saw GX also. There's a one to one mapping between many of the objects. For instance, instead of Tag objects, you have bundles. I believe they even tried to support GX's font technology and line layout. The Postscript way is to use little programs or display lists to draw shapes. The primitives contain only geometric information. Color, transforms, and other bits of scene state are contained in scene globals. The Taligent/GX way is to provide concrete objects that encapsulate a scene's state. In Postscript, Classic Quickdraw, and probably in Quickdraw Extended, you do things like: SetForeGroundColor(somecolor) Stroke(somePath) In Taligent and GX you do things like: somepathshape->SetColor(somecolor) somepathshape->Draw Given an appropriate amount of time and effort, you can make one method emulate the other, though it's a lot less work to make the latter emulate the former, than the other way around. Personally, I prefer the latter, more OOPish, but some developers do indeed prefer the former. I think the Appkit has you a bit confused in that it often behaves like the latter, but really the underlying imaging model is like the former. :> It'd be awesome. Ported to Obj-C or even just provided as a static C++ :> library. Full integration could take years to do right, but the :> functionality could be provided sooner. Apple's not going to do it, :> though. : : I'm not convinced something like it won't appear though. I am. It just wouldn't help the big developers much, so Apple's probably not going to invest the time and effort. :bits and pieces now, the (poorly named) NSBezierPath is one small step down :that road. We'll get more pieces as time goes by, but don't expect anything particularly cohesive. -Eric
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.next.programmer Subject: Re: You don't measure a language/framework by the number of books on it.. Re: future of mac programming Date: 2 Jun 1998 04:32:08 GMT Organization: MiscKit Development Message-ID: <6kvv88$ool$1@news.xmission.com> References: <6kn8qd$loc3@onews.collins.rockwell.com> <1998053003404000.XAA17471@ladder01.news.aol.com> <6kqn2d$3v9$1@news.digifix.com> <01bd8dd7$39eed380$04387880@test1> "Todd Heberlein" <todd@NetSQ.com> wrote: > Scott Anguish <sanguish@digifix.com> wrote > > This is such a bone-headed metric that I feel compelled to > > comment on it. > > Finding *good* metrics has traditionally been difficuly in the > computer field. [...] > the number of books can be an indication of where > the publishers and bookstores think the market is going. See, this brings us right back to Scott's point: Are you going to pick your programming language based upon what a publisher or bookstore thinks is the right language for the job, or will you consult a software engineer who actually _knows_ _how_ _to_ _program_? I see a lot of irony there--especially in the fact that, yes indeed, most people will trust the bookstore more than an experienced programmer. Go figure! > And out of the masses of > dimwits might come some geniuses who will create killer > apps. I guess it could be considered a statistical > approach to success. Yup, that's the irony right there, by golly. An infinite number of monkeys, indeed! Sadly, if you consider the amount of money Microsloth is pouring into R&D, the statistical approach simply isn't working. Guess they need to invest more... :-/ (If anyone thinks it is hard for Apple to keep up since they spend over an order of magnitude less on R&D, think again. Considering people I've met that work at Microsoft and looking at their tools, Apple is buying much better programmers and has much better tools...so they're spending their money in much smarter ways. It balances the equation and tips things toward Apple's side, in fact. It does pay off handsomely.) -- Later, -Don Yacktman don@misckit.com <a href="http://www.misckit.com/don.html">My home page</a>
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 2 Jun 1998 03:19:26 GMT Organization: Digital Fix Development Message-ID: <6kvqvu$t9d$1@news.digifix.com> References: <6kgu3m$nc0$10@ns3.vrx.net> <B193753E-2A1857@192.168.128.1> <6km2m1$s3m$5@ns3.vrx.net> <356EE7C0.2F1CF0FB@ctron.com> <6kn7ci$q5r$1@news.digifix.com> <6kt34d$h4l$1@nnrp1.dejanews.com> <6kt7c7$13g$1@news.digifix.com> <6ktsqk$dv7$1@nnrp1.dejanews.com> <6kv2lc$klv$1@news.digifix.com> <6kvm4h$qms$1@nnrp1.dejanews.com> In-Reply-To: <6kvm4h$qms$1@nnrp1.dejanews.com> On 06/01/98, quinlan@intergate.bc.ca wrote: >In article <6kv2lc$klv$1@news.digifix.com>, > sanguish@digifix.com (Scott Anguish) wrote: > >> Its not rational. >> >> If they're going to learn a whole new framework, why not learn >> one that will allow them to support both Windows and MacOS X? > >Scott, you're missing the different between "should" and "will". People don't >always make the rational choice. I'm not addressing "should" or "will". >Also, I believe, people perceive Apple to not have sufficiently commitment to >YB to make porting to YB a safe bet. Apple's commited to it.. unfortunately they can't really do much to change the 'perception' that some people have about YB short of just shipping it. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: "Todd Heberlein" <todd@NetSQ.com> Newsgroups: comp.sys.next.programmer Subject: Re: You don't measure a language/framework by the number of books on it.. Re: future of mac programming Date: 2 Jun 1998 03:35:29 GMT Organization: Net Squared, Inc. Message-ID: <01bd8dd7$39eed380$04387880@test1> References: <6kn8qd$loc3@onews.collins.rockwell.com> <1998053003404000.XAA17471@ladder01.news.aol.com> <6kqn2d$3v9$1@news.digifix.com> Scott Anguish <sanguish@digifix.com> wrote > This is such a bone-headed metric that I feel compelled to > comment on it. Finding *good* metrics has traditionally been difficuly in the computer field. Other famous measurements include: o MIPS - "Millions of Instructions Per Second" or "Meaningless Index of Processor Speed". o MHz - general clock rate of the CPU. o KLOC - Thousands (K) Lines Of Code (a popular measurement by IBM standards - bloatware implies progress). Seriously though, the number of books can be an indication of where the publishers and bookstores think the market is going. And out of the masses of dimwits might come some geniuses who will create killer apps. I guess it could be considered a statistical approach to success. Todd
From: id_est@interport.net (tse_di) Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.mac.programmer Subject: Re: How mature is Carbon? Date: Mon, 01 Jun 1998 23:49:57 -0400 Organization: none Message-ID: <id_est-0106982349570001@192.168.1.1> References: <01bd8cc5$31a42990$04387880@test1> <forrestDELETE-THIS!-3105981358070001@term1-10.vta.west.net> <tbrown-0106980111430001@mv131.axom.com> In article <tbrown-0106980111430001@mv131.axom.com>, tbrown@netset.com (Ted Brown) wrote: > In article <forrestDELETE-THIS!-3105981358070001@term1-10.vta.west.net>, > forrestDELETE-THIS!@west.net (Forrest Cameranesi) wrote: > > >Carbon is merely a subset of the Mac Toolbox, around 6000 calls (2000 less ....... > >MacOS, just without all the modern buzzwords. Chances are that some apps > >are already 100% Carbon-ready, although those chances are rather slim Well Given that Carbon requires navigation services in place of of traditional file services I think that means that every app is less that 100% carbon ready (baring that one Nintendo emulator)
From: quinlan@intergate.bc.ca Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.mac.programmer Subject: Re: How mature is Carbon? Date: Tue, 02 Jun 1998 09:36:24 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6l0h2o$u19$1@nnrp1.dejanews.com> References: <01bd8cc5$31a42990$04387880@test1> <forrestDELETE-THIS!-3105981358070001@term1-10.vta.west.net> <tbrown-0106980111430001@mv131.axom.com> <id_est-0106982349570001@192.168.1.1> In article <id_est-0106982349570001@192.168.1.1>, id_est@interport.net (tse_di) wrote: > Well Given that Carbon requires navigation services in place of of > traditional file services I think that means that every app is less that > 100% carbon ready (baring that one Nintendo emulator) I believe that it hasn't been decided yet if the current Standard*File calls are going to be removed or mearly devalued. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: quinlan@intergate.bc.ca Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Tue, 02 Jun 1998 09:34:41 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6l0gvh$tki$1@nnrp1.dejanews.com> References: <6kgu3m$nc0$10@ns3.vrx.net> <B193753E-2A1857@192.168.128.1> <6km2m1$s3m$5@ns3.vrx.net> <356EE7C0.2F1CF0FB@ctron.com> <6kn7ci$q5r$1@news.digifix.com> <6kt34d$h4l$1@nnrp1.dejanews.com> <6kt7c7$13g$1@news.digifix.com> <6ktsqk$dv7$1@nnrp1.dejanews.com> <6kv2lc$klv$1@news.digifix.com> <6kvm4h$qms$1@nnrp1.dejanews.com> <6kvqvu$t9d$1@news.digifix.com> In article <6kvqvu$t9d$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: > I'm not addressing "should" or "will". You stated that "If they're going to learn a whole new framework, why not learn one that will allow them to support both Windows and MacOS X." so you are suggesting, if I understand you correctly, that people who are porting to a new Windows framework should learn YB. That is what they "should" do. It is not necessarily what they "will" do. So giving these developers a solution that they will use might be better than giving them a superior solution that they won't use. > Apple's commited to it.. unfortunately they can't really do > much to change the 'perception' that some people have about YB short > of just shipping it. Exactly. And even shipping YB will not convince a lot of people. Apple has shipped a lot of products and then killed them. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: someone@somewhere.com (Clarence Locke) Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.mac.programmer Subject: Re: How mature is Carbon? Date: Tue, 02 Jun 1998 09:21:47 -0400 Organization: Dr Solomon's Software Message-ID: <someone-0206980921470001@38.176.138.44> References: <01bd8cc5$31a42990$04387880@test1> <forrestDELETE-THIS!-3105981358070001@term1-10.vta.west.net> <tbrown-0106980111430001@mv131.axom.com> <id_est-0106982349570001@192.168.1.1> <6l0h2o$u19$1@nnrp1.dejanews.com> > I believe that it hasn't been decided yet if the current Standard*File calls > are going to be removed or mearly devalued. Apple stated that an attempt will be made to map the Standard File calls to Navigation Services calls via a small piece of stub source code that will be made available to developers. The message was clear: Standard File is not going to be available in MacOS X.
From: Bill Bumgarner <bbum@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Tue, 2 Jun 1998 09:31:06 -0400 Organization: Spacelab.net Internet Access Message-ID: <Pine.NXT.3.96.980602092755.10028D-100000@pathos> References: <6kgu3m$nc0$10@ns3.vrx.net> <B193753E-2A1857@192.168.128.1> <6km2m1$s3m$5@ns3.vrx.net> <356EE7C0.2F1CF0FB@ctron.com> <6kn7ci$q5r$1@news.digifix.com> <6kt34d$h4l$1@nnrp1.dejanews.com> <6kt7c7$13g$1@news.digifix.com> <6ktsqk$dv7$1@nnrp1.dejanews.com> <6kv2lc$klv$1@news.digifix.com> <6kvm4h$qms$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII To: quinlan@intergate.bc.ca In-Reply-To: <6kvm4h$qms$1@nnrp1.dejanews.com> Not sufficiently committed? The Carbon demo of PhotoShop was running ON YELLOWBOX! Mac OS X is RHAPSODY WITH A REFINED AND TRANSPARENT MAC OS API [Carbon]. Everything Apple has announced in terms of Mac OS X has been market savvy in that it preserves the existing application base (or, at least, provides a migration path) and is technologically sound in that it puts the really killer Mach/BSD/YB environments as the muscle behind the marketing message. How does THAT not demonstrate a clear committment? Maybe because I spent the last eight+ years programming NeXTSTEP/OpenStep/Rhapsody and NOT watching Apple decide then undecide, I'm not scarred and see this as a committment that isn't going away.... :-) b.bum On Tue, 2 Jun 1998 quinlan@intergate.bc.ca wrote: > In article <6kv2lc$klv$1@news.digifix.com>, > sanguish@digifix.com (Scott Anguish) wrote: > > > Its not rational. > > > > If they're going to learn a whole new framework, why not learn > > one that will allow them to support both Windows and MacOS X? > > Scott, you're missing the different between "should" and "will". People don't > always make the rational choice. > > Also, I believe, people perceive Apple to not have sufficiently commitment to > YB to make porting to YB a safe bet. > > -----== Posted via Deja News, The Leader in Internet Discussion ==----- > http://www.dejanews.com/ Now offering spam-free web-based newsreading > >
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 2 Jun 1998 08:58:44 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6l0es4$8th$4@ns3.vrx.net> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku7hc$g29$2@ns3.vrx.net> <rex-0106981800560001@192.168.0.3> <6kuml2$rhg$1@ns3.vrx.net> <rex-0106981924540001@192.168.0.3> <6kuucg$2uf$2@ns3.vrx.net> <rex-0106982229380001@192.168.0.3> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: rex@smallandmighty.com In <rex-0106982229380001@192.168.0.3> Eric King claimed: > Not quite, the OS finds the binary through a mime type embedded in the > BMessage. Ahhhhhh. So then is it safe to assume that the purpose of ths shelf is to map this, load the binary and then provide a space for the app to run? > Though, I suppose it wouldn't be hard to cram the binary into > the message also, which might make sense if you were sending one over a > network. But there would be security issues to deal with. Sure. The big problem with this approach is that the data typically only makes sense to the binary that created it. Unless Be has included some form of data interchage standard (and I can't _find_ a mention of one on their [terribly useless] web pages) then the problem is that you always have to map back to the original binary, or the MIME mapped type. > Yep, that's pretty much the way it works. Some clever folks are using > it to patch the parts of the OS, since replicants run in their hosts > address space. Hmmm. What exactly is the mechanism? If I put one into my app, does it simply dynaload the other binary? How do they partition the app to get reasonable load times and footprints? > Provide a convenient user interface item. All you do is add a BShelf as > a child of your view and you're ready to recieve replicants. It handles > all of the instantiation work for you. Ok, that makes sense. This model would map easily onto YB as well, although you'd do it via some sort of NSProxyView. The problem is that apps are still "headed", in that the NSApp supports the event system and splitting that to another instance is not easy. > The background window of the Tracker is a big BShelf. Drop a replicant > on their and it starts to run. It's all done through shared libraries, > mime types, and BMessages. It's simple and works remarkably well. Can _any_ app become support for a BReplicant? Or do you have to spin off that support into a shared lib? > But I can see instances where Be's solution makes more sense. Since > their OS is so heavily threaded, it makes a sense to directly use > thinly-wrapped kernel messages to communicate. I doubt this is actually what it does. I'll bet that it's a coded message (ie, run NSCoder over a method call) that is then passed to a cue. The kernel likely has no more to do with their runtime (directly that is) than it does under YB. > This is probably what PDO is doing behind the scenes. Not from what I understand about it, it does just that, codes the invocation, and does what it always does - sends it to another object. Unknown to it the other object is an NSProxy subclass that does "something", like send it to another machine. This is a side effect of Obj-C's ability to send any message to any object, I'm not at all sure how to duplicate that _model_ under C++. > Actually, I looked and saw GX also. There's a one to one mapping > between many of the objects. For instance, instead of Tag objects, you > have bundles. I believe they even tried to support GX's font technology > and line layout. Sure, but they're doing the same thing for EQD. > The Postscript way is to use little programs or display lists to draw > shapes. Only if you draw in PS directly, a OOPS layer hides all of this (I know this, this is what I do for a living). The reason I say it looks like PS is the conceptual model of the API is one similar to PS. > I am. It just wouldn't help the big developers much *coff*. It would have reduced my project's timeline likely buy 1/4. More to the point it would make it more portable. Maury
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.mac.programmer Subject: Re: How mature is Carbon? Date: 2 Jun 1998 09:00:18 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6l0ev2$8th$5@ns3.vrx.net> References: <01bd8cc5$31a42990$04387880@test1> <forrestDELETE-THIS!-3105981358070001@term1-10.vta.west.net> <tbrown-0106980111430001@mv131.axom.com> <id_est-0106982349570001@192.168.1.1> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: id_est@interport.net In <id_est-0106982349570001@192.168.1.1> tse_di claimed: => Well Given that Carbon requires navigation services in place of of > traditional file services I think that means that every app is less that > 100% carbon ready (baring that one Nintendo emulator) Umm, doesn't it support SFP? Not StandardFile, but the newer version. Under OS8.5 there are _three_ file nav systems, under Carbon I understood there would be only two. Indeed the port of PhotoShop they demoed used the OS8-a-like one. Maury
From: jnutting@my-dejanews.com Newsgroups: comp.sys.next.programmer Subject: NSPopUpButton within NSTableView Date: Tue, 02 Jun 1998 14:15:03 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6l11d6$hd1$1@nnrp1.dejanews.com> Does anyone have an example of how to do this? I knew how to do it with NXTableView and the old popup, but NSTableView doesn't use a formatter anymore, and the new popup is a whole new ball game as well. Any clues would help. -- // Jack Nutting // jnutting"at"rebisoft.com // www.rebisoft.com // -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: NSPopUpButton within NSTableView Date: 2 Jun 1998 14:44:56 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6n83vc.9bh.rog@talisker.ohm.york.ac.uk> References: <6l11d6$hd1$1@nnrp1.dejanews.com> On Tue, 02 Jun 1998 14:15:03 GMT, jnutting@my-dejanews.com <jnutting@my-dejanews.com> wrote: > Does anyone have an example of how to do this? I knew how to do it > with NXTableView and the old popup, but NSTableView doesn't use a > formatter anymore, and the new popup is a whole new ball game as > well. Any clues would help. as far as i'm aware this is now impossible, as you can't get at the popup functionality without its associated popup button, which is a subclass of NSView and therefore can't be used anywhere that needs an NSCell. i tried to find a way of doing this some time ago and failed. if you succeed, i'd love to know how! cheers, rog.
From: "Jeremy Bettis" <jeremy@hksys.com> Newsgroups: comp.sys.next.programmer Subject: Re: NSPopUpButton within NSTableView Date: Tue, 2 Jun 1998 10:30:00 -0500 Organization: Internet Nebraska Message-ID: <6l15q3$4ic$1@owl.inetnebr.com> References: <6l11d6$hd1$1@nnrp1.dejanews.com> <slrn6n83vc.9bh.rog@talisker.ohm.york.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit -----BEGIN PGP SIGNED MESSAGE----- >On Tue, 02 Jun 1998 14:15:03 GMT, jnutting@my-dejanews.com <jnutting@my-dejanews.com> wrote: >> Does anyone have an example of how to do this? I knew how to do it >> with NXTableView and the old popup, but NSTableView doesn't use a >> formatter anymore, and the new popup is a whole new ball game as >> well. Any clues would help. > >as far as i'm aware this is now impossible, as you can't get at the >popup functionality without its associated popup button, which is a >subclass of NSView and therefore can't be used anywhere that needs an >NSCell. make a NSPopUpButton, configure it the way you like, and then steal it's cell. -----BEGIN PGP SIGNATURE----- Version: PGP 5.5.5 iQBVAwUBNXQadxmtiimFYrNNAQHgHgIAqYuz9HmF263TSqinsvxaflOWkbGTxNUO k7PzD+21NaWfUhl3WjodFOLtkasJzHyPBWvxKhlWQ98m9Pba1JMWEw== =JLBD -----END PGP SIGNATURE-----
From: cb@bartlet.df.lth.se (Christian Brunschen) Newsgroups: comp.sys.next.programmer Subject: Keyboard input & multiple-button mouse events Date: 2 Jun 1998 17:19:29 +0200 Organization: The Computer Society at Lund Message-ID: <6l1561$tfn$1@bartlet.df.lth.se> NNTP-Posting-User: cb Hi everyone, I am in the process of writing a Vnc client for OpenStep (doing all the work on my trusty old NeXTcube). For those who don't know, Vnc stands for 'Virtual Network Computer' and provides platform-independant remote display access - for full info, please look at <http://www.orl.co.uk/vnc/> I have two major problems right now. First problem: ============== Anyway, I need to intercept keyboard events and generate the corresponding messages for the Vnc protocol. Basically, in my NSView subclass I override - (void) keyDown:(NSEvent *)theEvent (and of course keyUp: as well), and things work quite well - I get all the keyboard events that the Application doesn't handle otherwise (command key equivalents basically), and I am successful at sending this to the remote display - things work. However, when trying to compuse characters - say, using 'Alt-u a' to generate 'ä' - I get the diaeresis character and then the 'a' character, each in separate events. Hm; not quite good. What to do, what to do. So I tried sending every character that I do not handle separately, to [self interpretKeyEvents:[NSArray arrayWithObjects:theEvent, nil]] in the hope that this would somehow compose the characters in question together and then hand them to me by calling - insertText:(NSString *)text but this doesn't work better - I still get each character one at a time. I have some other things I'm going to try - ie, checking each character for [[NSCharacterSet nonBaseCharacterSet] characterIsMember:theChar] , push all the non-base chars into a string, and only when I get a base char, push it in with the non-base chars ... but the question still remains, how do I get something in the system to coalesce my combination of modifiers (non-base chars) and the base char in question, into the closest character available ? I guess I could ask the string of (modifiers + base char) to be converted to ISOLatin1 encoding, but .... Anyone who can offer helpful hints, please do so :) Second problem: =============== OpenStep 'only' supports 2-button mice. I would like to include support for emulating a three-button mouse - by treating the two physical buttons as mouse buttons 1 and 3, and intercepting if you press _both_ mouse buttons to mean that you have pressed the 'middle' mouse button. However, it seems the event queue is 'serialized' to only handle events on one mouse button at a time; ie, if I press both buttons, I get the (right)mouseDown event corresponding to whichever mouse button I happened to hit first, then a series of corresponding (right)mouseDragged events, and finally a (right)mouseUp - and only then do I get the down & up events for the _other_ mouse button. Does anyone have any input on this ? Should I 'roll my own' event loop perhaps ? (I'd rather not, for obvious reasons.) Anyway, I am planning an alpha release of a leaking-memory-like-a-sieve-but-usable-for-casual-work versino fo my Vnc client. I do not have a server for it available, or for that matter in the works. One thing here: There is a Vnc server which is also an X server - ie, applications talk to it like they'd talk to any X server, but instead of drawing to a screen it draws to memory, and Vnc clients can then connect and talk to this 'virtual X server' - based on the XFree86 3.3.2 source tree. If someone could take a look at that and make it compile under OpenStep, there would be a semi-low-performance, but completely free, X server solution available for OpenStep to complement the shareware native port of X11R6.3 . I have no experience with something like that, and not much time either, so finishing the Vnc viewer is a higher priority for me, but perhaps there is someone reading this newsgroup who could take a look at it ? It does _not_ involve any hardware interfacing, or OpenStep programming afaik, since the 'hardware' it talks to is a memory buffer ... Best regards // Christian Brunschen
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: Re: DTS sample: Son of SillyBalls Message-ID: <1998060217570900.NAA15132@ladder03.news.aol.com> Date: 02 Jun 1998 17:57:09 GMT Organization: AOL http://www.aol.com References: <6kv9al$ku0$1@news.idiom.com> -jcr may or may not have said: :It's going at whatever rate the NSTimer is telling it to go. Right, but I want to skip the timer entirely. :Depends on what your goal is. Using an NSTimer, you can make the balls :appear at whatever rate you want, basically. Just set the desired interval :on the NSTimer. BTW, you *really*don't want the -drawRect: method of an :NSView to loop indefinitely. Becuase the view never updates the screen, right? -> Any of the NS experienced have an insight on a cooler, faster way? :If you just want to see those balls getting drawn the fastest way possible, :then you use a non-retained window, and draw the balls in a loop that quits :when it sees a mousedown event. I dug around some and I couldn't figure out what a non-retained window is or how to make one. Can you point me at the documentation I need to read? :-jcr
From: embuck@palmer.cca.rockwell.com (Erik M. Buck) Newsgroups: comp.sys.next.programmer Subject: Re: DTS sample: Son of SillyBalls Date: 2 Jun 1998 19:03:04 GMT Organization: Rockwell International Message-ID: <6l1i98$o0t1@onews.collins.rockwell.com> References: <6kv9al$ku0$1@news.idiom.com> <1998060217570900.NAA15132@ladder03.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: larrysb@aol.com In <1998060217570900.NAA15132@ladder03.news.aol.com> LarrySB wrote: > -jcr may or may not have said: > > > :It's going at whatever rate the NSTimer is telling it to go. > > Right, but I want to skip the timer entirely. > > > :Depends on what your goal is. Using an NSTimer, you can make the balls > :appear at whatever rate you want, basically. Just set the desired interval > :on the NSTimer. BTW, you *really*don't want the -drawRect: method of an > :NSView to loop indefinitely. > > Becuase the view never updates the screen, right? > You are free to add flushWindow calls if you want to flush to screen arbitrarily. See the source code to Backspace for and example of looping in -drawRect: > -> Any of the NS experienced have an insight on a cooler, faster way? Add a method - (void)runStep:(id)sender { [NSObject cancelPreviousPerformRequestsWithTarget:self selector:@selector(runStep:) object:nil]; [self performSelector:@selector(runStep:) withObject:nil afterDelay:0]; [self display]; } Call -runStep: once at the start of your application and it will continuously display as fast as it can. > > :If you just want to see those balls getting drawn the fastest way possible, > :then you use a non-retained window, and draw the balls in a loop that quits > :when it sees a mousedown event. > > I dug around some and I couldn't figure out what a non-retained window is or > how to make one. Can you point me at the documentation I need to read? > > :-jcr > > You are not looking very hard for documentation. See the NSWindow class and or Interface Builder.
From: "Charles W. Swiger" <chuck@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Followup-To: comp.sys.next.advocacy Date: 2 Jun 1998 20:40:01 GMT Organization: Spacelab.net Internet Access Message-ID: <6l1nv1$d7k$1@news.spacelab.net> References: <6kp9sv$cdl$1@crib.corepower.com> <1998053118593100.OAA15861@ladder01.news.aol.com> larrysb@aol.com (LarrySB) wrote: >Can anyone name 3 sku's of shrink-wrap software, on the shelves and available >at your local Fry's, CompUSA, Computer City, MicroCenter, etc, which was >written in OpenStep? > >You guys can sit here and derride obviously clunky Win32/MFC all you want. You >can even make it scary with statements like "eat your lunch". Larry, your article had no content relevant to comp.sys.next.programmer. Take it to a .advocacy group, please... -Chuck Charles Swiger | chuck@codefab.com | standard disclaimer ---------------+-------------------+-------------------- "Microsoft: we make the easy almost impossible."
From: yumi8 <yumi8@hotmail.com> Newsgroups: comp.sys.next.programmer Subject: anyone want a MediaStation 1.5 manual? Date: Tue, 02 Jun 1998 20:22:43 -0400 Organization: Ball State University Message-ID: <35749753.678BAFFF@hotmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I have a Next MediaStation 1.5 manual... sitting there. If you want it, make me an offer, I have no use for it email at: yumi8 at hotmail dot com
From: quinlan@intergate.bc.ca Newsgroups: comp.sys.next.programmer,comp.sys.mac.advocacy,comp.sys.mac.programmer Subject: Re: How mature is Carbon? Date: Tue, 02 Jun 1998 20:07:56 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6l1m2s$jo3$1@nnrp1.dejanews.com> References: <01bd8cc5$31a42990$04387880@test1> <forrestDELETE-THIS!-3105981358070001@term1-10.vta.west.net> <tbrown-0106980111430001@mv131.axom.com> <id_est-0106982349570001@192.168.1.1> <6l0h2o$u19$1@nnrp1.dejanews.com> <someone-0206980921470001@38.176.138.44> In article <someone-0206980921470001@38.176.138.44>, someone@somewhere.com (Clarence Locke) wrote: > Apple stated that an attempt will be made to map the Standard File calls > to Navigation Services calls via a small piece of stub source code that > will be made available to developers. The message was clear: Standard > File is not going to be available in MacOS X. Are they going to put this code into the universal headers or do you have to include this code as source yourself? The former would be smarter. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: axel.steininger@uibk.ac.at (Axel Steininger) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 03 Jun 1998 03:40:09 +0200 Organization: q.consult Message-ID: <axel.steininger-0306980340130001@ip02port19.highway.telekom.at> References: <6kks1f$95j$1@crib.bevc.blacksburg.va.us> <B19382E0-2D4CAC@192.168.128.1> In article <B19382E0-2D4CAC@192.168.128.1>, "Baskaran Subramaniam" <baskar@RemoveMe.snet.net> wrote: > On Thu, May 28, 1998 7:29 PM, Nathan Urban > <mailto:nurban@crib.bevc.blacksburg.va.us> wrote: > >In article <B19366A7-26AA48@192.168.128.1>, "Baskaran Subramaniam" < > >baskar@RemoveMe.snet.net> wrote: > > >> Sure it is easy for Nathan to tell the current MacOS developers to > forget > >> about their cash cows (currently shipping MacOS software) and start > >> rewriting their code using the YB API and in the process abandon the > >> current MacOS market. But that is complete nonsense. > > >You're right. Can we say "strawman"? I knew we could. > > I am not sure, what you mean here. Sorry :-( > > >I never suggested anything of the sort. The issue in this thread has > >always been: which is better for _new_ development? For someone starting > >from scratch with no MacOS code to leverage, YB is more compelling. > > Well, as a long time Mac developer, I don't like the idea of abandoning > the Mac installed base and start developing for Windows. Applications > written using YB API will run on all Intel machines than can run > Windows, but only on a puny fraction of the Mac hardware. > To even suggest that I do this, Apple (oops... I mean NeXT) must be > smoking something. > > Baskaran Who is talking about porting current MacOS apps to YB anyway? Most of the big MacOS app developers (Adobe, Macromedia, Quark, Microsoft, MetaCreations, etc...) already have a Windows version of their programs. They will update their apps to the Carbon API as soon as it's available and will then make new versions of their apps available to MacOS and Windows users simultaneously just like they did for the last couple of years. The real promise of YB is that developers with NEW apps who want to target a large part of the market finally have an alternative to learning Visual C++ or other Windows developer tools. OpenStep has always been very expensive, "elite" and exotic. So far OpenStep has only been an option for creators of specialized apps for a few enlightened companies that were lucky enough to have an IT professional who recommended OpenStep. But now I can imagine a lot of MacOS shareware programmers and small software companies that don't have to abandon a huge MacOS code base to switch to YB because of it's power and compatibility. Eventually they will spread the word in the Windows developer world as well. Having been bought by Apple was the best thing that could happen to OpenStep (not necessarily NextStep). Apple's PowerMac G3s are selling like hot cakes and I could imagine that a large part of the Mac world will have migrated at least to the current generation of PMacs by late 99. Those that haven't bought a new machine by then probably aren't interested in a new OS anyway. Any professional Mac user who doesn't upgrade to the current line of PowerMacs will certainly upgrade to the next line with G4 CPUs (like myself). And they will also migrate to MacOS X because this will be the only OS Apple will preinstall on their machines from late 99 on. So I really don't see a problem with the installed base for YB apps. Most Win users are ready for it now and most (serious) Mac users will be ready for it from late 99 on.
From: rex@smallandmighty.com (Eric King) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Wed, 03 Jun 1998 00:25:07 -0400 Organization: The Small & Mighty Group Message-ID: <rex-0306980025070001@192.168.0.3> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku7hc$g29$2@ns3.vrx.net> <rex-0106981800560001@192.168.0.3> <6kuml2$rhg$1@ns3.vrx.net> <rex-0106981924540001@192.168.0.3> <6kuucg$2uf$2@ns3.vrx.net> <rex-0106982229380001@192.168.0.3> <6l0es4$8th$4@ns3.vrx.net> In article <6l0es4$8th$4@ns3.vrx.net>, maury@remove_this.istar.ca (Maury Markowitz) wrote: :In <rex-0106982229380001@192.168.0.3> Eric King claimed: :> Not quite, the OS finds the binary through a mime type embedded in the :> BMessage. : : Ahhhhhh. So then is it safe to assume that the purpose of ths shelf is to :map this, load the binary and then provide a space for the app to run? The BShelf's main purpose is to accept drag BMessages and instantiate any replicants in them. The actual instantiation process is handled entirely by the OS. : Sure. The big problem with this approach is that the data typically only :makes sense to the binary that created it. Yep. Though you can easily define BMessaging protocols. For instance, there's one for passing color values between apps. : Unless Be has included some form :of data interchage standard (and I can't _find_ a mention of one on their :[terribly useless] web pages) There data interchange standard is the BMessage. They are used all over the place. : Hmmm. What exactly is the mechanism? If I put one into my app, does it :simply dynaload the other binary? Yup. It calls the replicants instantiate method, which reconstitutes. If you look closely at the Be API, just all GUI of the elements have constructors that take BMessages. A class just needs to inherit from BArchivable and implement a couple of methods for it to become a replicant. : How do they partition the app to get reasonable load times and footprints? Everything's done through shared libraries. In the BeOS any executable that exports the appropriate symbols can be used as a shared library, i.e. there's no significant difference between a shared library and an application. : Ok, that makes sense. This model would map easily onto YB as well, :although you'd do it via some sort of NSProxyView. :The problem is that apps :are still "headed", in that the NSApp supports the event system and splitting :that to another instance is not easy. Why would you need another instance of NSApp? : Can _any_ app become support for a BReplicant? Or do you have to spin off :that support into a shared lib? It's just a matter of exporting the appropriate symbols and attaching a small resource. It's not a complex process. : I doubt this is actually what it does. I'll bet that it's a coded message :(ie, run NSCoder over a method call) that is then passed to a cue. The :kernel likely has no more to do with their runtime (directly that is) than it :does under YB. Nope BMessages, really are just thin wrappers over BeOS kernel messages. It's the most convenient and least dangerous way for threads to communicate with each other. :> This is probably what PDO is doing behind the scenes. : : Not from what I understand about it, it does just that, codes the :invocation, and does what it always does - sends it to another object. If that other object happens to be a thread or in another address space, it would make a lot of sense to use a Mach port to send the data. This is the sort of thing ports are meant for. : This is a side effect of Obj-C's ability to :send any message to any object, I'm not at all sure how to duplicate that :_model_ under C++. C++ is so low-level, I think that you could probably *could* do it. As shown by David Stes' POC, Objective-C can be compiled down to normal C. I'm sure you could wrap the raw C calls up in C++ to make them look prettier. : Sure, but they're doing the same thing for EQD. : :> The Postscript way is to use little programs or display lists to draw :> shapes. : : Only if you draw in PS directly, Well that *is* the Postscript way. Because of a lot of work on the part of Next, and now Apple, it's not the Appkit way, but the Appkit != Postscript. : a OOPS layer hides all of this (I know :this, this is what I do for a living). But then you're really comparing the Taligent framework with the OO layer on top of Postscript, not Postscript itself. Postscript is a procedural language that includes immediate mode graphics operators. For instance a path isn't an 'object' its a little program. This is nothing at all like the classes in CommonPoint, or the pseudo-classes in QD 3D and GX. However, it is somewhat like a display list in OpenGL. :The reason I say it looks like PS is :the conceptual model of the API is one similar to PS. I think you're muddying things. The floating-point coordinate space is similar but really only one fairly minor issue. There are much larger issues here like how scene state is maintained, object models, support for metadata, choice of curve bases, color standards, etc. That said there are some Appkit classes, most notably NSBezierPath, that are similar, but NSBezierPath is *not* a part of Postscript. In fact, it flies in the face of the immediate mode nature of Postscript's graphics operators. :> I am. It just wouldn't help the big developers much : : *coff*. It would have reduced my project's timeline likely buy 1/4. More :to the point it would make it more portable. But it would also make it easier for smaller developers to compete with Adobe and the other large developers. This is not helping big developers, it's hurting. -Eric
From: sarid@particle.phys.nd.edu (Uri Sarid) Newsgroups: comp.sys.next.programmer Subject: gcc-2.8.x ported to nextstep 3.3 on intel? Date: 2 Jun 1998 23:52:01 GMT Organization: University of Notre Dame Message-ID: <6l2371$9oa@news.nd.edu> I'm trying to get gcc-2.8.x (I'd like x as high as possible) running on my nextstep 3.3 running on a pentium. I haven't been able to find binaries, so I downloaded the 2.8.0 source from mit and tried to build it. It's running into trouble: the config.log file reads configure:713: checking host system type configure:734: checking target system type configure:752: checking build system type configure:779: checking for gcc configure:808: checking for cc configure:856: checking whether the C compiler (cc ) works configure:870: cc -o conftest conftest.c 1>&5 ld: can't locate file for: -lcrt0.o configure: failed program was: #line 866 "configure" #include "confdefs.h" main(){return(0);} but, not having run cc in the past, I have no idea what's wrong. Does anybody know if there's a set of binaries (inluding the libraries) prepackaged for NS3.3FIP that I can just download and install, without doing a build? Thanks! Uri (sarid@particle.phys.nd.edu) -- Uri Sarid Department of Physics office: (219) 631-6823 University of Notre Dame fax: (219) 631-5332 Notre Dame, IN 46556 (if fax fails, try -5952) e-mail: sarid@particle.phys.nd.edu http://www.phys.nd.edu/physics/ http://www.phys.nd.edu/physics/faculty/sarid.html http://www.nd.edu/~sarid/ (personal home page)
From: larrysb@aol.com (LarrySB) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Message-ID: <1998060306023100.CAA09914@ladder03.news.aol.com> Date: 03 Jun 1998 06:02:31 GMT Organization: AOL http://www.aol.com References: <axel.steininger-0306980340130001@ip02port19.highway.telekom.at> Charles Swiger | chuck@codefab.com | squeaked: :Larry, your article had no content relevant to :comp.sys.next.programmer. :Take it to a .advocacy group, please... :-Chuck So who appointed you the net cop? Almost every damn message in the whole thread belongs in an advocacy newgroup. If you want a moderated board, go make one, or sign up to one of the mailing lists. -- Dr. Nuketopia Read the Blue Glow in Tubes FAQ at http://www.persci.com/~larrysb This address gets lots-o-spam. Please note that your letter is *not* spam in the subject line.
From: arti@lava.DOTnet (Art Isbell - remove "DOT") Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 3 Jun 1998 08:08:49 GMT Organization: LavaNet, Inc. Distribution: world Message-ID: <6l30ah$2fj@mochi.lava.net> References: <axel.steininger-0306980340130001@ip02port19.highway.telekom.at> <1998060306023100.CAA09914@ladder03.news.aol.com> > Almost every damn message in the whole > thread belongs in an advocacy newgroup. Good point! I read this group for OPENSTEP and YB programming information, not for a discussion of Common Point, BeOS, MFC, C++, Visual Age, etc., etc. which seem to dominate this group lately. All of these topics have more appropriate groups. So please take them there. -- Art Isbell NeXT/MIME Mail: arti@lavaDOTnet IDX Systems Corporation Voice/Fax: +1 808 394 0511 (for whom I don't speak) Voice Mail: +1 808 394 0495 Healthcare Info Technology US Mail: Honolulu, HI 96825-2638
From: Samuel Boivin <Samuel.Boivin@inria.fr> Newsgroups: comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.programmer Subject: Bad Size for a 2GB HD on Cube Date: Wed, 03 Jun 1998 11:51:51 +0200 Organization: I.N.R.I.A. Message-ID: <35751CB7.FDFB0BBC@inria.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi everybody, Here is my problem, that has already been resolved, but I can't remember where ?!? I bought a Quantum 2GB Hard Disk Drive for my NeXT Cube(040/25/NS3.3), and it is well recognized. During boot, it is identified as a 2051MB Disk. When I use 'BuildDisk' to build it, it seems to be OK: the disk is identified as a 2051MB Disk and is correctly constructed(I mean: no error during construction). But the problem is: if I look at my Workspace, I can see: 840MB free !!! The disk is recognized as a 1GB !! This is confirmed by: if I use 'df /dev/rsd1a' under Terminal.app, it gives a size of 1GB, 160MB occuppied, 16%full, etc... So my questions is the following: How to force my cube to correctly initialize my 2GB HD, in order to use it under Workspace or Terminall.app as a REAL 2GB Hard Disk drive ? Thank you for your future help !! Sam. PS: I tried to build 2 partitions of 1GB, but nothing changed ! I only have ONE 1GB partition and I can't find the second !! Where is it ? -- ||===================================================================|| || Samuel Boivin || || I.N.R.I.A - Rocquencourt || || Tel: 01-39-63-51-86 Fax: 01-39-63-57-71 || || http://www-syntim.inria.fr/syntim/recherche/boivin/index-eng.html || ||===================================================================||
From: rdieter@math.unl.edu (Rex Dieter) Newsgroups: comp.sys.next.programmer Subject: Re: gcc-2.8.x ported to nextstep 3.3 on intel? Date: 3 Jun 1998 13:45:22 GMT Organization: University of Nebraska-Lincoln Message-ID: <6l3k1i$85e$1@unlnews.unl.edu> References: <6l2371$9oa@news.nd.edu> In article <6l2371$9oa@news.nd.edu> sarid@particle.phys.nd.edu (Uri Sarid) writes: > I'm trying to get gcc-2.8.x (I'd like x as high as possible) running > on my nextstep 3.3 running on a pentium. I've been maintaining a gcc port for nextstep, currently at version 2.7.2.3. I've been waiting for g77 to become available for gcc-2.8.x (since we use it here a lot) before doing a 2.8.x port for nextstep. Perhaps I should try out egcs (URL (?) http://egcs.cygnus.com/) which has g77 built-in. -- Rex A. Dieter rdieter@math.unl.edu (NeXT/MIME OK) Computer System Manager http://www.math.unl.edu/~rdieter/ Mathematics and Statistics University of Nebraska-Lincoln
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.pen,comp.s Subject: cmsg cancel <6kos5f$ssc$3@news.vossnet.de> Control: cancel <6kos5f$ssc$3@news.vossnet.de> Date: 03 Jun 1998 14:30:43 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6kos5f$ssc$3@news.vossnet.de> Sender: "Uwe Vogt" <tora@vossnet.de> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: William.Clocksin@CL.cam.ac.uk Newsgroups: comp.sys.next.programmer Subject: Query on Mach Messaging Date: 3 Jun 1998 14:30:42 GMT Organization: University of Cambridge, England Message-ID: <6l3mmi$ia7$1@lyra.csx.cam.ac.uk> I noticed in the demo example SortingInAction that Mach messages are constructed and sent to another thread. It is perfectly understandable, and it demonstrably works. However, because the message body is in the same place in memory used again and again, what prevents the sender from trashing a previous message with a new message before the receiver has handled the first message? If the messenging system passes a pointer, I can't see how the integrity of the message body can remain intact in a multi-threaded environment. The memory is not protected with a mutex as far as I can tell. Or is the message content copied out of memory when the message is sent (as opposed to when it is received, because the sender could trash it immediately after sending it)? And if it is copied, how much space does the kernel make available for message copies? Any explanation gratefully received. Thank you. William Clocksin Computer Laboratory University of Cambridge
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer Subject: Re: Query on Mach Messaging Date: 3 Jun 1998 15:12:42 GMT Organization: Technical University of Berlin, Germany Message-ID: <6l3p5a$q11$1@news.cs.tu-berlin.de> References: <6l3mmi$ia7$1@lyra.csx.cam.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit William.Clocksin@CL.cam.ac.uk writes: >I noticed in the demo example SortingInAction that Mach messages are >constructed and sent to another thread. It is perfectly understandable, and >it demonstrably works. However, because the message body is in the same >place in memory used again and again, what prevents the sender from >trashing a previous message with a new message before the receiver has >handled the first message? If the messenging system passes a pointer, I >can't see how the integrity of the message body can remain intact in a >multi-threaded environment. The memory is not protected with a mutex as far >as I can tell. Or is the message content copied out of memory when the >message is sent (as opposed to when it is received, because the sender >could trash it immediately after sending it)? And if it is copied, how >much space does the kernel make available for message copies? Any >explanation gratefully received. Thank you. From memory: - The mach kernel will copy the messages for you. - There is effectively no size limit for messages because data can be sent 'out-of-line' in which case the memory is mapped into the appropriate address space with 'copy-on-write' protection. - Small messages can be sent inline. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: fukuda@ifor17.ethz.ch (Komei Fukuda) Newsgroups: comp.sys.next.programmer Subject: Re: gcc-2.8.x ported to nextstep 3.3 on intel? MIME-Version: 1.0 Content-Type: text/plain; charset=iso-2022-jp Content-Transfer-Encoding: 7bit Cc: rdieter@math.unl.edu References: <6l2371$9oa@news.nd.edu> <6l3k1i$85e$1@unlnews.unl.edu> Message-ID: <35756b1a.0@epflnews.epfl.ch> Date: 3 Jun 98 15:26:18 GMT Organization: EPFL In <6l3k1i$85e$1@unlnews.unl.edu> Rex Dieter wrote: > In article <6l2371$9oa@news.nd.edu> sarid@particle.phys.nd.edu (Uri Sarid) > writes: > > I'm trying to get gcc-2.8.x (I'd like x as high as possible) running > > on my nextstep 3.3 running on a pentium. > > I've been maintaining a gcc port for nextstep, currently at version > 2.7.2.3. I've been waiting for g77 to become available for gcc-2.8.x > (since we use it here a lot) before doing a 2.8.x port for nextstep. > Perhaps I should try out egcs (URL (?) http://egcs.cygnus.com/) which has > g77 built-in. I have tried to build gcc-2.8.1 with libg++-2.8.1.1a without success. I have essentially the same error when I tried to build egcs-1.0.3a. I will attatch a log below. My nextstep 3.3J machine has a pentium, and I used gnu make. I will appreciate it very much if anyone teach me how to fix the problem. To my surprise, I could build egcs-1.0.3a on my black machines (Turbo and non-Turbo) without problems. So I happily use gcc-2.8.1 with STL libraries on beautiful original boxes :-) Komei ------ thinkblue> cd libstdc++-2.8.1.1 thinkblue> ls COPYING config/ include/ ltmain.sh COPYING.LIB config-ml.in install-sh* missing* ChangeLog config.guess* libiberty/ mkinstalldirs* INSTALL config.sub* libio/ move-if-change* Makefile.in configure* libstdc++/ symlink-tree* README configure.in ltconfig* texinfo/ thinkblue> mv ../libg++-2.8.1.1a g++ thinkblue> ls COPYING config-ml.in install-sh* mkinstalldirs* COPYING.LIB config.guess* libiberty/ move-if-change* ChangeLog config.sub* libio/ symlink-tree* INSTALL configure* libstdc++/ texinfo/ Makefile.in configure.in ltconfig* README g++/ ltmain.sh config/ include/ missing* thinkblue> mv ../gcc-2.8.1 gcc thinkblue> ls COPYING config-ml.in include/ missing* COPYING.LIB config.guess* install-sh* mkinstalldirs* ChangeLog config.sub* libiberty/ move-if-change* INSTALL configure* libio/ symlink-tree* Makefile.in configure.in libstdc++/ texinfo/ README g++/ ltconfig* config/ gcc/ ltmain.sh thinkblue> mkdir objdir thinkblue> cd objdir thinkblue> cd .. thinkblue> ls COPYING config-ml.in include/ missing* COPYING.LIB config.guess* install-sh* mkinstalldirs* ChangeLog config.sub* libiberty/ move-if-change* INSTALL configure* libio/ objdir/ Makefile.in configure.in libstdc++/ symlink-tree* README g++/ ltconfig* texinfo/ config/ gcc/ ltmain.sh thinkblue> patch <INSTALL Hmm... Looks like a unified diff to me... The text leading up to this was: . . . Patching file gcc/Makefile.in using Plan A... Hunk #1 succeeded at 415 (offset 4 lines). Hunk #2 succeeded at 560 (offset 4 lines). Hunk #3 succeeded at 2117 (offset 5 lines). Hunk #4 succeeded at 2401 (offset 4 lines). done thinkblue> cd objdir thinkblue> ../configure Configuring for a i386-next-nextstep3 host. Created "Makefile" in /superuser/Archive/libstdc++-2.8.1.1/objdir Links are now set up to build a native compiler for i386-next-nextstep3 thinkblue> ls Makefile config.status* libiberty/ texinfo/ config.cache gcc/ libraries/ thinkblue> cd gcc thinkblue> make bootstrap . . much deleted here (without errors) . ln -s ../../gcc/cplus-dem.c cxxmain.c cc -traditional-cpp -c -DMAIN -DIN_GCC -g -DHAVE_CONFIG_H -I. -I../../gcc -I../../gcc/config \ -DVERSION=\"2.8.1\" cxxmain.c In file included from ../../gcc/demangle.h:24, from cxxmain.c:35: ../../gcc/gansidecl.h:86: warning: `bcopy' redefined /NextDeveloper/Headers/ansi/string.h:119: warning: this is the location of the previous definition ../../gcc/gansidecl.h:90: warning: `bzero' redefined /NextDeveloper/Headers/ansi/string.h:125: warning: this is the location of the previous definition ../../gcc/gansidecl.h:94: warning: `bcmp' redefined /NextDeveloper/Headers/ansi/string.h:122: warning: this is the location of the previous definition ../../gcc/gansidecl.h:98: warning: `rindex' redefined /NextDeveloper/Headers/ansi/string.h:116: warning: this is the location of the previous definition ../../gcc/gansidecl.h:102: warning: `index' redefined /NextDeveloper/Headers/ansi/string.h:112: warning: this is the location of the previous definition rm -f cxxmain.c echo "int xxy_us_dummy;" >tmp-dum.c ./xgcc -B./ -S tmp-dum.c tmp-dum.c:2: parse error at null character make[1]: *** [s-under] Error 1 make[1]: Leaving directory `/superuser/Archive/libstdc++-2.8.1.1/objdir/gcc' make: *** [bootstrap] Error 2 ---
From: Komei.Fukuda@epfl.ch (Komei Fukuda) Newsgroups: comp.sys.next.programmer Subject: Re: gcc-2.8.x ported to nextstep 3.3 on intel? MIME-Version: 1.0 Content-Type: text/plain; charset=iso-2022-jp Content-Transfer-Encoding: 7bit Cc: Komei.Fukuda@epfl.ch References: <6l2371$9oa@news.nd.edu> <6l3k1i$85e$1@unlnews.unl.edu> <35756b1a.0@epflnews.epfl.ch> Message-ID: <3575768f.0@epflnews.epfl.ch> Date: 3 Jun 98 16:15:11 GMT Organization: EPFL Sorry, I used a wrong e-mail address in my previous posting. Just to add some real content, my black next machine returns: -------- manext2% g++ -v Reading specs from /usr/local/lib/gcc-lib/m68k-next-nextstep3/egcs-2.90.27/specs gcc version egcs-2.90.27 980315 (egcs-1.0.2 release) manext2% g77 -v Reading specs from /usr/local/lib/gcc-lib/m68k-next-nextstep3/egcs-2.90.27/specs gcc version egcs-2.90.27 980315 (egcs-1.0.2 release) -------- I am not using fortran but the compiler seems to be working. I use g++ for Rational (infinite precision) arithmetic, and the new gcc compiler is working well for this purpose. Yet, executables tend to be much larger than those created by versions 2.7.x. I do not know why, but I am happy to be able to use C++ standard templates now. Komei Fukuda IFOR, ETH Zentrum, CH-8092 Zurich, Switzerland fax +41-1-632 1025 tel +41-1-632 4023 or 4028 email:fukuda@ifor.math.ethz.ch http://www.ifor.math.ethz.ch/staff/fukuda/fukuda.html Address in Lausanne: DMA, EPFL, CH-1015 Lausanne, Switzerland fax +41-21-693 55 70 tel +41-21-693 25 36 email : Komei.Fukuda@epfl.ch
From: embuck@palmer.cca.rockwell.com (Erik M. Buck) Newsgroups: comp.sys.next.programmer Subject: Re: Query on Mach Messaging Date: 3 Jun 1998 16:29:05 GMT Organization: Rockwell International Message-ID: <6l3tkh$ntc2@onews.collins.rockwell.com> References: <6l3mmi$ia7$1@lyra.csx.cam.ac.uk> <6l3p5a$q11$1@news.cs.tu-berlin.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: marcel@cs.tu-berlin.de In <6l3p5a$q11$1@news.cs.tu-berlin.de> Marcel Weiher wrote: > From memory: > > - The mach kernel will copy the messages for you. > - There is effectively no size limit for messages > because data can be sent 'out-of-line' in which > case the memory is mapped into the appropriate > address space with 'copy-on-write' protection. > - Small messages can be sent inline. > > Marcel > Mach messages are always copy-on-write and thus have the equivalent of a mutex provided automatically by the kernel. Mach messages can also be distributed over a network. However, just because NeXT provided this example does not mean that was the best way to do the job. I think NeXT just wanted an excuse do demo Mach messages. Are there any other demos of messages ? A better way to solve that particular problem is with an NSData instance shared between threads. The locking is transparent; the implementation is simpler, and the performance is probably even better due to fewer kernel calls. Of course I could be wrong. This topic is way outside my expertise.
From: satoru@candenext.lsa.berkeley.edu Newsgroups: comp.sys.next.programmer Subject: Re: gcc-2.8.x ported to nextstep 3.3 on intel? Date: 3 Jun 1998 17:36:11 GMT Organization: University of California, Berkeley Message-ID: <6l41ib$ond$1@agate.berkeley.edu> References: <6l2371$9oa@news.nd.edu> <6l3k1i$85e$1@unlnews.unl.edu> <35756b1a.0@epflnews.epfl.ch> fukuda@ifor17.ethz.ch (Komei Fukuda) wrote: >echo "int xxy_us_dummy;" >tmp-dum.c >./xgcc -B./ -S tmp-dum.c >tmp-dum.c:2: parse error at null character If you look the content of tmp-dum.c, there is additional line with nothing in it. I edited the file and making gcc went successful beyond (if I remember correctly) on 3.3 for Intel. I no longer have the binary since I moved to 4.2 and gcc got broken. I'll wait to hear from Rex for a port of 2.8.x meanwhile I'm living with his 2.7.2.3 port :) I didn't try to make libg++-2.8.1.1a. Hope this helps. -- Satoru Uzawa, satoru@candenext.lsa.berkeley.edu (NeXTmail welcome)
From: Dan Wellman <wellman@students.uiuc.edu> Newsgroups: comp.sys.next.programmer Subject: EOF autorelease Date: 3 Jun 1998 20:05:31 GMT Organization: University of Illinois at Urbana-Champaign Message-ID: <6l4aab$ff2$1@vixen.cso.uiuc.edu> In the following code auto-generated by EOF we see autorelease being used on a variable we want to dispose. Why does this code use autorelease instead of release? I thought that it was better to use release instead of autorelease? Is there a case when it's better to use autorelease -- even when we're not returning a variable? - (void)setMyVal:(NSString *)value { [self willChange]; [MyVal autorelease]; MyVal = [value retain]; } Thanks for your help in advance! dan -- Dan Wellman <> wellman@uiuc.edu <> http://www.cen.uiuc.edu/~wellman/ "A million thoughts in one night can't be wrong" - Cause & Effect
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: Query on Mach Messaging Date: 3 Jun 98 09:39:06 Organization: Is a sign of weakness Message-ID: <SCOTT.98Jun3093906@slave.doubleu.com> References: <6l3mmi$ia7$1@lyra.csx.cam.ac.uk> In-reply-to: William.Clocksin@CL.cam.ac.uk's message of 3 Jun 1998 14:30:42 GMT In article <6l3mmi$ia7$1@lyra.csx.cam.ac.uk>, William.Clocksin@CL.cam.ac.uk writes: I noticed in the demo example SortingInAction that Mach messages are constructed and sent to another thread. It is perfectly understandable, and it demonstrably works. However, because the message body is in the same place in memory used again and again, what prevents the sender from trashing a previous message with a new message before the receiver has handled the first message? If the messenging system passes a pointer, I can't see how the integrity of the message body can remain intact in a multi-threaded environment. The memory is not protected with a mutex as far as I can tell. Or is the message content copied out of memory when the message is sent (as opposed to when it is received, because the sender could trash it immediately after sending it)? And if it is copied, how much space does the kernel make available for message copies? I'd guess the kernel marks it copy-on-write. Thus, if nobody modifies it, there's only one copy in memory (perhaps referenced from multiple virtual memory spaces). If anyone writes it, they get their own copy to munge (though, to be honest, I don't know if you can write to received messages. Never thought about it, I guess). Keep in mind, though, that messages may be modified by the kernel itself. For instance, port rights have to be translated from the sender's port space to the receiver's port space. [Note that there's not a problem with messages between hosts. In that case, the shared memory transfer is used to get the message to the nmserver (net message server), which is the proxy for all ports off the local machine. nmserver bundles things up, ships it to another nmserver, which unbundles things, figures out which local port to use, and lets the kernel do the shared memory trick again.] -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.next.programmer Subject: Re: SUBMISSION: NSPPL Framework Date: 4 Jun 1998 00:06:49 GMT Organization: MiscKit Development Message-ID: <6l4oep$ml5$2@news.xmission.com> References: <6l4h14$k8u$1@news.digifix.com> <6l4ifm$4jv$1@cronkite.cygnus.com> jrudd@cygnus.com (John Rudd) wrote: > In <6l4h14$k8u$1@news.digifix.com> Scott Anguish wrote: > > Apple has released the sources to the NSPPL class. The NSPPL class was > > removed from Foundation when Rhapsody DR2 was released. > > I've just been looking through the foundation spec, and can't find an NSPPL > class. Anyone got information? Was it something new to Rhapsody and I > should therefore not be concerned yet and no public information is available? > :-} It is in OPENSTEP 4.2...and not part of "strict" OPENSTEP, since it is a later addition. Seeing as you can just download it from Apple, there's little point in me emailing you docs or headers. :-) It _is_ a pretty nifty class. Sort of reminiscent of the IXKit's low level storage classes, but refined to make property lists persistent. I haven't really used it, though, so I can't comment too much on the performance aspects. -- Later, -Don Yacktman don@misckit.com <a href="http://www.misckit.com/don.html">My home page</a>
From: eric@skatter.USask.Ca Newsgroups: comp.sys.next.programmer Subject: User frameworks on OPENSTEP Enterprise for Windows NT Date: 3 Jun 1998 20:08:25 GMT Organization: University of Saskatchewan Message-ID: <6l4afp$ae8$1@tribune.usask.ca> I'm a long-time UNIX/NeXTSTEP/OPENSTEP programmer who's trying to move some applications over to OPENSTEP/NT. I've managed to make a couple of applications work, but I seem to have hit a stumbling block when trying to use my own frameworks. Here are the details -- I hope this isn't too long-winded. I've got a bunch of classes for controlling various pieces of equipment here. Because the classes are used by several different applications I want to make the classes into a framework. On OPENSTEP/Mach this worked with no problems, I just created a new project, set it up as a framework, dragged in my class .h and .m files, built and installed it in /LocalLibrary/Frameworks/SalStepper.framework. In my appliction projects I just dragged and dropped the /LocalLibrary/Frameworks/SalStepper.framework in the Frameworks suitcase and the application builds and runs with no problems. My efforts at getting this working on NT have been less successful. Concern: When I compile my framework I get a whole bunch of warnings like: libtool: warning: -compatibility_version ignored libtool: warning: -current_version ignored libtool: warning: -install_name ignored F:/OpenStep/SalStepper/derived_src/SalStepper.def : warning LNK4087: CONSTANT keyword is obsolete; use DATA F:/OpenStep/SalStepper/derived_src/SalStepper.def : warning LNK4087: CONSTANT keyword is obsolete; use DATA F:/OpenStep/SalStepper/derived_src/SalStepper.def : warning LNK4087: CONSTANT keyword is obsolete; use DATA . . . . The build says that it succeeds, though, and installs the framework in: f:/LocalDeveloper/Frameworks/SalStepper.framework/ So far so good: When I compile my application which uses the framework I get no warnings nor errors. Showstopper: When I try to run the application (from the Project builder gdb panel) I get: (gdb) run Starting program: F:/OpenStep/StepperTest/StepperTest.app/StepperTest.exe 34020000:C:\NEXT\NEXTLIBRARY\EXECUTABLES\AppKit.dll 32010000:C:\NEXT\NEXTLIBRARY\EXECUTABLES\Foundation.dll 31000000:C:\NEXT\NEXTLIBRARY\EXECUTABLES\nextpdo.dll 78000000:C:\WINNT\system32\MSVCRT.dll 77f00000:C:\WINNT\system32\KERNEL32.dll 77e70000:C:\WINNT\system32\USER32.dll 77ed0000:C:\WINNT\system32\GDI32.dll 77dc0000:C:\WINNT\system32\ADVAPI32.dll 77e10000:C:\WINNT\system32\RPCRT4.dll 77c40000:C:\WINNT\system32\SHELL32.dll 71030000:C:\WINNT\system32\COMCTL32.dll 776d0000:C:\WINNT\System32\WSOCK32.dll 776b0000:C:\WINNT\System32\WS2_32.dll 776a0000:C:\WINNT\System32\WS2HELP.dll 5f600000:C:\WINNT\System32\WINSPOOL.DRV 77d80000:C:\WINNT\system32\comdlg32.dll 77b20000:C:\WINNT\system32\ole32.dll Program received signal SIGSEGV, Segmentation fault. 0x5500 in ?? () (gdb) I would greatly appreciate any suggestions as to what I'm doing wrong here... BTW -- I tried rebuilding my application by just dragging and dropping all my framework source files into the application and it builds and runs properly! This is fine for a single application, but I've got lots of applications that use the same classes. I would *really* rather use a framework! Thanks, -- Eric Norum eric@skatter.usask.ca Saskatchewan Accelerator Laboratory Phone: (306) 966-6308 University of Saskatchewan FAX: (306) 966-6058 Saskatoon, Canada.
From: nospam+nospam+nospam+this+works@luomat.peak.org (Timothy J Luoma) Newsgroups: comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.programmer Subject: Re: Bad Size for a 2GB HD on Cube Followup-To: comp.sys.next.hardware MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit References: <35751CB7.FDFB0BBC@inria.fr> Message-ID: <FJgd1.95$BE5.569527@news.rdc1.nj.home.com> Date: Wed, 03 Jun 1998 18:37:25 GMT NNTP-Posting-Date: Wed, 03 Jun 1998 11:37:25 PDT Organization: @Home Network NOTE: FOLLOWUPS TO COMP.SYS.NEXT.HARDWARE This is not really a software problem, and certainly not a programmer problem. Followups set appro. In <35751CB7.FDFB0BBC@inria.fr> Samuel Boivin wrote: > During boot, it is identified as a 2051MB Disk. When I use 'BuildDisk' > to build it, it seems to be OK: the disk is identified as a 2051MB Disk > and is correctly constructed(I mean: no error during construction). > But the problem is: if I look at my Workspace, I can see: 840MB free > !!! The disk is recognized as a 1GB !! This is confirmed by: if I use > 'df /dev/rsd1a' under Terminal.app, it gives a size of 1GB, 160MB > occuppied, 16%full, etc... This makes me think that it was made into two partitions of 1gig each. I still believe that 2047mb is as much as you can get on one partition. Anyway, the second partition may not have been mounted. Try this: /etc/mount -t 4.3 -o rw,noquota /dev/sd1b /Partition2 Note: the sd1b might be different.... the 'sd1' part should be OK, but the letter might even be "h" If you get it to work, you'll need to edit /etc/fstab to automount the other partition something like this: /dev/sd0a / 4.3 rw,noquota,noauto 0 1 /dev/sd1a /Partition1 4.3 rw,noquota 0 2 /dev/sd1b /Partition2 4.3 rw,noquota 0 3 Note that the sd1b should be the same as whatever you used with the mount command above. I'm also assuming you have an sd0a which is your regular boot drive that you used to Build this new drive > PS: I tried to build 2 partitions of 1GB, but nothing changed ! I only > have ONE 1GB partition and I can't find the second !! Where is it ? Same problem as above, it may just need to be mounted by hand and then have the entry added to /etc/fstab TjL
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: EOF autorelease Date: 3 Jun 98 14:17:57 Organization: Is a sign of weakness Message-ID: <SCOTT.98Jun3141757@slave.doubleu.com> References: <6l4aab$ff2$1@vixen.cso.uiuc.edu> In-reply-to: Dan Wellman's message of 3 Jun 1998 20:05:31 GMT In article <6l4aab$ff2$1@vixen.cso.uiuc.edu>, Dan Wellman <wellman@students.uiuc.edu> writes: In the following code auto-generated by EOF we see autorelease being used on a variable we want to dispose. Why does this code use autorelease instead of release? I thought that it was better to use release instead of autorelease? Is there a case when it's better to use autorelease -- even when we're not returning a variable? - (void)setMyVal:(NSString *)value { [self willChange]; [MyVal autorelease]; MyVal = [value retain]; } Thanks for your help in advance! If MyVal==value, then you'd have to code it instead as: -(void)setMyVal:(NSString *)value { [self willChange]; if( MyVal!=value) { [MyVal release]; MyVal=[value retain]; } } Using -autorelease makes it somewhat more clear. And keep in mind that there's a possibility that value has been retained by MyVal - so in my code snippet above, the [MyVal release] could release the last retain of value! [Admittedly, that's unlikely to happen in this particular case - unless there's a funky NSString implementation involved!] Many people also code -dealloc as: -(void)dealloc { [MyVal autorelease]; [MyOtherVal autorelease]; [super dealloc]; } This time, the consideration is that by using -autorelease, you brush certain ref-after-free type bugs under the rug. You aren't really _fixing_ them, you're just making them moot. NSAutoReleasePool appears to have been coded so that newly autoreleased objects are added to the _end_ of the _current_ autorelease pool, and the pool is released from the beginning. So if the -dealloc above was initiated by a release of an autorelease pool, the newly autoreleased objects will also be released during the same autorelease pool's release. [Specifically, if there's one object in the pool, with the above -dealloc method, after it's called there will be two objects, which each will also be autoreleased. If they autorelease other objects, those will also go to the end and get autoreleased themselves.] Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: SUBMISSION: NSPPL Framework Date: 4 Jun 1998 03:44:38 GMT Organization: Digital Fix Development Message-ID: <6l5576$qmr$1@news.digifix.com> References: <6l4h14$k8u$1@news.digifix.com> <6l4ifm$4jv$1@cronkite.cygnus.com> In-Reply-To: <6l4ifm$4jv$1@cronkite.cygnus.com> On 06/03/98, John Rudd wrote: >In <6l4h14$k8u$1@news.digifix.com> Scott Anguish wrote: >> Apple has released the sources to the NSPPL class. The NSPPL class was >> removed from Foundation when Rhapsody DR2 was released. >> > >I've just been looking through the foundation spec, and can't find an NSPPL >class. Anyone got information? Was it something new to Rhapsody and I >should therefore not be concerned yet and no public information is available? >:-} The docs were in the 4.2 and DR1 distributions, and are included in the archive on ftp.apple.com. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: jrudd@cygnus.com (John Rudd) Newsgroups: comp.sys.next.programmer Subject: Re: SUBMISSION: NSPPL Framework Date: 3 Jun 1998 22:24:54 GMT Organization: Cygnus Solutions Message-ID: <6l4ifm$4jv$1@cronkite.cygnus.com> References: <6l4h14$k8u$1@news.digifix.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: sanguish@digifix.com In <6l4h14$k8u$1@news.digifix.com> Scott Anguish wrote: > Apple has released the sources to the NSPPL class. The NSPPL class was > removed from Foundation when Rhapsody DR2 was released. > I've just been looking through the foundation spec, and can't find an NSPPL class. Anyone got information? Was it something new to Rhapsody and I should therefore not be concerned yet and no public information is available? :-} -- John "kzin" Rudd jrudd@cygnus.com http://www.cygnus.com/~jrudd Intel: Putting \"I want a pair of Daisy Eagle semi-auto paintball pistols the backward in \in shoulder rigs. Who cares if you win the game as long backward compatible\as you can John Woo as you dive over obstacles?" - anon
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 3 Jun 1998 18:32:37 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6l44s5$5dc$7@ns3.vrx.net> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku7hc$g29$2@ns3.vrx.net> <rex-0106981800560001@192.168.0.3> <6kuml2$rhg$1@ns3.vrx.net> <rex-0106981924540001@192.168.0.3> <6kuucg$2uf$2@ns3.vrx.net> <rex-0106982229380001@192.168.0.3> <6l0es4$8th$4@ns3.vrx.net> <rex-0306980025070001@192.168.0.3> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: rex@smallandmighty.com In <rex-0306980025070001@192.168.0.3> Eric King claimed: > The BShelf's main purpose is to accept drag BMessages and instantiate > any replicants in them. The actual instantiation process is handled > entirely by the OS. Ok. > Yep. Though you can easily define BMessaging protocols. For instance, > there's one for passing color values between apps. Same as OS then. Both need wide XML or Fe. > There data interchange standard is the BMessage. They are used all over > the place. That's not a data interchange format, that's a message interchage _system_. The meaning of the message itself is contextual. That's bad. That means that if the creator of the message is not available, the message is meaningless. This is one of my big complaints about OS, and now Be too it seems. > Yup. It calls the replicants instantiate method, which reconstitutes. Sure, but is this a side effect of the responder chain? > If you look closely at the Be API, just all GUI of the elements have > constructors that take BMessages. A class just needs to inherit from > BArchivable and implement a couple of methods for it to become a > replicant. That doesn't mean anything though, the same is just as true of OpenStep (in fact, I have no doubt that lots of this was lifted from OS). The issue is that the applications are typically monolythic, so while this works just as well under OS it requires the original app/lib to be there to handle the messages. This is not at all like OD, well not in theory anyway. > Everything's done through shared libraries. In the BeOS any executable > that exports the appropriate symbols can be used as a shared library, i.e. > there's no significant difference between a shared library and an > application. This is a _bad_ thing. This is the OS model (I seem to be saying that a lot). The issue here is whether apps are headless as they are in OD, or headed as they are in OS, and from what I've seen to date, in BeOS as well. > :are still "headed", in that the NSApp supports the event system and splitting > :that to another instance is not easy. > > Why would you need another instance of NSApp? Where do events go? > : Can _any_ app become support for a BReplicant? Or do you have to spin off > :that support into a shared lib? > > It's just a matter of exporting the appropriate symbols and attaching a > small resource. It's not a complex process. Yes it _is_, I don't think you understand the problem. > Nope BMessages, really are just thin wrappers over BeOS kernel > messages. It's the most convenient and least dangerous way for threads to > communicate with each other. Then how does it do remote instantation? How about message cueing/balancing? TP? > If that other object happens to be a thread or in another address > space, it would make a lot of sense to use a Mach port to send the data. > This is the sort of thing ports are meant for. Nooooooooooooo! You're missing the level of abstraction in the middle! > C++ is so low-level, I think that you could probably *could* do it. As > shown by David Stes' POC, Objective-C can be compiled down to normal C. Uhh, that's the point. > I'm sure you could wrap the raw C calls up in C++ to make them look > prettier. Har, wrap C in C++ and make it prettier?!? :-) > Because of a lot of work on the part > of Next, and now Apple, it's not the Appkit way, but the Appkit != > Postscript. My point exactly. Please keep in mind I'm developing a graphics app. > But then you're really comparing the Taligent framework with the OO > layer on top of Postscript, There is not OO layer (at least not in OS), there's a functional one. It's bad enough, but I'm getting my work done. > I think you're muddying things. The floating-point coordinate space is > similar but really only one fairly minor issue. _I_ don't think so. > That said there are some Appkit classes, most notably NSBezierPath, > that are similar, but NSBezierPath is *not* a part of Postscript. But that's the point. DPS is a display engine, nothing more. > But it would also make it easier for smaller developers to compete with > Adobe and the other large developers. This is not helping big developers, > it's hurting. Ahhhhhh. Well then, let's hope they do it!! Maury
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: cancel <356D5070.977F0F6B@Square.nl> Control: cancel <356D5070.977F0F6B@Square.nl> Date: Thu, 04 Jun 1998 08:40:03 +0200 Organization: NLnet Message-ID: <6l5fgq$rb1$1@news.utrecht.NL.net> References: <356D5070.977F0F6B@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit This message was cancelled from within Mozilla.
From: rex@smallandmighty.com (Eric King) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Thu, 04 Jun 1998 04:33:23 -0400 Organization: The Small & Mighty Group Message-ID: <rex-0406980433240001@192.168.0.3> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku7hc$g29$2@ns3.vrx.net> <rex-0106981800560001@192.168.0.3> <6kuml2$rhg$1@ns3.vrx.net> <rex-0106981924540001@192.168.0.3> <6kuucg$2uf$2@ns3.vrx.net> <rex-0106982229380001@192.168.0.3> <6l0es4$8th$4@ns3.vrx.net> <rex-0306980025070001@192.168.0.3> <6l44s5$5dc$7@ns3.vrx.net> In article <6l44s5$5dc$7@ns3.vrx.net>, maury@remove_this.istar.ca (Maury Markowitz) wrote: : That's not a data interchange format, that's a message interchage _system_. In the BeOS, it's fast becoming both. : The meaning of the message itself is contextual. True, but you can get a some meaning from the names and types of BMessage items. Furthermore you can attach as much metainfo as you want in the message. Developers still should establish protocols though. : The issue is that the applications are typically monolythic, so while this :works just as well under OS it requires the original app/lib to be there to :handle the messages. I dug into the BeBook on this issue, and no it doesn't. It simply requires that something understands what's in the BMessage. : This is not at all like OD, well not in theory anyway. What they really need is something like Bento. :> :are still "headed", in that the NSApp supports the event system and :splitting :> :that to another instance is not easy. :> :> Why would you need another instance of NSApp? : : Where do events go? I suppose under OpenStep, it might be a bit trickier, since it has a more monolithic event model. In the BeOS, since windows run in their own threads, they have their own private event loops. All events are sent via BMessages. Depending on their type, event messages are either sent to a window or the BApplication running. Since both are descended from BLoopers, you can install BMessageFilters to intercept and preprocess messages. Be's Will Adams used BMessageFilters to show how you can create a simple Be app with a GUI and no need to subclass any of the user interface or application kit classes. Overall, I think the BeOS event handling system is actually a pretty cool design that avoids a lot of problems that C++ could cause. : Yes it _is_, I don't think you understand the problem. Ahh, I think I understand now. Do you mean can any app be used as the 'engine' for a given replicant? Yes. In a replicant BMessage, there exists an item named 'add-on' which specifies the mime-signature of the app/add-on needed to instantiate the replicant. There's another item named 'class' which holds the class name of the replicant. The specified add-on must export the symbols for the class specified in the replicant for instantiation to work. Though, nothing is preventing you from changing the class name also, or just using the instance variables and other information stored in the replicant's BMessage directly. As an example, suppose app 'A' and app 'B' could both export classes named 'goo' with identical methods. Now let's say a 'goo' replicant from 'A' is sent to app 'C'. If you change the mime signature stored in the 'add-on' item to that of app 'B', app 'C' will use code from app 'B' to instantiate the replicant, not app 'A'. One thing I really like about the BeOS is that its messaging system is very explicit. From a language standpoint, it's not as convenient to use as Obj-C, but it's pretty flexible. : Nooooooooooooo! You're missing the level of abstraction in the middle! Okay, fine there's a level of abstraction in the middle. But at some point in the bowels of the system they are probably using kernel ports for sending actual data back and forth. Since PDO works across threads, applications, and networks it makes sense for local communication to use ports, network communication to use sockets, etc. Of course, the actual implementation is hidden very well in OpenStep. However, in the BeOS, it is not hidden that well. : Har, wrap C in C++ and make it prettier?!? :-) Okay, Prettier might be a stretch. ;) How about more convenient? : There is not OO layer (at least not in OS), there's a functional one. It's actually both. Some drawing tasks are handled in classes others require the functional interface. :It's bad enough, but I'm getting my work done. I always thought it really odd, that Next worked so hard to abstract and objectify their OS, but didn't bother to wrap up the basic drawing functions in even simple classes. Of course some folks would argue that it's better to have a functional immediate mode interface than a retained object-oriented one. I'm not one of those folks, but the issue does come up. : But that's the point. DPS is a display engine, nothing more. No, it was quite a bit more. It can do all sorts of 'weird' things like access files, play sounds, etc. It's really a Postscript programming environment. -Eric
From: holger@object-factory.REMOVETHIS.com (Holger Hoffstaette) Newsgroups: comp.sys.next.programmer Subject: Re: User frameworks on OPENSTEP Enterprise for Windows NT Date: 4 Jun 1998 07:46:54 GMT Organization: the unstoppable code machine Message-ID: <6l5jde$p9g$1@leonie.object-factory.com> References: <6l4afp$ae8$1@tribune.usask.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit eric@skatter.USask.Ca wrote: > Showstopper: When I try to run the application (from the Project builder gdb > panel) I get: > > (gdb) run > Starting program: > F:/OpenStep/StepperTest/StepperTest.app/StepperTest.exe > 34020000:C:\NEXT\NEXTLIBRARY\EXECUTABLES\AppKit.dll > 32010000:C:\NEXT\NEXTLIBRARY\EXECUTABLES\Foundation.dll > 31000000:C:\NEXT\NEXTLIBRARY\EXECUTABLES\nextpdo.dll > 78000000:C:\WINNT\system32\MSVCRT.dll > 77f00000:C:\WINNT\system32\KERNEL32.dll > 77e70000:C:\WINNT\system32\USER32.dll > 77ed0000:C:\WINNT\system32\GDI32.dll > 77dc0000:C:\WINNT\system32\ADVAPI32.dll > 77e10000:C:\WINNT\system32\RPCRT4.dll > 77c40000:C:\WINNT\system32\SHELL32.dll > 71030000:C:\WINNT\system32\COMCTL32.dll > 776d0000:C:\WINNT\System32\WSOCK32.dll > 776b0000:C:\WINNT\System32\WS2_32.dll > 776a0000:C:\WINNT\System32\WS2HELP.dll > 5f600000:C:\WINNT\System32\WINSPOOL.DRV > 77d80000:C:\WINNT\system32\comdlg32.dll > 77b20000:C:\WINNT\system32\ole32.dll > > Program received signal SIGSEGV, Segmentation fault. > 0x5500 in ?? () > (gdb) > You can see that your framework DLL isn't pulled in by the M$ linker. That's because it's a little weak in the dynaloading department, and needs a little help. Put something like this into your app_main.m: #if defined (WIN32) #import <YourFramework/SomeHeaderInTheFramework.h> void _referenceAllFrameworks(void) { id dummyInstance = [OneOfYourClasses new]; } #endif WIN32 This will cause your framework to be loaded, with an explicit linker dependency in the main module. I think there's also a NeXTanswer on the topic. Holger -- Object Architect, Object Factory GmbH | It's like a jungle sometimes, @work: holger"at"object-factory.com | it makes me wonder @home: hhoff"at"ragnarok.en.uunet.de | how I keep from going under
From: Samuel Boivin <Samuel.Boivin@inria.fr> Newsgroups: comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.programmer Subject: Re: Bad Size for a 2GB HD on Cube Date: Thu, 04 Jun 1998 12:28:17 +0200 Organization: I.N.R.I.A. Message-ID: <357676C1.60EC08E4@inria.fr> References: <35751CB7.FDFB0BBC@inria.fr> <FJgd1.95$BE5.569527@news.rdc1.nj.home.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Timothy J Luoma <nospam+nospam+nospam+this+works@luomat.peak.org> CC: Samuel.Boivin@inria.fr Timothy J Luoma wrote: > > POSTED TO USENET GROUP(S): comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.programmer > > NOTE: FOLLOWUPS TO COMP.SYS.NEXT.HARDWARE > > This is not really a software problem, and certainly not a programmer > problem. In my opinion, everything that stops you in your programmation under NeXTStep is a NeXT-programmer-problem... A programmer problem is not necessarly a pure code problem. Moreover, the best answers to a problem generally come from NeXT programmers... This could be a NeXT-software-problem since the reason coulde be a bad NS3.3 installation... for example. But I will remember your remarks for NeXT time ;-). > This makes me think that it was made into two partitions of 1gig each. > I still believe that 2047mb is as much as you can get on one partition. In fact, I just found the answer to my problem yesterday, after re-reading the 'NeXT Answers'... and you are right. NeXT Cube does not accept partitions bigger than 2047MB. Then I created two: 2026MB and 25 MB (apparently the smallest that you can create), and it works, except I didn't manage to mount the second (?). > Anyway, the second partition may not have been mounted. > /etc/mount -t 4.3 -o rw,noquota /dev/sd1b /Partition2 I tried this: /etc/mount /dev/sd1b /HD3 It gave: I/O ERROR. > Note: the sd1b might be different.... the 'sd1' part should be OK, but the > letter might even be "h" OK, I will try this. > If you get it to work, you'll need to edit /etc/fstab to automount the other > partition something like this: > /dev/sd0a / 4.3 rw,noquota,noauto 0 1 > /dev/sd1a /Partition1 4.3 rw,noquota 0 2 > /dev/sd1b /Partition2 4.3 rw,noquota 0 3 I already did that, except the last line, since I don't manage to mount the second partition manually. > I'm also assuming you have an sd0a which is your regular boot drive that you > used to Build this new drive Yes, it is. Thank you very much for your answers. It will certainly help me to finish the job and to mount the second partition correctly. Thank you again, Samuel.
From: mmalcolm crawford <malcolm@plsys.co.uk> Newsgroups: comp.sys.next.programmer,comp.sys.next.advocacy Subject: Re: whats up with zoneraiders?!?/! Date: 4 Jun 1998 10:32:30 GMT Organization: P & L Systems Message-ID: <6l5t3u$adf$4@ironhorse.plsys.co.uk> References: <6j04ej$nvc$1@newsfep4.sprintmail.com> <6j0bma$l3h$3@ha2.rdc1.nj.home.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: nospam+yes-this-is-a-valid_address@luomat.peak.org In <6j0bma$l3h$3@ha2.rdc1.nj.home.com> Timothy Luoma wrote: > In <6j04ej$nvc$1@newsfep4.sprintmail.com> macghod@concentric.net wrote: > > Why in gods name would zoneraiders be ppc only? > [...] ("Zone Warrior") > My guess is the only people who know the real answer are the people who > wrote it. > A fair comment, so I asked them (the developer's actually English, and came over to our offices a few months ago)... ... he said that the reason's because a lot of data (models,textures etc) are pre-generated on MacOS and are therefore big-endian. He could rewrite stuff, but he doesn't have a PC to test things on (an important consideration here), and there are a couple of other issues which impinge... Best wishes, mmalc.
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 4 Jun 1998 12:24:40 GMT Organization: Technical University of Berlin, Germany Message-ID: <6l63m8$t8u$1@news.cs.tu-berlin.de> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku7hc$g Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit rex@smallandmighty.com (Eric King) writes: [..] > I suppose under OpenStep, it might be a bit trickier, since it has a >more monolithic event model. In the BeOS, since windows run in their own >threads, they have their own private event loops. All events are sent via >BMessages. All events in OS are eventually sent as Objective-C messages, which I would sayt is actually slighltly *more* general than the Be approach. The AppKit handling of user events is fairly monolithic, but don't let that confuse you. The even system, based on standard messages, NSRunLoops and the responder chain, is very flexible. [example deleted] > Overall, I think the BeOS event handling system is actually a pretty >cool design that avoids a lot of problems that C++ could cause. Yes, the problems that use of Objective-C avoids altogether... :-) > I always thought it really odd, that Next worked so hard to abstract >and objectify their OS, but didn't bother to wrap up the basic drawing >functions in even simple classes. Of course some folks would argue that >it's better to have a functional immediate mode interface than a retained >object-oriented one. I'm not one of those folks, but the issue does come >up. Having an OO interface is wonderful. However, for some drawing tasks it is essential to be able to get to the metal, just like it is sometimes necessary to code in straight C. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: Trevin Beattie <*trevin*@*xmission*.*com*> Newsgroups: comp.sys.next.programmer Subject: Re: gcc-2.8.x ported to nextstep 3.3 on intel? Date: Thu, 04 Jun 1998 13:03:34 +0000 Organization: XMission Internet (801 539 0852) Message-ID: <6l65hb$k5q$1@news.xmission.com> References: <6l2371$9oa@news.nd.edu> <6l3k1i$85e$1@unlnews.unl.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Rex Dieter wrote: > > I've been maintaining a gcc port for nextstep, currently at version > 2.7.2.3. I've been waiting for g77 to become available for gcc-2.8.x > (since we use it here a lot) before doing a 2.8.x port for nextstep. > Perhaps I should try out egcs (URL (?) http://egcs.cygnus.com/) which has > g77 built-in. I was able to compile gcc 2.8.1 from MIT sources, but it required using Gnu make 3.76 and some environment variables which I don't remember offhand. Also, I had to set --with-gnu-as and --with-gnu-ld in config, and manually copy NS' as and ld to /usr/local/lib/gcc-lib/{arch}/.... Anyway, the point is that getting gcc compiled and running is easy. The hard part (I haven't been able to do it yet; no time) is getting binutils to run, so I can use the latest as, ld, and other such development tools. I haven't even seen any _old_ versions of precompiled binutils, last time I checked. Also, I would like to see a new library and set of headers that are fully compliant with current ANSI and POSIX standards. -- Trevin Beattie "Do not meddle in the affairs of wizards, *To*reply*to*this* for you are crunchy and good with ketchup." *message,*remove*the* --unknown *asterisks*from*my*email*address.*
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6l4dim$o4m$14032@news.cybercity.dk> Control: cancel <6l4dim$o4m$14032@news.cybercity.dk> Date: 04 Jun 1998 14:10:45 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6l4dim$o4m$14032@news.cybercity.dk> Sender: S A F<myemail@any.where.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 4 Jun 1998 09:27:12 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6l5p9g$ebr$6@ns3.vrx.net> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku7hc$g29$2@ns3.vrx.net> <rex-0106981800560001@192.168.0.3> <6kuml2$rhg$1@ns3.vrx.net> <rex-0106981924540001@192.168.0.3> <6kuucg$2uf$2@ns3.vrx.net> <rex-0106982229380001@192.168.0.3> <6l0es4$8th$4@ns3.vrx.net> <rex-0306980025070001@192.168.0.3> <6l44s5$5dc$7@ns3.vrx.net> <rex-0406980433240001@192.168.0.3> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: rex@smallandmighty.com In <rex-0406980433240001@192.168.0.3> Eric King claimed: > In the BeOS, it's fast becoming both. Sorry, but that's a non-sequitur, it's the same on other OS's. Only the MacOS and Win still use specific mechanisms for specific data interchange. For custom data movement, whether that's to disk, onto a pasteboard, or to the network somewhere, OS uses NSCoders. The fact that Be is "fast becoming" that was is not surprising. However this does not solve the _problem_. > : The meaning of the message itself is contextual. > > True, but you can get a some meaning from the names and types of > BMessage items. Furthermore you can attach as much metainfo as you want in > the message. Developers still should establish protocols though. Exactly. > : The issue is that the applications are typically monolythic, so while this > :works just as well under OS it requires the original app/lib to be there to > :handle the messages. > > I dug into the BeBook on this issue, and no it doesn't. It simply > requires that something understands what's in the BMessage. What, where, and who loads it? Are the messages typed? How? MIME? Where's the reverse mapping take place? To what? > What they really need is something like Bento. Yes, they need Fe, or XML. Both have their specific good and bad points. Fe is definitely easier to use as a developer, XML is a better interchange format. > : Where do events go? > > I suppose under OpenStep, it might be a bit trickier, since it has a > more monolithic event model. In the BeOS, since windows run in their own > threads, they have their own private event loops. All events are sent via > BMessages. Hmmmm, who registeres what events to get and to whom? For instance does the window get all events, or does it indicate which it wants, and to what entity? > Depending on their type, event messages are either sent to a window or > the BApplication running. Since both are descended from BLoopers, you can > install BMessageFilters to intercept and preprocess messages. Who do you register them with? The OS or the window server? > : Yes it _is_, I don't think you understand the problem. > > Ahh, I think I understand now. Do you mean can any app be used as the > 'engine' for a given replicant? Yes. Noooo, that's an "of course". Any app on OS can open any bundle too. The issue is whether that app can then load _specific_ bits of functionality out of the creator without loading the whole app. > In a replicant BMessage, there exists an item named 'add-on' which > specifies the mime-signature of the app/add-on needed to instantiate the > replicant. THAT answers my question. This is rather similar to OpenStep's info.plist's which are dropped into bundles. Problems... a) the data is private b) there's no "sub information" available for other substitutes to use c) the ID either works or it fails completely Sorry Eric, I'm well aware that this system doesn't work, because that's what I have. > As an example, suppose app 'A' and app 'B' could both export classes > named 'goo' with identical methods. Now let's say a 'goo' replicant from > 'A' is sent to app 'C'. If you change the mime signature stored in the > 'add-on' item to that of app 'B', app 'C' will use code from app 'B' to > instantiate the replicant, not app 'A'. As an example of the problem, consider app A which has a custome view subclass that contains a BMP and a flag that it should be inverted. Now that's placed in B. If A is missing, you get nothing, even though the data itself is still there. Without agreement on the items inside the data, the flexibility of this solutions is seriously limited. That's why OD started as a file format, rather than adding one. > One thing I really like about the BeOS is that its messaging system is > very explicit. From a language standpoint, it's not as convenient to use > as Obj-C, but it's pretty flexible. That's because it _has_ a model. Most systems ignore this, OpenStep doesn't need one. > I always thought it really odd, that Next worked so hard to abstract > and objectify their OS, but didn't bother to wrap up the basic drawing > functions in even simple classes. Agreed, I don't understand the specifics of this either. Sure, speed could be an issue, but CPU's have grown so much that any possible issue there in the past isn't there now. > it's better to have a functional immediate mode interface than a retained > object-oriented one. I'm not one of those folks, but the issue does come > up. It's best to have both. Maury
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 4 Jun 1998 09:28:13 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6l5pbd$ebr$7@ns3.vrx.net> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6l63m8$t8u$1@news.cs.tu-berlin.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: marcel@cs.tu-berlin.de In <6l63m8$t8u$1@news.cs.tu-berlin.de> Marcel Weiher claimed: > Having an OO interface is wonderful. However, for some drawing tasks > it is essential to be able to get to the metal, just like it is > sometimes necessary to code in straight C. Sure, but that doesn't explain why there isn't a OOPS lib. Maury
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 4 Jun 1998 20:32:01 GMT Organization: Technical University of Berlin, Germany Message-ID: <6l7081$cma$1@news.cs.tu-berlin.de> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0. Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit maury@remove_this.istar.ca (Maury Markowitz) writes: >In <6l63m8$t8u$1@news.cs.tu-berlin.de> Marcel Weiher claimed: >> Having an OO interface is wonderful. However, for some drawing tasks >> it is essential to be able to get to the metal, just like it is >> sometimes necessary to code in straight C. > Sure, but that doesn't explain why there isn't a OOPS lib. It sure doesn't. Maybe Lawson and Lawrence lobbied them to have one. :-))) Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: rex@smallandmighty.com (Eric King) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Thu, 04 Jun 1998 17:08:21 -0400 Organization: The Small & Mighty Group Message-ID: <rex-0406981708220001@192.168.0.3> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku7hc$g29$2@ns3.vrx.net> <rex-0106981800560001@192.168.0.3> <6kuml2$rhg$1@ns3.vrx.net> <rex-0106981924540001@192.168.0.3> <6kuucg$2uf$2@ns3.vrx.net> <rex-0106982229380001@192.168.0.3> <6l0es4$8th$4@ns3.vrx.net> <rex-0306980025070001@192.168.0.3> <6l44s5$5dc$7@ns3.vrx.net> <rex-0406980433240001@192.168.0.3> <6l5p9g$ebr$6@ns3.vrx.net> In article <6l5p9g$ebr$6@ns3.vrx.net>, maury@remove_this.istar.ca (Maury Markowitz) wrote: : However this does not solve the _problem_. Okay, now I have a clearer idea of the problem. You want metainformation included in the archival system such that the original application code is not needed to use what's stored in the archive. I think you're being a bit too quick to compare Be's system with the NSCoder system. They are different. In OpenStep, I can see no way of inspecting the instance variables of an archive and getting their names, without actually instantiating the objects themselves. OpenStep doesn't provide methods to do this. BMessages have methods to do just that. For your example of a replicant view containing a picture. The replicant BMessage must contain the various instance variables needed by the view to instantiate itself. However, the developer of the replicant view has several choices as to how to encode the picture. They could just leave it as a hunk of data bound to some arbitrary name e.g. 'anInstanceVariableForAPicture' or they could do something smart and have the picture bound to a name that contained its mime-type. Because of Obj-C, it looks like developers have to do a lot less work to archive an object in OpenStep, but this archival mechanism is currently very opaque. The BeOS, because of its C++ basis, forces developers to do more work, but in doing so there's plenty of opportunity to store the metainformation that you want in the resulting BMessage. Furthermore, since BMessages are readily inspectible, replicants are also inspectible. (People have already written BMessage inspectors.) The information in a replicant message can be looked at without loading any associated add-on code. Be needs to establish protocols with developers as to how certain types of media should be archived in replicants. They *do* realize this and I'm sure they'll eventually get around to it, once things begin to settle down from the Intel port. For instance, they might advise replicant developers to embed a small XML string containing a description of all of the media types embedded in their replicant's BMessage. This would allow any developer to inspect and work with a foreign BMessage as long as it contained known types of data. You could even have descriptions and proxies for the unknown types of data. This is what you want, right? : Hmmmm, who registeres what events to get and to whom? For instance does :the window get all events, or does it indicate which it wants, and to what :entity? All events get sent to the appropriate BLooper, be it an application, a window, or just some random thread sitting around doing something. The BLooper can either handle the message itself or forward it to one of its views. Message filters can intercept messages being sent to BLoopers and do what the want with them. : Who do you register them with? The OS or the window server? The app server creates and manages all of this stuff. Everything is registered with it. : THAT answers my question. This is rather similar to OpenStep's :info.plist's which are dropped into bundles. Problems... Yes, it's similar, but you've also got named and typed items representing instance variables, as well as other types of state information, stored in the replicant BMessage. As far as I can tell the ability to readily access this information without instantiation isn't present in NSArchives. All you can do is get at the class name and NSBundles just tell you which binary holds the code and resources associated with the archived object. :a) the data is private In the BeOS, it is public. You're perfectly free to browse around inside a BMessage. (Since BMessages are used in so many places, this is actually a useful debugging tool...) :b) there's no "sub information" available for other substitutes to use There's no reason why it couldn't be easily added. The mechanisms are there, the protocols are not. :c) the ID either works or it fails completely Only if you're using the current implementation of BShelf. : Sorry Eric, I'm well aware that this system doesn't work, because that's :what I have. Be's system is different and it is still maturing. They can't tackle all problems at once. They've got a very good foundation with their replicant system. They will get around to building on it. : If A is missing, you get nothing, even though the data :itself is still there. That doesn't have to be true with Be's system. : That's why OD started as a file format, rather than adding one. This is also why replicants were based on BMessages. Every item in a BMessage is named and typed and has established methods for access. I'd still like to have a format like Bento to keep track of revisions and such, but the current BMessage system is workable. : Agreed, I don't understand the specifics of this either. Sure, speed could :be an issue, I think marketing is the real root. Most Next developers just were not making graphics apps. Next had limited resources and such a thing was put pretty low on their list of priorities. -Eric
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 4 Jun 1998 16:54:45 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6l6jgl$4i6$2@ns3.vrx.net> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku7hc$g29$2@ns3.vrx.net> <rex-0106981800560001@192.168.0.3> <6kuml2$rhg$1@ns3.vrx.net> <rex-0106981924540001@192.168.0.3> <6kuucg$2uf$2@ns3.vrx.net> <rex-0106982229380001@192.168.0.3> <6l0es4$8th$4@ns3.vrx.net> <rex-0306980025070001@192.168.0.3> <6l44s5$5dc$7@ns3.vrx.net> <rex-0406980433240001@192.168.0.3> <6l5p9g$ebr$6@ns3.vrx.net> <rex-0406981708220001@192.168.0.3> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: rex@smallandmighty.com In <rex-0406981708220001@192.168.0.3> Eric King claimed: > Okay, now I have a clearer idea of the problem. You want > metainformation included in the archival system such that the original > application code is not needed to use what's stored in the archive. Exactly. Not in all cases of course, but for a generalized BShelf, it's a requirement. > I think you're being a bit too quick to compare Be's system with the > NSCoder system. They are different. In OpenStep, I can see no way of > inspecting the instance variables of an archive and getting their names, > without actually instantiating the objects themselves. Good point, and this exact problem is why just about everyone I know (including myself) uses PPL's instead. For instance you can open our documents in any text editor and see the vars. We've seem multiple posts in these forums where other people did the same thing, and our DocPL classes are a popular download from our web site. Better yet XML is very similar to PPL's in some ways, notably in concept, so I'll be able to write a XML version as soon as the VML/PGML war dies down. When I was at Apple for WWDC, I mentioned this issue on several occasions, that although the coder system was great for what it did (compact representations) it should really be a subclass of a more generalized solution based off of PL's which can also be spooled as a PPL/XML or for that matter, Bento/Fe. However they did not seem to be terribly interested in doing this just yet. > provide methods to do this. BMessages have methods to do just that. That is indeed different. > For your example of a replicant view containing a picture. The > replicant BMessage must contain the various instance variables needed by > the view to instantiate itself. How does one avoid namespace collisions? > view has several choices as to how to encode the picture. They could just > leave it as a hunk of data bound to some arbitrary name e.g. > 'anInstanceVariableForAPicture' or they could do something smart and have > the picture bound to a name that contained its mime-type. Sure, we use FileWrappers. NSImages do this for you too. > Be needs to establish protocols with developers as to how certain types > of media should be archived in replicants. They *do* realize this and I'm > sure they'll eventually get around to it, once things begin to settle down > from the Intel port. One can hope. > data. This is what you want, right? Yes. > All events get sent to the appropriate BLooper, be it an application, a > window, or just some random thread sitting around doing something. Ok, that's cool. So BLooper forms the root of the event chain rather than the App. This is good. Do all messages have to go through the BLooper, or can you call into the other threads as you can under OS? > The app server creates and manages all of this stuff. Everything is > registered with it. How does the mapping from the window to BLooper happen? Is the Window an owner of one? Or is the app the owner? Or does the BLooper own the window? > Yes, it's similar, but you've also got named and typed items > representing instance variables So do plists. > That doesn't have to be true with Be's system. Doesn't _have_ to be true with any system, yet sadly it is with all of them. Well, OpenDoc aside. > This is also why replicants were based on BMessages. I wouldn't say the lineage is that clear, it sounds like everything is built on them, no? > I think marketing is the real root. Most Next developers just were not > making graphics apps. I would have to disagree, graphics apps seem to form the majority of commercial apps on the NeXT platform. Maury
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.next.programmer,comp.sys.next.advocacy Subject: Re: future of mac programming Date: 4 Jun 1998 22:31:02 GMT Organization: MiscKit Development Message-ID: <6l7776$qti$2@news.xmission.com> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku7hc$g29$2@ns3.vrx.net> <rex-0106981800560001@192.168.0.3> <6kuml2$rhg$1@ns3.vrx.net> <rex-0106981924540001@192.168.0.3> <6kuucg$2uf$2@ns3.vrx.net> <rex-0106982229380001@192.168.0.3> <6l0es4$8th$4@ns3.vrx.net> <rex-0306980025070001@192.168.0.3> <6l44s5$5dc$7@ns3.vrx.net> <rex-0406980433240001@192.168.0.3> <6l5p9g$ebr$6@ns3.vrx.net> <rex-0406981708220001@192.168.0.3> rex@smallandmighty.com (Eric King) wrote: > Okay, now I have a clearer idea of the problem. You want > metainformation included in the archival system such that the original > application code is not needed to use what's stored in the archive. > I think you're being a bit too quick to compare Be's system with the > NSCoder system. They are different. In OpenStep, I can see no way of > inspecting the instance variables of an archive and getting their names, > without actually instantiating the objects themselves. OpenStep doesn't > provide methods to do this. BMessages have methods to do just that. You can do it with a custom NSCoder; the idea of coders is sufficiently general that you can get what you want...but a little bit of work is required beyond the simplistic implementation you get in Foundation. I've seen coders that put the archive in property list format, which is an ascii format based on key/value dictionaries...so you *can* name the data members in the flattened objet hierarchy. I do wish that Apple included a few coders like this in the Foundation Kit, however. And why the **** is this Be vs. OPENSTEP pissing match in csn.programmer when it belongs in csn.advocacy? When (if) anyone follows up to this, please trim out csn.programmer. Please put this thread where it belongs... -- Later, -Don Yacktman don@misckit.com <a href="http://www.misckit.com/don.html">My home page</a>
From: nurban@crib.corepower.com (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Window ordering Date: 4 Jun 1998 20:07:32 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6l7cs4$vcg$1@crib.corepower.com> If I have a list of NSWindows, can I tell how they are all ordered relative to each other? -level doesn't do what I want, I want to tell which windows are on top of which, even if they're all at the same window level.
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer,comp.sys.next.advocacy Subject: Re: future of mac programming Date: 4 Jun 1998 19:37:01 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6l6t0t$c3k$1@ns3.vrx.net> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku7hc$g29$2@ns3.vrx.net> <rex-0106981800560001@192.168.0.3> <6kuml2$rhg$1@ns3.vrx.net> <rex-0106981924540001@192.168.0.3> <6kuucg$2uf$2@ns3.vrx.net> <rex-0106982229380001@192.168.0.3> <6l0es4$8th$4@ns3.vrx.net> <rex-0306980025070001@192.168.0.3> <6l44s5$5dc$7@ns3.vrx.net> <rex-0406980433240001@192.168.0.3> <6l5p9g$ebr$6@ns3.vrx.net> <rex-0406981708220001@192.168.0.3> <6l7776$qti$2@news.xmission.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: don@misckit.com In <6l7776$qti$2@news.xmission.com> Don Yacktman claimed: > You can do it with a custom NSCoder; the idea of coders is sufficiently > general that you can get what you want...but a little bit of work is required > beyond the simplistic implementation you get in Foundation. I've seen coders > that put the archive in property list format, which is an ascii format based > on key/value dictionaries...so you *can* name the data members in the > flattened objet hierarchy. I do wish that Apple included a few coders like > this in the Foundation Kit, however. The problem is that the NSCoder API doesn't have room in the current implementation. If they're going to do this, they should do it NOW because the API's harden. > And why the **** is this Be vs. OPENSTEP pissing match Hardly a pissing match. Quite a welcome change for me and Eric I'd say. Maury
From: nospam+nospam+nospam+this+works@luomat.peak.org (Timothy J Luoma) Newsgroups: comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.programmer Subject: Re: Bad Size for a 2GB HD on Cube Followup-To: comp.sys.next.hardware MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit References: <35751CB7.FDFB0BBC@inria.fr> <FJgd1.95$BE5.569527@news.rdc1.nj.home.com> <357676C1.60EC08E4@inria.fr> Message-ID: <eeBd1.145$BE5.1173933@news.rdc1.nj.home.com> Date: Thu, 04 Jun 1998 17:57:30 GMT NNTP-Posting-Date: Thu, 04 Jun 1998 10:57:30 PDT Organization: @Home Network NOTE: FOLLOWUPS TO COMP.SYS.NEXT.HARDWARE In <357676C1.60EC08E4@inria.fr> Samuel Boivin wrote: > Timothy J Luoma wrote: > > > > POSTED TO USENET GROUP(S): > > comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.programmer > > > > NOTE: FOLLOWUPS TO COMP.SYS.NEXT.HARDWARE > > > > This is not really a software problem, and certainly not a programmer > > problem. > > In my opinion, everything that stops you in your programmation under > NeXTStep is a NeXT-programmer-problem... a) You should not ignore the Followups-To setting when replying to a post b) beyond your opinion, there is a list of guidelines which have been used in the NeXT community for years based on the charters for the groups. It can be found here: http://www.stepwise.com/Resources/Newsgroups/roadmap.html In it you will find brief summaries for the groups, including [9]comp.sys.next.programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. So csn.programmer is for programmers to discuss programming issues, and formatting a HD is not a programmer issue. Crossposting is generally looked down upon, especially without use of a Followups-To line. The groups are low volume enough that a post to one group will not be lost. > A programmer problem is not necessarly a pure code problem. But the programmer group is not the right group for what comes down to a hardware (or probably better, sysadmin) question > Moreover, the best answers to a problem generally come from NeXT programmers... So we should all post all our questions to csn.programmer ? I daresay that would not be well accepted by those who are trying to use the groups for its stated purpose of programming issues, not just because these are the folks who really know their stuff. > This could be a NeXT-software-problem since the reason coulde be a bad > NS3.3 installation... for example. Possibly, but I still tent to think that csn.sysadmin would have been a better choice. > But I will remember your remarks for NeXT time ;-). Thanks... it will save bandwidth worldwide ;-) > > This makes me think that it was made into two partitions of 1gig each. > > I still believe that 2047mb is as much as you can get on one partition. > > In fact, I just found the answer to my problem yesterday, after > re-reading the 'NeXT Answers'... and you are right. > NeXT Cube does not accept partitions bigger than 2047MB. Then I created > two: 2026MB and 25 MB (apparently the smallest that you can create), and > it works, except I didn't manage to mount the second (?). I would try '/dev/sd1h' as the name for the second partition.... Actually, I wonder if it needs to be formatted somehow before being ready to be mounted? Perhaps someone else will know the answer to that one. TjL
From: mark@sapphire.oaai.com (Mark Onyschuk) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 4 Jun 1998 14:36:22 GMT Organization: M. Onyschuk and Associates Inc. Message-ID: <slrn6ndtt0.f4a.mark@sapphire.oaai.com> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku7hc$g29$2@ns3.vrx.net> <rex-0106981800560001@192.168.0.3> <6kuml2$rhg$1@ns3.vrx.net> <rex-0106981924540001@192.168.0.3> <6kuucg$2uf$2@ns3.vrx.net> <rex-0106982229380001@192.168.0.3> <6l0es4$8th$4@ns3.vrx.net> <rex-0306980025070001@192.168.0.3> <6l44s5$5dc$7@ns3.vrx.net> <rex-0406980433240001@192.168.0.3> <6l5p9g$ebr$6@ns3.vrx.net> In article <6l5p9g$ebr$6@ns3.vrx.net>, Maury Markowitz wrote: >> I always thought it really odd, that Next worked so hard to abstract >> and objectify their OS, but didn't bother to wrap up the basic drawing >> functions in even simple classes. > > Agreed, I don't understand the specifics of this either. Sure, speed could >be an issue, but CPU's have grown so much that any possible issue there in >the past isn't there now. It was a speed issue - in past certain things that today are classes used to be structs that you'd access by reference and that you cer- tainly would not instantiate hundreds of times in an event loop. And some things that today are methods on first-class objects used to be function calls. Providing an OO wrapper overtop of Display PostScript on an '030 cube would have been unacceptably slow. For that matter, it would probably have been unacceptably slow until quick Pentium boxes became the machine of choice for running NeXTstep. Of course by that time, NeXT had already developed a more narrow focus on "mission critical" enterprise apps where the cost- benefit of such a change to the graphics layer would pretty much rule it out after a nanosecond of consideration. Now it looks like with Rhapsody and OSX, a lot of these loose ends are being attended to. I expect NSBezierPath and friends to become more and more studly as time goes on. Regards, Mark
From: Robert La Ferla <rlaferla_@_cac.stratus.com> Newsgroups: comp.sys.next.programmer Subject: IB Connections Inspector under OPENSTEP Date: 4 Jun 1998 22:19:37 GMT Organization: Stratus Computer Inc, Marlboro MA Message-ID: <6l76hp$hte@transfer.stratus.com> In the OPENSTEP InterfaceBuilder connections inspector, you can select "Outlets" or various "Associations" from the aspect pop up list. How does one go about creating a new connections inspector aspect? We would like to create a connections inspector for a subclass of NSFormatter *BUT* we want to be able to make/edit connections from the inspector panel. Thanks, Robert PS. Reply to the From address and remove the "_"s.
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: Window ordering Date: 5 Jun 1998 02:36:50 GMT Organization: Idiom Communications Message-ID: <6l7lk2$bji$1@news.idiom.com> References: <6l7cs4$vcg$1@crib.corepower.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: nurban@crib.corepower.com Nathan Urban may or may not have said: -> If I have a list of NSWindows, can I tell how they are all ordered -> relative to each other? -level doesn't do what I want, I want to tell -> which windows are on top of which, even if they're all at the same -> window level. Not quite sure of this, but I think you'll need to query their global window numbers, and then get their order from the DPS interpreter. This may require a short pswrap. Check in the red book. -jcr
From: rex@smallandmighty.com (Eric King) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: Thu, 04 Jun 1998 21:16:41 -0400 Organization: The Small & Mighty Group Message-ID: <rex-0406982116430001@192.168.0.3> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku7hc$g29$2@ns3.vrx.net> <rex-0106981800560001@192.168.0.3> <6kuml2$rhg$1@ns3.vrx.net> <rex-0106981924540001@192.168.0.3> <6kuucg$2uf$2@ns3.vrx.net> <rex-0106982229380001@192.168.0.3> <6l0es4$8th$4@ns3.vrx.net> <rex-0306980025070001@192.168.0.3> <6l44s5$5dc$7@ns3.vrx.net> <rex-0406980433240001@192.168.0.3> <6l5p9g$ebr$6@ns3.vrx.net> <rex-0406981708220001@192.168.0.3> <6l6jgl$4i6$2@ns3.vrx.net> In article <6l6jgl$4i6$2@ns3.vrx.net>, maury@remove_this.istar.ca (Maury Markowitz) wrote: :In <rex-0406981708220001@192.168.0.3> Eric King claimed: :> Okay, now I have a clearer idea of the problem. You want :> metainformation included in the archival system such that the original :> application code is not needed to use what's stored in the archive. : : Exactly. Not in all cases of course, but for a generalized BShelf, it's a :requirement. I think Be would very much like to have a generalized BShelf, but it would make a lot more sense to wait until the new Translation and Media kits are complete before creating one. That way even if an app couldn't find the appropriate binary for a replicant and it couldn't understand the embedded data, perhaps the OS could translate the data into something it could understand. : Good point, and this exact problem is why just about everyone I know :(including myself) uses PPL's instead. For instance you can open our :documents in any text editor and see the vars. But what happens when you have binary data? My problem with using XML for this task, is that it doesn't handle binary well at all. Incidentally, isn't NSPPL being replaced by something new? :matter, Bento/Fe. However they did not seem to be terribly interested in :doing this just yet. Well, they are rather busy at the moment. You might have to wait until the Mac OS X time frame or later. I also get the impression that anything that sounds OpenDOCish is looked down upon. : How does one avoid namespace collisions? Collisions where? : Ok, that's cool. So BLooper forms the root of the event chain rather than :the App. This is good. Yes, the BApplication class is really just a special BLooper. It connects with the App Server and does some other setup work. Since it's already running in a thread to begin with BApplications don't start a new thread like other BLoopers. : Do all messages have to go through the BLooper, or A BLooper's reason for being, is to sit around and catch BMessages and do something appropriate when it gets them. It can also send messages to other BLoopers. :can you call into the other threads as you can under OS? I'm not quite sure what you're asking. Specifically what do you want to do, and how do you do it under OpenStep? : How does the mapping from the window to BLooper happen? Is the Window an :owner of one? Or is the app the owner? Or does the BLooper own the window? BWindow is a subclass of BLooper. It *is* the BLooper. BLoopers run in a give BApplication's address space. It really isn't all that complex, and is explained pretty well in the Interface Kit section of the BeBook. http://www.be.com/documentation/be_book/The%20Interface%20Kit/Window.html and http://www.be.com/documentation/be_book/The%20Application%20Kit/Looper.html :> That doesn't have to be true with Be's system. : : Doesn't _have_ to be true with any system, yet sadly it is with all of :them. Well, OpenDoc aside. I think Be will eventually get around to doing the right thing. But remember they're the 'Media OS' so they've got a lot of work ahead to live up to that billing. This issue can be put off for a while. : I wouldn't say the lineage is that clear, it sounds like everything is :built on them, no? Well, they are a useful construct. Sort of a cross between a Bag, a Dictionary, and an Array. OOP is about reuse right? :) :> I think marketing is the real root. Most Next developers just were not :> making graphics apps. : : I would have to disagree, graphics apps seem to form the majority of :commercial apps on the NeXT platform. That may be so, but most Next developers were not making commercial apps, they were developing inhouse apps. -Eric
From: lloyddean@earthlink.net (Lloyd D. Ollmann, Jr.) Newsgroups: comp.sys.next.programmer Subject: POV Ray Sources??? Date: Fri, 05 Jun 1998 00:44:32 -0700 Organization: EarthLink Network, Inc. Message-ID: <lloyddean-0506980044320001@ip53.seattle7.wa.pub-ip.psi.net> Does anyone know if/where I might find sources for either a NextSTEP or OpenSTEP version of POV-Ray and a viewer application?
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: IEEE floating point rounding mode control under Openstep? Date: 5 Jun 1998 11:15:36 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6nfkqr.b0a.rog@talisker.ohm.york.ac.uk> i'm trying to port a strtod function to openstep under intel and have run across the following comment in the code. i'm pretty clueless when it comes to the depths of floating point arithmetic; can anyone tell me if a) what rounding mode the FPU is in by default (perhaps it's in extended precision mode by default?) b) is it possible to access the IEEE control words (i'd prefer not to have to write any assembler, but if that's what it takes, i'll do it) c) might this have implications for other floating-point functions elsewhere in my code? here's the comment: /* On a machine with IEEE extended-precision registers, it is * necessary to specify double-precision (53-bit) rounding precision * before invoking strtod or dtoa. If the machine uses (the equivalent * of) Intel 80x87 arithmetic, the call * _control87(PC_53, MCW_PC); * does this with many compilers. Whether this or another call is * appropriate depends on the compiler; for this to work, it may be * necessary to #include "float.h" or another system-dependent header * file. */ i've found i386/fpu.h, but that header doesn't declare any relevant functions... cheers, rog.
From: nospam+nospam+nospam+this+works@luomat.peak.org (Timothy J Luoma) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.programmer Subject: Re: DHCP for OPENSTEP? Followup-To: comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit References: <6kg47t$d8i@fridge.shore.net> <TyNd1.996$On1.3750778@ptah.visi.com> Message-ID: <UxSd1.254$BE5.1666092@news.rdc1.nj.home.com> Date: Fri, 05 Jun 1998 13:39:00 GMT NNTP-Posting-Date: Fri, 05 Jun 1998 06:39:00 PDT Organization: @Home Network NOTE: FOLLOWUPS TO COMP.SYS.NEXT.PROGRAMMER In <TyNd1.996$On1.3750778@ptah.visi.com> David Young wrote: > In <6kg47t$d8i@fridge.shore.net> Robert La Ferla wrote: > > Looks the source code is available for DHCP from: > > http://www.isc.org/dhcp.html > > Anyone port it to OPENSTEP yet? > > Yes. Me. > > It's in the ISC distribution now. Binaries are available from others. FYI the distribution (Version 2, Beta 1, Patchlevel 0) claims it will compile for NeXTStep but trying to do so gave me immediate fatal errors: cc -g -I.. -I../includes -c raw.c ./includes/cf/nextstep.h:73: warning: redefinition of macro _PATH_DHCLIENT_PID ./includes/cf/nextstep.h:73: warning: is the location of the previous definition ../includes/dhcp.h:58: undefined type, found `u_int8_t' ../includes/dhcp.h:59: undefined type, found `u_int8_t' ../includes/dhcp.h:60: undefined type, found `u_int8_t' ../includes/dhcp.h:61: undefined type, found `u_int8_t' ../includes/dhcp.h:62: undefined type, found `u_int32_t' ../includes/dhcp.h:63: undefined type, found `u_int16_t' ../includes/dhcp.h:64: undefined type, found `u_int16_t' ../includes/sysconf.h:46: undefined type, found `u_int32_t' ../includes/sysconf.h:47: undefined type, found `u_int32_t' ../includes/sysconf.h:73: undefined type, found `u_int8_t' ../includes/sysconf.h:112: undefined type, found `u_int8_t' ../includes/sysconf.h:113: undefined type, found `u_int8_t' ../includes/sysconf.h:114: undefined type, found `u_int8_t' ../includes/sysconf.h:157: undefined type, found `u_int32_t' ../includes/sysconf.h:160: undefined type, found `u_int32_t' ../includes/sysconf.h:161: undefined type, found `u_int16_t' ../includes/sysconf.h:162: undefined type, found `u_int16_t' ../includes/sysconf.h:165: undefined type, found `u_int8_t' ../includes/sysconf.h:166: undefined type, found `u_int8_t' ../includes/sysconf.h:288: undefined type, found `u_int8_t' ../includes/sysconf.h:289: undefined type, found `u_int8_t' ../includes/sysconf.h:329: undefined type, found `u_int32_t' ../includes/sysconf.h:360: undefined type, found `u_int32_t' ../includes/sysconf.h:477: undefined type, found `u_int16_t' ../includes/sysconf.h:478: undefined type, found `u_int16_t' ../includes/sysconf.h:492: undefined type, found `u_int8_t' ../includes/sysconf.h:770: undefined type, found `u_int32_t' ../includes/sysconf.h:771: undefined type, found `int32_t' ../includes/sysconf.h:772: undefined type, found `u_int16_t' ../includes/sysconf.h:773: undefined type, found `int16_t' ../includes/sysconf.h:774: syntax error, found `u_int32_t' ../includes/sysconf.h:774: illegal function definition, found `)' ../includes/sysconf.h:775: syntax error, found `int32_t' ../includes/sysconf.h:775: illegal function definition, found `)' ../includes/sysconf.h:776: syntax error, found `u_int16_t' ../includes/sysconf.h:776: illegal function definition, found `)' ../includes/sysconf.h:777: syntax error, found `int16_t' ../includes/sysconf.h:777: illegal function definition, found `)' ../includes/sysconf.h:781: syntax error, found `u_int32_t' ../includes/sysconf.h:781: illegal function definition, found `)' ../includes/sysconf.h:783: undefined type, found `u_int32_t' ../includes/sysconf.h:843: undefined type, found `u_int32_t' ../includes/sysconf.h:843: syntax error, found `u_int32_t' ../includes/sysconf.h:843: illegal function definition, found `)' ../includes/sysconf.h:844: undefined type, found `u_int32_t' ../includes/sysconf.h:847: syntax error, found `u_int32_t' ../includes/sysconf.h:847: illegal function definition, found `)' ../includes/sysconf.h:899: undefined type, found `u_int8_t' ../includes/sysconf.h:900: undefined type, found `u_int8_t' ../includes/sysconf.h:915: undefined type, found `u_int8_t' ../includes/sysconf.h:919: undefined type, found `u_int8_t' ../includes/sysconf.h:926: syntax error, found `struct' ../includes/sysconf.h:926: illegal function definition, found `)' *** Exit 1 Stop. *** Exit 1 Stop. which I am not sure how to solve OpenStep compiled the 'client' and 'server' but failed to compile the 'relay' Making all in relay cc -g -I.. -I../includes -c dhcrelay.c ./includes/cf/nextstep.h:73: warning: redefinition of macro _PATH_DHCLIENT_PID ./includes/cf/nextstep.h:72: warning: this is the location of the previous definition cc -o dhcrelay dhcrelay.o ../common/libdhcp.a /bin/ld: Undefined symbols: _lexchar _lexline _tlname _token_line *** Exit 1 Stop. *** Exit 1 Stop. Given the growing common use of DHCP I'd like to get a DHCP client at least for NeXTStep onto PEAK. TjL
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Re: future of mac programming Date: 5 Jun 1998 10:05:42 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6l8ftm$gdt$13@ns3.vrx.net> References: <6k61dt$ae5$1@crib.bevc.blacksburg.va.us> <1998052507544800.DAA25183@ladder01.news.aol.com> <6kbt81$1da$1@crib.bevc.blacksburg.va.us> <6kc2go$5tr$3@ns3.vrx.net> <jesjones-ya02408000R2605982143170001@news.accessone.com> <6kgtva$nc0$9@ns3.vrx.net> <jesjones-ya02408000R2705982033120001@news.accessone.com> <6kjagc$q9p$5@ns3.vrx.net> <rex-2805981519470001@192.168.0.3> <6kjtpo$9da$3@ns3.vrx.net> <rex-2805981924060001@192.168.0.3> <6km2cm$s3m$4@ns3.vrx.net> <rex-0106980320490001@192.168.0.3> <6ku7hc$g29$2@ns3.vrx.net> <rex-0106981800560001@192.168.0.3> <6kuml2$rhg$1@ns3.vrx.net> <rex-0106981924540001@192.168.0.3> <6kuucg$2uf$2@ns3.vrx.net> <rex-0106982229380001@192.168.0.3> <6l0es4$8th$4@ns3.vrx.net> <rex-0306980025070001@192.168.0.3> <6l44s5$5dc$7@ns3.vrx.net> <rex-0406980433240001@192.168.0.3> <6l5p9g$ebr$6@ns3.vrx.net> <rex-0406981708220001@192.168.0.3> <6l6jgl$4i6$2@ns3.vrx.net> <rex-0406982116430001@192.168.0.3> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: rex@smallandmighty.com In <rex-0406982116430001@192.168.0.3> Eric King claimed: > I think Be would very much like to have a generalized BShelf, but it > would make a lot more sense to wait until the new Translation and Media > kits are complete before creating one. What are these? And why do they have to call them "kits". It's like when people are ripping off OS they can't even be bothered to hide it. Heck. Taligent as a Responder. Grrrrr. > find the appropriate binary for a replicant and it couldn't understand the > embedded data, perhaps the OS could translate the data into something it > could understand. Let me guess, this will be called "TFilter"? ;-) > But what happens when you have binary data? We push it out to strings. Obj-C has a nice set of functions to do this. Hold on... ok, here's a portion of one of our documents that described the document layout itself: { windowFrame = "{x = 568; y = 289; width = 514; height = 490}"; zoomFactor = 1.000000; } Notice that the floats and ints become "english" text. Quite handy that, only six lines of code to build the document. > My problem with using XML > for this task, is that it doesn't handle binary well at all. Incidentally, > isn't NSPPL being replaced by something new? No one knows. Keep in mind the difference between a PL and a PPL, the former is a data format "in memory" (a subclass of NSData IIRC) but the later is a document format based on it. I kinda liked NSPPL's in theory, although I never actually used them personally. The missing link is that the NSCoder protocol does not collect enough info in order to build a PPL - it only includes enough to make an NSData. It would take a change to NSCoder to fix this, and I've long been of the opinion that they should. > Well, they are rather busy at the moment. You might have to wait until > the Mac OS X time frame or later. I also get the impression that anything > that sounds OpenDOCish is looked down upon. Yes, tainted by association. Still this no longer seems to be a death knell, KeyChains are finally back!!! > : How does one avoid namespace collisions? > > Collisions where? When coding down subclasses typically have the chance to right a "readable" name to the document. How does one avoid using the same one someone else did? It's not easy for the compiler to enforce this. > Yes, the BApplication class is really just a special BLooper. It > connects with the App Server and does some other setup work. Since it's > already running in a thread to begin with BApplications don't start a new > thread like other BLoopers. Ok. Under OS it's similar, but the NSRunLoop is a member of the app, as opposed to being a subclass of one. You can make your own and swap them and such. > A BLooper's reason for being, is to sit around and catch BMessages and > do something appropriate when it gets them. It can also send messages to > other BLoopers. Ok, now here I definitely prefer the Obj-C solution. Basically what they've done here is objectized their runtime. I don't know if I like that. Runtimes should do runtime things, objects should do object things. > I'm not quite sure what you're asking. Specifically what do you want to > do, and how do you do it under OpenStep? You just call the method. The runtime takes care of the rest. > BWindow is a subclass of BLooper. Ewwwwwww, I _definitely_ don't like that! > http://www.be.com/documentation/be_book/The%20Application%20Kit/Looper.html I'll give this a read over. > I think Be will eventually get around to doing the right thing. But > remember they're the 'Media OS' so they've got a lot of work ahead to live > up to that billing. This issue can be put off for a while. Sure, everyone's got priorities. > That may be so, but most Next developers were not making commercial > apps, they were developing inhouse apps. Now, yes. Maybe this was always on the books and got shuffled out? Maury
Subject: This bit me From: "Robert A. Decker" <comrade@umich.edu> Newsgroups: comp.sys.next.programmer MIME-Version: 1.0 Message-ID: <B19D84EF-30731@141.214.128.36> Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Date: Fri, 05 Jun 1998 15:16:05 GMT NNTP-Posting-Date: Fri, 05 Jun 1998 11:16:05 EDT I didn't think this would happen when you're just doing comparisons: The output of the following code is: oops. x>y correct. x isn't greater than z ************************** int x, z; unsigned int y; x = -97; y = 300; z = 300; if (x > y) { NSLog(@"oops. x>y"); } else { NSLog(@"correct. x isn't greater than y"); } if (x > z) { NSLog(@"oops. x>z"); } else { NSLog(@"correct. x isn't greater than z."); } *************************** rob -- <mailto: "Robert A. Decker" comrade@umich.edu> <http://hmrl.cancer.med.umich.edu/Rob/index.ssi> Programmer Analyst - Health Media Research Lab University of Michigan Comprehensive Cancer Center "Get A Life" quote #5: "In my heart I know you're right, but my perfectly functioning brain says you're a horse's ass." -Bob Elliott
From: Lars Immisch <lars@ibp.de> Newsgroups: comp.sys.next.programmer Subject: Re: This bit me Date: Fri, 05 Jun 1998 20:01:12 +0200 Organization: Immisch, Becker & Partner Message-ID: <35783268.45B8@ibp.de> References: <B19D84EF-30731@141.214.128.36> <slrn6ng56j.b0a.rog@talisker.ohm.york.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Roger Peppe wrote: -97; > > y = 300; > > if (x > y) { > > NSLog(@"oops. x>y"); > > that's one of the reasons i avoid unsigned ints like the plague. > what does one bit matter anyway? Because suddenly your maximum supported partition size is 2 Gig instead of 4? Reminds you of something? ;-) Cheers, Lars -- mailto:immisch@pobox.com http://pobox.com/~immisch Yesterdays yellow yoyo can make you yawn today
From: simpson@nospam.cts.com (Michael Simpson) Newsgroups: comp.sys.next.programmer Subject: Re: [Q] NSSplitView in a NSSplitView? Date: Fri, 05 Jun 1998 18:34:32 GMT Organization: QUALCOMM, Incorporated; San Diego, CA, USA Message-ID: <35783a08.352732522@news> References: <357837ba.352143024@news> On Fri, 05 Jun 1998 18:28:53 GMT, simpson@nospam.cts.com (Michael Simpson) wrote: >I'm new to IB and Objective C programming. I'm a seasoned commercial >developer , Mac and Windows, undertaking the transition to Yellow Box >and Objective C. > >First thing, I was trying to put a split view, with two views, and >another view into another splitview. When I do this, the split views >are stacked vertically. I need the inner split view to be vertical >and the other split view, with the inner split view to be horizontal. >How do I do this in IB? > >Also, is there a Faq for objective C, PB and IB available somewhere? Never mind on this. Scott Anquish had posted something elsewhere that gave me a start. Still need info on splitviews in splitviews. Michael >Thanks in advance, >Michael
From: "Charles W. Swiger" <chuck@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: This bit me Date: 5 Jun 1998 19:03:29 GMT Organization: Spacelab.net Internet Access Message-ID: <6l9fe1$s8m$2@news.spacelab.net> References: <B19D84EF-30731@141.214.128.36> "Robert A. Decker" <comrade@umich.edu> wrote: > I didn't think this would happen when you're just doing comparisons: > >The output of the following code is: >oops. x>y >correct. x isn't greater than z > >************************** >int x, z; >unsigned int y; > >x = -97; >y = 300; >z = 300; > >if (x > y) { > NSLog(@"oops. x>y"); >} else { > NSLog(@"correct. x isn't greater than y"); [ ... ] Welcome to the sometimes surprising world of digital arithmetic using two's-compliment representation of signed integers. There are some really good reasons why this happens, and it might be to your advantage to figure them out if you wish to program computers. Once you are capable of explaining to a third party what the N and V processor flags are, you should be fine.... :-) -Chuck Charles Swiger | chuck@codefab.com | standard disclaimer ---------------+-------------------+-------------------- "Microsoft: we make the easy almost impossible."
From: eric@skatter.USask.Ca Newsgroups: comp.sys.next.sysadmin,comp.sys.next.programmer Subject: Re: DHCP for OPENSTEP? Date: 5 Jun 1998 19:36:27 GMT Organization: University of Saskatchewan Message-ID: <6l9hbr$65a$1@tribune.usask.ca> References: <6kg47t$d8i@fridge.shore.net> <TyNd1.996$On1.3750778@ptah.visi.com> <UxSd1.254$BE5.1666092@news.rdc1.nj.home.com> I've got a package of source and fat (NeXT/Intel) executables. Note that you can't compile the Berkeley Packet Filter stuff on anything but NeXTSTEP 3.3. In particular, you can't compile any loadable-kernel stuff on OPENSTEP 4.2. Sigh. Anyhow, the executables are at: ftp://skatter.usask.ca/pub/eric/OPENSTEP_DHCP.tgz and the source is at ftp://skatter.usask.ca/pub/eric/OPENSTEP_DHCP_src.tgz -- Eric Norum eric@skatter.usask.ca Saskatchewan Accelerator Laboratory Phone: (306) 966-6308 University of Saskatchewan FAX: (306) 966-6058 Saskatoon, Canada.
From: "John F. Kennedy" <jfk@rlw.com> Newsgroups: comp.sys.next.programmer Subject: Where can YellowBox for Windows be obtained. Date: Fri, 5 Jun 1998 13:06:23 -0700 Message-ID: <6l9j58$s08$1@news0-alterdial.uu.net> I am a Windows programmer, and I am investigating the potential to do crossplatform programming with a Mac. I am evaluating Create which has been compiled for the YellowBox, but I can not find the necessary files. Where can I find YellowBox for Windows? Thanks, John Kennedy
From: rpk@lotatg.lotus.com (Robert P. Krajewski) Newsgroups: comp.sys.next.programmer Subject: Re: Where can YellowBox for Windows be obtained. Date: Fri, 05 Jun 1998 16:42:14 -0400 Organization: Lotus Development Corporation Message-ID: <rpk-0506981642150001@198.114.84.134> References: <6l9j58$s08$1@news0-alterdial.uu.net> qh6[f-]Im&PfyJ:K1?7L:,Bs^wp(!UpGZrVk^KgxrV62lp\}g8'UJF+ig;l`p."~|0gf S!%/I8*pUGvUFhUb70)wj9a%F^Xw]A7:?mA8#$q#A`(LNv5&(\^R#V*`~0wxu[,PMWkH y6&Cosh$;MSvhc:RL|:#H In article <6l9j58$s08$1@news0-alterdial.uu.net>, "John F. Kennedy" <jfk@rlw.com> wrote: >Where can I find YellowBox for Windows? I can't tell you where, but starting from www.developer.apple.com would be a good start. It must exist, because some software mentioned on comp.sys.next.announce can be used with it.
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: This bit me Date: 5 Jun 1998 15:54:54 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6ng56j.b0a.rog@talisker.ohm.york.ac.uk> References: <B19D84EF-30731@141.214.128.36> On Fri, 05 Jun 1998 15:16:05 GMT, "Robert A. Decker" <comrade@umich.edu> wrote: [edited for brevity] > I didn't think this would happen when you're just doing comparisons: > > The output of the following code is: > oops. x>y > > int x, z; > unsigned int y; > > x = -97; > y = 300; > if (x > y) { > NSLog(@"oops. x>y"); that's one of the reasons i avoid unsigned ints like the plague. what does one bit matter anyway? rog.
From: Stephen Peters <portnoy@ai.mit.edu> Newsgroups: comp.sys.next.programmer Subject: Re: POV Ray Sources??? Date: 05 Jun 1998 07:28:10 -0400 Organization: Massachvsetts Institvte of Technology Sender: portnoy@crispy-critters Message-ID: <us5yavc6qb9.fsf@ai.mit.edu> References: <lloyddean-0506980044320001@ip53.seattle7.wa.pub-ip.psi.net> lloyddean@earthlink.net (Lloyd D. Ollmann, Jr.) writes: > Does anyone know if/where I might find sources for > either a NextSTEP or OpenSTEP version of POV-Ray and > a viewer application? I've been putting one together -- it's in the very early stages now -- and am about to port it to OpenStep and Rhapsody. At the current moment it's pretty unstable, but if you can wait a while I might be able to put something on the web soon. I also have a simple version of povray which works like x-povray, displaying to the screen. That's binary-only, but it works. -- Stephen L. Peters portnoy@ai.mit.edu PGP fingerprint: BFA4 D0CF 8925 08AE 0CA5 CCDD 343D 6AC6 "Poodle: The other white meat." -- Sherman, Sherman's Lagoon
From: simpson@nospam.cts.com (Michael Simpson) Newsgroups: comp.sys.next.programmer Subject: [Q] NSSplitView in a NSSplitView? Date: Fri, 05 Jun 1998 18:28:53 GMT Organization: QUALCOMM, Incorporated; San Diego, CA, USA Message-ID: <357837ba.352143024@news> I'm new to IB and Objective C programming. I'm a seasoned commercial developer , Mac and Windows, undertaking the transition to Yellow Box and Objective C. First thing, I was trying to put a split view, with two views, and another view into another splitview. When I do this, the split views are stacked vertically. I need the inner split view to be vertical and the other split view, with the inner split view to be horizontal. How do I do this in IB? Also, is there a Faq for objective C, PB and IB available somewhere? Thanks in advance, Michael
From: Stephen Peters <portnoy@ai.mit.edu> Newsgroups: comp.sys.next.programmer Subject: Re: POV Ray Sources??? Date: 06 Jun 1998 00:24:46 -0400 Organization: Massachvsetts Institvte of Technology Sender: portnoy@crispy-critters Message-ID: <us5yavbkvht.fsf@ai.mit.edu> References: <lloyddean-0506980044320001@ip53.seattle7.wa.pub-ip.psi.net> <us5yavc6qb9.fsf@ai.mit.edu> Stephen Peters <portnoy@ai.mit.edu> writes: [...] > I also have a simple version of povray which works like x-povray, > displaying to the screen. That's binary-only, but it works. Actually, I just realized I *do* have the POV 1.0.2 diffs for this quick-and-dirty port of povray. If you like, I can send them to you. Send me email if you're interested. -- Stephen L. Peters portnoy@ai.mit.edu PGP fingerprint: BFA4 D0CF 8925 08AE 0CA5 CCDD 343D 6AC6 "Poodle: The other white meat." -- Sherman, Sherman's Lagoon
From: "scott nichol" <scottnichol@earthlink.net> Newsgroups: comp.sys.next.programmer Subject: gcc 2.7.2 on DR2 (build problems)... Date: Sat, 06 Jun 1998 08:36:38 -0400 Organization: EarthLink Network, Inc. Message-ID: <6lbd6b$5ro$1@bolivia.it.earthlink.net> Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit hello all: i've recently downloaded the gcc package from apple's web site and tried to install it on my DR2. i've run the configure script with host=powerpc-apple-rhapsody and then a make for LANGUAGES="c c++ objective-c", but it keeps failing. it seems that the distribution installs two libraries, one called cp (which is were the c++ lexical analyzer is stored (lex.c))and one called obcp (which is where the objective-c++ lexcial analyzer is stored). but the problem seems to be that the lex.c in the objective-c++ directory just includes the lex.c from the c++ directory. this causes conflicts when trying to compile it because there are lexical tokens (OBJECTNAME) that are referenced in the objective-c++ parser that are not defined in the c++ lexical analyzer. the following compile error occurs: cc -traditional-cpp -c -DIN_GCC -DOBJCPLUS -g -DNEXT_RELEASE_MAJOR=`((hostinfo | egrep 'NeXT|Rhapsody'; echo $RC_RELEASE) | tr '\12' ' '; echo $RC_OSVERSION) | sed -e 's/.*\([0-9]\)\.[0-9][JRm:].*/\1/'` -DNEXT_RELEASE_MINOR=`((hostinfo | egrep 'NeXT|Rhapsody'; echo $RC_RELEASE) | tr '\12' ' '; echo $RC_OSVERSION) | sed -e 's/.*[0-9]\.\([0-9]\)[JRm:].*/\1/'` -DAPPLE_CC=`cd .; vers_string -f cc | sed -e 's/[-A-Za-z_]*//' | sed -e 's/\.[0-9.]*//'` -I. -I.. -I. -I./.. -I./../config gc.c cc -traditional-cpp -c -DIN_GCC -DOBJCPLUS -g -DNEXT_RELEASE_MAJOR=`((hostinfo | egrep 'NeXT|Rhapsody'; echo $RC_RELEASE) | tr '\12' ' '; echo $RC_OSVERSION) | sed -e 's/.*\([0-9]\)\.[0-9][JRm:].*/\1/'` -DNEXT_RELEASE_MINOR=`((hostinfo | egrep 'NeXT|Rhapsody'; echo $RC_RELEASE) | tr '\12' ' '; echo $RC_OSVERSION) | sed -e 's/.*[0-9]\.\([0-9]\)[JRm:].*/\1/'` -DAPPLE_CC=`cd .; vers_string -f cc | sed -e 's/[-A-Za-z_]*//' | sed -e 's/\.[0-9.]*//'` -I. -I.. -I. -I./.. -I./../config lex.c In file included from ../cp/../obcp/hash.h:9, from ../cp/lex.c:307, from lex.c:1: ../cp/../obcp/hash-next.h:154: `OBJECTNAME' undeclared here (not in a function) ../cp/../obcp/hash-next.h:154: too many errors, bailing out make[1]: *** [lex.o] Error 1 make: *** [cc1objplus] Error 2 localhost:130# so, am i doing something wrong with the configuration? or is it just that the distribution from apple does not work? if you've had success with this, please let me know what i'm doing wrong. thanks, scott nichol
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: This bit me Date: 6 Jun 1998 13:20:34 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6nigh6.bsv.rog@talisker.ohm.york.ac.uk> References: <B19D84EF-30731@141.214.128.36> <slrn6ng56j.b0a.rog@talisker.ohm.york.ac.uk> <35783268.45B8@ibp.de> On Fri, 05 Jun 1998 20:01:12 +0200, Lars Immisch <lars@ibp.de> wrote: > > that's one of the reasons i avoid unsigned ints like the plague. > > what does one bit matter anyway? > > Because suddenly your maximum supported partition size is 2 Gig instead > of 4? Reminds you of something? ;-) that's an unusual case and one that should really be dealt with by using 64-bit ints. 4 gig is still too small... of course when we're pushing the 10 exabyte limit, i might have cause to think again, but i doubt it! rog.
From: cejensen@winternet.com (Christian Jensen) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.programmer Subject: Re: DHCP for OPENSTEP? Date: 6 Jun 1998 18:40:45 GMT Organization: TJP, Inc. Message-ID: <6lc2fd$ji1$1@blackice.winternet.com> References: <6kg47t$d8i@fridge.shore.net> <TyNd1.996$On1.3750778@ptah.visi.com> <UxSd1.254$BE5.1666092@news.rdc1.nj.home.com> <6l9hbr$65a$1@tribune.usask.ca> NNTP-Posting-Date: 6 Jun 1998 18:40:45 GMT You (eric@skatter.USask.Ca <eric@skatter.USask.Ca>) wrote > Note that you can't compile the Berkeley Packet Filter stuff on anything but > NeXTSTEP 3.3. In particular, you can't compile any loadable-kernel stuff on > OPENSTEP 4.2. Sigh. Did you compile only the BPF/LKS stuff on NS 3.3? Or all of it, in which case I can assume that this will work on NEXTSTEP 3.3 as well as OpenStep? Thanks, --Chris ************************** Chris Jensen cejensen@winternet.com MIME, Sun, NeXTMail OK "Sacred cows make the best hamburger." --Mark Twain
From: nunya@baker.com Newsgroups: comp.sys.next.programmer Date: Sat, 06 Jun 1998 12:00:03 PDT Subject: For YOU!! Organization: Email Platinum v.3.1b Message-ID: <3578c187.0@news> --------------------------------------------------------------------------------------------------------- This message is sent 'as a paid for service' by Johnson Advertising. Our Clients pay for advertising on a per response basis. --------------------------------------------------------------------------------------------------------- One of our Clients is giving away a free phone card worth 50 free minutes to all United States residents who send a #10 self-addressed stamped envelope to Johnson Advertising Post Office Box 309 Mount Union, Pennsylvania 17066
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <3578c187.0@news> Control: cancel <3578c187.0@news> Date: 06 Jun 1998 19:58:43 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.3578c187.0@news> Sender: nunya@baker.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: arti@lava.DOTnet (Art Isbell - remove "DOT") Newsgroups: comp.sys.next.programmer Subject: Re: This bit me Date: 7 Jun 1998 04:08:54 GMT Organization: LavaNet, Inc. Distribution: world Message-ID: <6ld3om$1uq@mochi.lava.net> References: <B19D84EF-30731@141.214.128.36> <slrn6ng56j.b0a.rog@talisker.ohm.york.ac.uk> rog@ohm.york.ac.uk (Roger Peppe) wrote: > On Fri, 05 Jun 1998 15:16:05 GMT, "Robert A. Decker" <comrade@umich.edu> wrote: > [edited for brevity] > > I didn't think this would happen when you're just doing comparisons: > > > > The output of the following code is: > > oops. x>y > > > > int x, z; > > unsigned int y; > > > > x = -97; > > y = 300; > > if (x > y) { > > NSLog(@"oops. x>y"); > > that's one of the reasons i avoid unsigned ints like the plague. > what does one bit matter anyway? I bit doesn't usually matter, but I prefer to specify unsigned if the value of the integer cannot be negative for documentation reasons. AppKit, Foundation, etc. do the same with sizes, for example, as unsigned integers. I also prefer to choose the size of the integer type based on the valid range of values. I realize that C-based languages don't enforce these choices so problems can occur as a result, but there are ways to avoid most of these problems. Objective-C is wonderfully self-documenting as is, so I like to extend this as much as possible because I don't tend to write very good documentation otherwise :-) -- Art Isbell NeXT/MIME Mail: arti@lavaDOTnet IDX Systems Corporation Voice/Fax: +1 808 394 0511 (for whom I don't speak) Voice Mail: +1 808 394 0495 Healthcare Info Technology US Mail: Honolulu, HI 96825-2638
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <3654896587219@digifix.com> Date: 7 Jun 1998 03:49:08 GMT Organization: Digital Fix Development Message-ID: <3379897192022@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1993. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: nospam+nospam+nospam+this+works@luomat.peak.org (Timothy J Luoma) Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: /NextDeveloper/Headers/bsd/netinet/ip.h:127: undefined type, found `n_long' Followup-To: comp.sys.next.programmer MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <Kwoe1.350$BE5.2545918@news.rdc1.nj.home.com> Date: Sun, 07 Jun 1998 04:18:50 GMT NNTP-Posting-Date: Sat, 06 Jun 1998 21:18:50 PDT Organization: @Home Network NOTE: FOLLOWUPS TO COMP.SYS.NEXT.PROGRAMMER This one has bitten me on a few files I've tried to compile lately /NextDeveloper/Headers/bsd/netinet/ip.h:127: undefined type, found `n_long' /NextDeveloper/Headers/bsd/netinet/ip.h:130: undefined type, found `n_long' The file refers to ``n_long'' but apparently doesn't initialize or define it? Can someone help with a workaround? I checked dejanews and only found others in the similar position, no answers TjL
From: Trevin Beattie <*trevin*@*xmission*.*com*> Newsgroups: comp.sys.next.programmer Subject: Re: This bit me Date: Sun, 07 Jun 1998 12:26:31 +0000 Organization: XMission Internet (801 539 0852) Message-ID: <6le0eo$a2q$1@news.xmission.com> References: <B19D84EF-30731@141.214.128.36> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Robert A. Decker wrote: > > I didn't think this would happen when you're just doing comparisons: > > The output of the following code is: > oops. x>y > correct. x isn't greater than z > > ************************** > int x, z; > unsigned int y; > > x = -97; > y = 300; > z = 300; ... That reminds me of a bunch of library code I had written, which when I tried to compile resulted in several warnings about mixing signed and unsigned types (in my case, int vs. size_t). I knew the values of the numbers _shouldn't_ be negative in all cases, nor should they be greater than 2^31, but it's still bad coding (at least in the compiler's opinion!) The problem is that the complete range of values supported by both signed and unsigned types cannot be represented by either one alone, so the compiler must choose which type to use when doing arithmetic or making comparisons. In your case, we see that it changes signed ints to unsigned, so -97 becomes 4294967199 -- definitely greater than 300! The 'correct' solution is to typecast one of the numbers to however you want the comparison to be done, such as: if (x > (signed int) y) On the other hand, if negative values are legal for x and values greater than 2^31 are legal for y, you would probably want to do another test, such as: if ((x >= 0) && /* negative is always less than unsigned */ or: if ((y <= INT_MAX) && /* > INT_MAX is always greater than signed */ -- Trevin Beattie "Do not meddle in the affairs of wizards, *To*reply*to*this* for you are crunchy and good with ketchup." *message,*remove*the* --unknown *asterisks*from*my*email*address.*
From: thf00@hotmail.com (Thomas F. Unke) Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: Re: /NextDeveloper/Headers/bsd/netinet/ip.h: 127: undefined type, found `n_long' Date: Sun, 7 Jun 1998 12:50:46 GMT Organization: Dogbert Consulting Sender: thomas@gamelan.okay.net (thomas) Message-ID: <1998Jun7.125046.850@gamelan.okay.net> References: <Kwoe1.350$BE5.2545918@news.rdc1.nj.home.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Cc: nospam+nospam+nospam+this+works@luomat.peak.org In <Kwoe1.350$BE5.2545918@news.rdc1.nj.home.com> Timothy J Luoma wrote: > NOTE: FOLLOWUPS TO COMP.SYS.NEXT.PROGRAMMER > > > This one has bitten me on a few files I've tried to compile lately > > /NextDeveloper/Headers/bsd/netinet/ip.h:127: undefined type, found `n_long' > /NextDeveloper/Headers/bsd/netinet/ip.h:130: undefined type, found `n_long' > > The file refers to ``n_long'' but apparently doesn't initialize or define it? Problems like this can be solved quickly with: /usr/include> grep n_long **/*.h architecture/i386/alignment.h:get_align_long(void *ivalue) architecture/i386/alignment.h:put_align_long(unsigned long ivalue, void *ovalue) architecture/m68k/alignment.h:get_align_long(void *ivalue) architecture/m68k/alignment.h:put_align_long(unsigned long ivalue, void *ovalue) bsd/netinet/in_systm.h:typedef u_long n_long; /* long as received from the net */ bsd/netinet/ip.h: n_long ipt_time[1]; bsd/netinet/ip.h: n_long ipt_time; Thus we see that n_long is defined in # include <netinet/in_systm.h>
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <oYkApJek9GA.127@news.powernethk.com> Control: cancel <oYkApJek9GA.127@news.powernethk.com> Date: 08 Jun 1998 01:42:51 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.oYkApJek9GA.127@news.powernethk.com> Sender: tsevbeci@hello.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: This bit me Date: 8 Jun 1998 11:00:29 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6nnh2g.cpr.rog@talisker.ohm.york.ac.uk> References: <B19D84EF-30731@141.214.128.36> <slrn6ng56j.b0a.rog@talisker.ohm.york.ac.uk> <6ld3om$1uq@mochi.lava.net> On 7 Jun 1998 04:08:54 GMT, Art Isbell - remove "DOT" <arti@lava.DOTnet> wrote: > rog@ohm.york.ac.uk (Roger Peppe) wrote: > > On Fri, 05 Jun 1998 15:16:05 GMT, "Robert A. Decker" <comrade@umich.edu> > wrote: > > > x = -97; > > > y = 300; > > > if (x > y) { > > > NSLog(@"oops. x>y"); > > > > that's one of the reasons i avoid unsigned ints like the plague. > > what does one bit matter anyway? > > I bit doesn't usually matter, but I prefer to specify unsigned if the > value of the integer cannot be negative for documentation reasons. AppKit, > Foundation, etc. do the same with sizes, for example, as unsigned integers. i know that this would seem to be the obvious way to go about things, but little gotchas like the one above mean that it's generally less hassle to use signed values throughout. if the code i write depends on certain properties of a value (e.g. greater than zero, in a certain range, etc) then i will use assert to make sure that that's the case. for instance: int x; x = fn(); assert(x >= 0); is likely to lead to far less problems should a buggy fn() return a negative value than: unsigned x; x = fn(); asserts are a great way of making "active documentation" that tells you immediately something goes wrong... cheers, rog.
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: [Q] NSTask and NSTaskDidTerminateNotification Date: Mon, 08 Jun 1998 12:34:07 +0200 Organization: Square B.V. Message-ID: <357BBE1F.C0ADF37@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I'm having the Monday morning blues today. I'm trying to do such a simple thing but it won't work and I can't understand why. :-( I create a NSTask and launch it. When the task has terminated a new task may be scheduled, depending on an array. This morning I did _not_ receive the NSTaskDidTerminateNotification. I could not understand why. After some time I *did* receive the notification. I did not understand why, but what the heck. Guess, now I don't receive the notification anymore and I'm lost somewhere. I keep asking myself what did I change that triggers this behaviour. Why does an NSTask sometimes send the notification and sometimes not? Anybody? Some small code snippets, the object contains an NSTask instance variable 'task' and an NSMutableArray called 'entries'. If I uncomment the 'waitUntilExit' everything works as excpected. - (void)_launchConvertorWithPath:(NSString *)launchPath arguments:(NSArray *)arguments { NSParameterAssert(launchPath != nil); NSParameterAssert(arguments != nil); NSAssert1([ [NSFileManager defaultManager] fileExistsAtPath: launchPath], @"External convertor not found at path %@",launchPath); NSAssert1([ [NSFileManager defaultManager] isExecutableFileAtPath: launchPath], @"External convertor is not executable at path %@",launchPath); task = [ [ [NSTask alloc] init] retain]; [task setLaunchPath: launchPath]; [task setArguments: arguments]; [ [NSNotificationCenter defaultCenter] addObserver: self selector: @selector(taskDidTerminate:) name: NSTaskDidTerminateNotification object: task]; [task launch]; //[task waitUntilExit]; NSAssert([task isRunning],@"Task not running."); } - (void)taskDidTerminate:(NSNotification *)notification { NSAssert([notification object] == task,@"Received notification of other task."); [ [NSNotificationCenter defaultCenter] removeObserver: self name: NSTaskDidTerminateNotification object: task]; task = nil; // Tasks can only be used once. if([entries count] > 0) [self _processNextEntry]; } -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto:mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+ Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" and "to trick into believing or accepting as genuine something false and often preposterous"
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: cancel <357BBE1F.C0ADF37@Square.nl> Control: cancel <357BBE1F.C0ADF37@Square.nl> Date: Mon, 08 Jun 1998 17:10:16 +0200 Organization: NLnet Message-ID: <6lguu4$gl$1@news.utrecht.NL.net> References: <357BBE1F.C0ADF37@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit This message was cancelled from within Mozilla.
From: Eric Hermanson <eric@alum.mit.edu> Newsgroups: comp.sys.next.programmer Subject: Java Yucky Date: Mon, 08 Jun 1998 11:07:51 -0700 Organization: Digital Universe Corporation Message-ID: <357C2877.880F61D9@alum.mit.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NNTP-Posting-Date: 8 Jun 1998 18:08:29 GMT I've been diving into Java and I've noticed some things I don't like: 1. No header files? All code shoved into one file!? What's up with that? Are header files a possibility in Java, or is one expected to use an external application like JavaBrowser to view the API hierarchy? 2. Because there are no header files, and because of full method signatures missing from the Java syntax, one is forced to look at the documentation to figure out what a method and it's corresponding arguments do. JavaBrowser isn't even really helpful in this respect unless you already know the API by heart. At least with Obj-C I can sort of figure out what the methods do by reading the method name without reading the documentation. 3. What's all this 'getXXX' stuff? In OpenStep, the accessor methods are simply: variableName, setVariableName. In Java, people seem to take the "C++" approach by using the convention: getVariableName, setVariableName. What's up with that? When I think of 'get', I think in terms of putting something into a buffer (getCString) - I don't tend to think in terms of accessing a variable. Is there an advantage to the 'get' approach that I'm missing? 4. No categories. No posing. Enough said. 5. Java Syntax. One word: yuck! About the only useful things I've seen from Java so far are the *.net libraries (I wish these were available under Obj-C), and the fact that you can compile your code once and ship it everywhere. Oh how I wish Apple would make an Obj-C virtual machine, or a Obj-C--> Java bytecode compiler (not being a compiler person, these things might be impossible, but if not, I'd like to see them). Eric
From: "Michelle L. Buck" <buck.erik@mcleod.net> Newsgroups: comp.sys.next.programmer Subject: Re: Java Yucky Date: Mon, 8 Jun 1998 20:13:32 -0500 Organization: McleodUSA - http://www.mcleodusa.net Message-ID: <6li2b5$12t22@odie.mcleod.net> References: <357C2877.880F61D9@alum.mit.edu> Eric Hermanson wrote in message <357C2877.880F61D9@alum.mit.edu>... >I've been diving into Java and I've noticed some things I don't like: > >1. No header files? All code shoved into one file!? What's up with that? Are >header files a possibility in Java, or is one expected to use an external >application like JavaBrowser to view the API hierarchy? > Java fans site this as an advantage and YES. >2. Because there are no header files, and because of full method signatures >missing from the Java syntax, one is forced to look at the documentation to >figure out what a method and it's corresponding arguments do. JavaBrowser isn't >even really helpful in this respect unless you already know the API by heart. >At least with Obj-C I can sort of figure out what the methods do by reading the >method name without reading the documentation. > This is unfortunate. >3. What's all this 'getXXX' stuff? In OpenStep, the accessor methods are >simply: variableName, setVariableName. In Java, people seem to take the "C++" >approach by using the convention: getVariableName, setVariableName. What's up >with that? When I think of 'get', I think in terms of putting something into a >buffer (getCString) - I don't tend to think in terms of accessing a variable. >Is there an advantage to the 'get' approach that I'm missing? > This is just different style. Change your expectations. This naming of accessors is at least as reasonable as NeXT's. The important thing is consistency. >4. No categories. No posing. Enough said. These deficiencies may or may not be security related. I will miss categories! > >5. Java Syntax. One word: yuck! > In one word marketing. Objective-C designers wisely chose to accentuate the difference between a message and a structure derefference/function pointer dereference because they are semantically different. In early C++ there was no semantic difference and the exposure of implementation mechanism lives to this day. C++ clearly had the mind share so Sun made Java look like C++ in order to be comfortable for C++ programmers. The irony is that Java is semantically more similar to Objective-C and the similarity between Java syntax and C++ syntax (in spite of different semantics) can be very misleading. > >About the only useful things I've seen from Java so far are the *.net libraries >(I wish these were available under Obj-C), and the fact that you can compile >your code once and ship it everywhere. Oh how I wish Apple would make an Obj-C >virtual machine, or a Obj-C--> Java bytecode compiler (not being a compiler >person, these things might be impossible, but if not, I'd like to see them). > >Eric > The Beans API is awesome. It is better than NeXT's IB palette concepts in implementation. This is one area where we should all copy/imitate Java. Java also neatly solves name space issues still lurking in Objective-C.
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <357b6791.1@news> Control: cancel <357b6791.1@news> Date: 09 Jun 1998 03:37:39 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.357b6791.1@news> Sender: nunya@baker.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: Re: Java Yucky Date: Tue, 09 Jun 1998 08:48:52 +0200 Organization: Square B.V. Message-ID: <357CDAD4.C5866363@Square.nl> References: <357C2877.880F61D9@alum.mit.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Eric Hermanson <eric@alum.mit.edu> > > 1. No header files? All code shoved into one file!? What's up with that? Are > header files a possibility in Java, or is one expected to use an external > application like JavaBrowser to view the API hierarchy? Personally I think that is great. It is like the module concept of Delphi. When you deploy a library you only have to release the (compiled) class files, not the headers. In Objective-C I always forget to synchronize the header files due to time pressure. I've never had that problem in Java. You annotate the documentation within the code and run the doctool and your documentation is ready! > 2. Because there are no header files, and because of full method signatures > missing from the Java syntax, one is forced to look at the documentation to > figure out what a method and it's corresponding arguments do. JavaBrowser isn't > even really helpful in this respect unless you already know the API by heart. > At least with Obj-C I can sort of figure out what the methods do by reading the > method name without reading the documentation. In Objective-C you have to look at the .h files _or_ the documentation, in Java you look at the html documentation with all classes hyperlinked. There's no difference. > 3. What's all this 'getXXX' stuff? In OpenStep, the accessor methods are > simply: variableName, setVariableName. In Java, people seem to take the "C++" > approach by using the convention: getVariableName, setVariableName. What's up > with that? When I think of 'get', I think in terms of putting something into a > buffer (getCString) - I don't tend to think in terms of accessing a variable. > Is there an advantage to the 'get' approach that I'm missing? No advantage. Just different approaches. > 4. No categories. No posing. Enough said. Security reasons. Java is marketed as 'Secure' and the Virtual Machine controls execution by a SecurityManager. Especially posing may intruduce all kinds of risks. > 5. Java Syntax. One word: yuck! When I started to do Java programming I really missed the named parameters. I like the Objective-C _syntax_ more than the Java syntax. But as the customer will _never_ see the program's code, and it doesn't even _care_ about the language-syntax I look at what can be done with the language and the libraries. Although there is still a speed problem I'm impressed with the libraries. > About the only useful things I've seen from Java so far are the *.net libraries *Most* things that are in the OpenStep libraries are in the Java libraries. Maurice le Rutte. P.S. I'm afraid this will end in a "I love/hate Java" thread. -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto:mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+ Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" and "to trick into believing or accepting as genuine something false and often preposterous"
From: "John Zollinger" <john.zollinger@arkona.com> Newsgroups: comp.sys.next.programmer Subject: Re: Java Yucky Date: Tue, 9 Jun 1998 08:55:40 -0600 Organization: Arkona Message-ID: <6ljhss$m4j$1@news.xmission.com> References: <357C2877.880F61D9@alum.mit.edu> <357CDAD4.C5866363@Square.nl> Maurice le Rutte wrote in message <357CDAD4.C5866363@Square.nl>... [...] >> 2. Because there are no header files, and because of full method signatures >> missing from the Java syntax, one is forced to look at the documentation to >> figure out what a method and it's corresponding arguments do. JavaBrowser isn't >> even really helpful in this respect unless you already know the API by heart. >> At least with Obj-C I can sort of figure out what the methods do by reading the >> method name without reading the documentation. > >In Objective-C you have to look at the .h files _or_ the documentation, >in Java you look at the html documentation with all classes hyperlinked. >There's no difference. Wrongo. Objective-C code is somewhat self documenting, Java code is not. For example, which of these statements is more clear and don't require you to look at the docs to change the height param? Java: foo.setSize(10, 20, 20, 100, 50, 50); Objective-C: [foo setX: 10 y: 20 z:20 width: 100 height: 50 depth: 50]; With the Java one, I have no choice but to go look at the documentation. Even though I may be able to guess which one is the height, I can't be sure. With the Objective-C example, even if I had no idea what class foo was, if I needed to change the width, I could do it in about 2 seconds, with confidence, and without looking at any documentation. Also, with this syntax, it makes it much more helpful to search through code for specific things. With Java, I have to know pretty much exactly what method I'm after before I start searching. Sure you have to type a bit more with Objective-C, but man, it is much easier to read after you write it. [...] >> 5. Java Syntax. One word: yuck! > >When I started to do Java programming I really missed the named >parameters. I like the Objective-C _syntax_ more than the Java syntax. >But as the customer will _never_ see the program's code, and it doesn't >even _care_ about the language-syntax I look at what can be done with >the language and the libraries. Although there is still a speed problem >I'm impressed with the libraries. True, but support/maintenance costs your customer does. I think the support/maintenance costs of Java are far greater than Objective-C. But, of course, I have no numbers to back that up. :-) >> About the only useful things I've seen from Java so far are the *.net libraries > >*Most* things that are in the OpenStep libraries are in the Java >libraries. Yes. But the OPENSTEP libraries aren't nearly as buggy, and they are more cleanly designed. :-) AND, the format of NeXT/Apple's documentation is far superior. >Maurice le Rutte. > >P.S. I'm afraid this will end in a "I love/hate Java" thread. Spoke too soon. While I program in Java about 90% of my time now, I still miss Objective-C a lot, it is a great language. Even more so, I miss AppKit, EOF, and Interface Builder (not to mention the OS). :-) I definitely feel spoiled under OPENSTEP, I feel punished in Java. Someday, perhaps if I'm good... Happy coding! John Zollinger Arkona, Inc. john.zollinger@arkona.com
Subject: Re: Java Yucky From: "Robert A. Decker" <comrade@umich.edu> Newsgroups: comp.sys.next.programmer References: <357CDAD4.C5866363@Square.nl> MIME-Version: 1.0 Message-ID: <B1A2C7DF-3C612@141.214.128.36> Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Date: Tue, 09 Jun 1998 15:02:55 GMT NNTP-Posting-Date: Tue, 09 Jun 1998 11:02:55 EDT What I want to know is why the Java compilers are so damn messed up! Java is a fully dynamic language with runtime lookup when you're sending messages to objects. However, those damn compilers are always making me type everything all over the damn place! It's ugly and doubles the amount of code I have to write. Another thing that increases the amount of code is always checking to make sure an object isn't null! rob -- <mailto: "Robert A. Decker" comrade@umich.edu> <http://hmrl.cancer.med.umich.edu/Rob/index.ssi> Programmer Analyst - Health Media Research Lab University of Michigan Comprehensive Cancer Center "Get A Life" quote #8: "Wow. A doctor. Maybe you can help me with another little problem I'm having. What exactly does it mean when you wake up every morning in a pool of your own vomit?" -Chris Elliott
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.next.programmer Subject: Re: Java Yucky Date: 9 Jun 1998 17:15:46 GMT Organization: MiscKit Development Message-ID: <6ljqk2$87j$1@news.xmission.com> References: <357C2877.880F61D9@alum.mit.edu> <357CDAD4.C5866363@Square.nl> <6ljhss$m4j$1@news.xmission.com> "John Zollinger" <john.zollinger@arkona.com> wrote: > Sure you have to type a bit more with Objective-C, but man, it is much > easier to read after you write it. You could even say that the time spent over ObjC in typing Java method calls is more than made up for by the time spent searching the Java docs. And then some! Then again, maybe you don't save any typing after all: "Robert A. Decker" <comrade@umich.edu> wrote: > Another thing that increases the amount of code is always checking > to make sure an object isn't null! So in the final analysis, IMHO Java doesn't save any typing at all. <sarcasm> All it saves you from is self-documenting, maintainable code. And thank heavens it also saves you from time-tested, elegantly designed APIs. I'd hate to have to admit that my buggy programs were a result of bugs in MY code; I just _love_ being able to point at buggy VMs and frameworks instead! </sarcasm> I used Java for a while. I'm using Objective-C again, and there's a reason. Java is simply too immature--buggy, lacking important features, and the design is inconsistent and inelegant. That said, it does show promise and ten years from now may be good enough to compete with Objective-C on technical grounds instead of being the winner just because of a slick marketing veneer. There are several things about Java I do like very much; I just wish there were more of them. :-) How's this for faint praise: it is some of the best software I've ever seen leave Sun's premises. -- Later, -Don Yacktman don@misckit.com <a href="http://www.misckit.com/don.html">My home page</a>
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: How to catch "message release sent to freed object=0xXXXXXX" Date: Wed, 10 Jun 1998 11:45:10 +0200 Organization: Square B.V. Message-ID: <357E55A6.C248192F@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Is it possible to catch the "objc: FREED(id): message release sent to freed object=0x492870" exception? On Windows-NT the user gets an awful '0x0000DEAD' message on the screen. I'd like to catch it and present it in a more 'humane' way. Maurice le Rutte -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto:mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+ Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" and "to trick into believing or accepting as genuine something false and often preposterous"
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: Java Yucky Date: 10 Jun 1998 07:49:52 GMT Organization: All USENET -- http://www.Supernews.com Message-ID: <6lldr0$j8r$1@supernews.com> References: <357C2877.880F61D9@alum.mit.edu> <357CDAD4.C5866363@Square.nl> <6ljhss$m4j$1@news.xmission.com> <6ljqk2$87j$1@news.xmission.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: don@misckit.com Don Yacktman may or may not have said: [stuff about java] -> How's this for faint praise: it is some of the best software I've ever seen -> leave Sun's premises. Sad to say, I can point to one item that left Sun's premises, (which was also done primarily by James Gosling,) which was nothing short of brilliant: The Network Extensible Window System (NeWS). Unfortunately, the only GUI that Sun ever implemented using this marvelous, elegant platform was OpenLook (talk about wretched!) Now, if we could just get our hands on the best NeWS implementation there ever was (rumor has it that it was the version that SGI did) and build GNUStep on it, then we'd be a lot farther on our quest for the holy grail of development systems. Sun used to sell great sofware. They lost their way, right around the time they lost John Gilmore. sys V, X Windows, Java... Mediocre crap, all of it. -jcr
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: am i mad or has NSTimer got a nasty bug? Date: 10 Jun 1998 07:58:13 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6nsf4m.ff1.rog@talisker.ohm.york.ac.uk> i've been using an NSTimer to schedule events every so often (i want to send 200 events/second) and there's something very odd going on, because it really doesn't like scheduling events between 100 and 1000 times a second. if i schedule an event to occur every 0.01 seconds, no problem, i get about 99.4 events/sec. if i schedule an event to occur every 0.001 seconds, also no problem, i get about 985 events/second. however if i try most values in between, i just get 100 events/sec. is this a known bug, or a nasty feature? here's some code that demonstrates the problem (just compile with "cc tsttimer.m -framework AppKit"). grateful for any workarounds! cheers, rog. --------------- tsttimer.m snip here ----------------------- #import <AppKit/AppKit.h> @implementation TimerHandler: NSObject { NSTimer *timer; NSDate *starttime; } // list of timer intervals to try out. double timevals[] = {0.1, 0.01, 0.005, 0.0015, 0.0011, 0.001, 0}; #define NTIMEVALS (sizeof(timevals) / sizeof(timevals[0])) int idx; static int count; - (void)startTimer { if (idx >= NTIMEVALS) exit(0); printf("starting timer with interval %g (%g events/sec)\n", timevals[idx], 1/timevals[idx]); timer = [NSTimer scheduledTimerWithTimeInterval:timevals[idx++] target:self selector:@selector(doEvent:) userInfo:nil repeats:YES]; [NSTimer scheduledTimerWithTimeInterval:5 target:self selector:@selector(nextTimer:) userInfo:nil repeats:NO]; starttime = [[NSDate date] retain]; count = 0; } - (void)nextTimer:sender { NSTimeInterval elapsed = -[starttime timeIntervalSinceNow]; printf("%g events/sec\n", count/elapsed); [timer invalidate]; timer = nil; [starttime release]; starttime = nil; [self startTimer]; } - (void)doEvent:sender { count++; } @end int main(int argc, char **argv) { NSAutoreleasePool *pool; id menu, item, obj; pool = [[NSAutoreleasePool alloc] init]; [NSApplication sharedApplication]; obj = [[TimerHandler alloc] init]; [obj startTimer]; [NSApp run]; [pool release]; return 0; }
From: cms@macisp.net (cms) Newsgroups: comp.sys.next.programmer,comp.sys.next.software Subject: Help Needed with ( CAPer ) Date: Wed, 10 Jun 1998 08:40:03 -0400 Organization: CMS Message-ID: <cms-1006980840040001@209.26.71.138> We are having printing problem using the terminal window to print.. Frank doesn't know what wrong..with CAPer.. and why its doing this. I can print from the GUI.. But can't print from the terminal window. very strange.. We haven't hear back from frank a 2 days..regarding this problem. please help. Richard
From: ahoesch@smartsoft.de Newsgroups: comp.sys.next.programmer Subject: SmartBase 1.5b for OPENSTEP Mach Date: 10 Jun 1998 13:59:57 GMT Organization: Offenes Netz Luebeck e.V. Message-ID: <6lm3gt$5gn@merkur.on.on-luebeck.de> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit SUBMISSION SmartBase 1.5b for OPENSTEP ftp.smartsoft.de/SmartBase/SmartBase.1.5b.I.b.tar.gz SmartBase has been developed to simplify the development and maintenance of database dependent software systems based on OPENSTEP, Apples new next generation operating system. This goal is achieved by centralizing the business logic for more flexibility, by providing support for remote object instanciation in order to support the design of client/server architectures and by introducing an advanced object paradigm allowing to create data models at a very high abstraction level. Š Centralized business logic Š Remote object support Š Advanced object paradigm SmartBase does not provide data storage mechanisms on its own but makes use of an ordinary relational database management system for this purpose. You can think of SmartBase as an objective extension for your database. Because SmartBase is based on Apples Enterprise Objects Framework it principly can be used with any RDBMS, that's supported by EOF or comes with an EOF-Adapter. SmartBase is built for Intel. You need to have OPENSTEP Mach and EOF installed to use this product. Moreover you need a RDBMS. The current release only supports OpenBase (Version 5.1.8 or higher). See http://www.openbase.com to get a demo version of OpenBase for Mach. RELEASE NOTES Changes for version 1.5b Release ---------------------------------------------- - The role/history mechanism has completely been redesigned. Both, roles and history attributes can now reasonably be used for qualification on SQL level by making use of views. Moreover you can cause an object in memory to reflect its condition at a given time. After fetching all persons that have been employees with salary > 2000 on June 3rd 1998, you can send each of these objects a setObjectTime: message in order to let its role and history attributes reflect the condition of the object at the specified time. Changes for version 1.3 Release ---------------------------------------------- - OBModeler has been redesigned to support cut & paste, modification of classes and attributes after their creation and comments for classes and attributes. Note that the obmodel file format has been changed. You can read old models with the new OBModeler but not vice versa. - Many minor bugs have been fixed. Changes for version 1.2 Release ---------------------------------------------- - Access Privileges can now correctly be set also for complex attributes (including attributes of type DATA). The responsible bug has been fixed. - Some more minor bugs have been fixed. Changes for version 1.1 Release ---------------------------------------------- - Only classbundles that have been changed since the last time the application has run are downloaded from the server again. This reduces network traffic and allows SmartBase to be used with very slow connections too. - Objects cannot only be instanciated on client computers but also on the server now - Some minor bugs have been fixed - Changes have been made to the framework libraries in order to realize the remote object mechanism. - The SmartBase demos have been updated to reflect the framework changes. INSTALLATION Simply untar SmartBase.1.5b.I.tar.gz and install all five packages on your workstation. Read the documentation in /LocalLibrary/Documentation/SmartBase for further details. If you encounter any problems or have suggestions regarding the further development of SmartBase, do not hesitate to contact us. We look forward to hear from you. For further information, contact: Smartsoft Development Email: info@smartsoft.de Special thanks to Scott Keith from OpenBase Inc. for his kind support!!!
From: ahoesch@smartsoft.de Newsgroups: comp.sys.next.programmer Subject: Test SmartBase and get a license for free! Date: 10 Jun 1998 14:02:05 GMT Organization: Offenes Netz Luebeck e.V. Message-ID: <6lm3kt$5gn@merkur.on.on-luebeck.de> Hello developers, we are very proud to announce SmartBase 1.5b for OPENSTEP Mach. SmartBase is an objective extention for relational database management and intends to essentially simplify the modeling and application development process for database dependent applications written in Objective-C. Because SmartBase is based on Apples Enterprise Objects Framework, it principly can be used with any RDBMS that is supported by EOF or comes with an EOF-adapter. Our product has undergone many tests and improvements in the last months and we believe it to have reached a state suitable for serious application development now. Anyway we want to continue our hard work to make it even better. That's why we want you to test SmartBase and give us as much feedback and ideas for the further development as possible. Your efforts will be honored with free licenses. Please visit our Web-site www.smartsoft.de to learn more about the goals of SmartBase or simply get a copy of our product. ftp.smartsoft.de/SmartBase/SmartBase.1.5b.I.b.tar.gz We look forward to hear from you! Smartsoft Development E-Mail: info@smartsoft.de
From: Joe Panico <jpanico@ml.com> Newsgroups: comp.sys.next.programmer Subject: Re: Java Yucky Date: Wed, 10 Jun 1998 10:02:44 -0400 Organization: Merrill Lynch Message-ID: <357E9204.38B48FA3@ml.com> References: <357C2877.880F61D9@alum.mit.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Eric Hermanson wrote: > 3. What's all this 'getXXX' stuff? In OpenStep, the accessor methods are > simply: variableName, setVariableName. In Java, people seem to take the "C++" > approach by using the convention: getVariableName, setVariableName. What's up > with that? When I think of 'get', I think in terms of putting something into a > buffer (getCString) - I don't tend to think in terms of accessing a variable. > Is there an advantage to the 'get' approach that I'm missing? This is a simple design pattern related to JavaBeans. getXXX/setXXX are "blessed" method prefixes that bean enabled tools know to look for as ivar accessor methods. The NeXT approach is ambiguous in this respect.
From: Stephen Peters <portnoy@ai.mit.edu> Newsgroups: comp.sys.next.programmer Subject: Re: POV Ray Sources??? Date: 10 Jun 1998 13:50:06 -0400 Organization: Massachvsetts Institvte of Technology Sender: portnoy@ANTI-MATTER Message-ID: <uzpflf8oh.fsf@ai.mit.edu> References: <lloyddean-0506980044320001@ip53.seattle7.wa.pub-ip.psi.net> <us5yavc6qb9.fsf@ai.mit.edu> <lloyddean-0806981701040001@ip210.seattle11.wa.pub-ip.psi.net> lloyddean@earthlink.net (Lloyd D. Ollmann, Jr.) writes: > What I'm looking for is the sources to both POV-Ray (the latest version) > and a viewer application. > > Being familiar with the Mac version of the sources I'm > interested in learning what changes are required in order > to work under Rhapsody Yellow box. It's a learning experience > I'm looking for, not the end product. OK, I'll try to mail you a copy of my diffs for the quick-and-dirty version that works like X11 Povray (i.e., it's a povray converter that can display to the screen while rendering). If you haven't gotten it in a couple of days, send me email; I probably just forgot. Effectively, all I did for that one was add two or three files and update the Makefile. Mind you, I've never seen the Mac version -- only X11 and Windows, so I don't know what extra features the Mac port provides. Hopefully, I'll be able to take some time and polish my NeXTSTEP/Rhapsody POVRay IDE soon. -- Stephen L. Peters portnoy@ai.mit.edu PGP fingerprint: BFA4 D0CF 8925 08AE 0CA5 CCDD 343D 6AC6 "Poodle: The other white meat." -- Sherman, Sherman's Lagoon
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: [Q] NSSplitView in a NSSplitView? Date: 11 Jun 1998 21:30:49 GMT Organization: Digital Fix Development Message-ID: <6lpia9$6rh$1@news.digifix.com> References: <6lo3f8$hn2$1@news.apple.com> <_yXf1.26$cn.1326368@news.ipass.net> In-Reply-To: <_yXf1.26$cn.1326368@news.ipass.net> On 06/11/98, Dave Coyle wrote: >Mark Bessey wrote: >> >> >> You can't. There's no inspector provided by IB to allow you to set the >> orientation. There's also a bug in SplitView that causes it to not unarchive >> it's verticality correctly sometimes. >> > >No, but it's VERY easy to write an IB palette to do this. >Hands up everyone who's done so? Should be at least a dozen of us. >;-) > >The verticality can be handled with a simple swapSides: category. >Note that this only seems to happen in IB, if the nib's saved in the >right orientation, your app will not be affected. Gee, its a darn shame that everyone is re-writing this themselves. Anyone thought about releasing it? -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: Java Yucky Date: 12 Jun 1998 03:33:22 GMT Organization: Digital Fix Development Message-ID: <6lq7i2$dmi$1@news.digifix.com> References: <357C2877.880F61D9@alum.mit.edu> <6lo5rt$vbg$1@news.apple.com> <3580706C.746C7AD7@alum.mit.edu> In-Reply-To: <3580706C.746C7AD7@alum.mit.edu> On 06/11/98, Eric Hermanson wrote: >Mark Bessey wrote: > >> I never thought I'd say this but: > >> Objective-C is a dead end, Java is the future. Once I get a Java to native >> code compiler for Rhapsody/MacOS X (something like SuperCede for Windows), >> I'll probably never use Objective-C again. > >Oh god. This, coming from an Apple engineer is truly scary! > Mark isn't an engineer. As I recall, he's in Product Marketing or quality assurance... Mark?? -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: t=*ryan@tensor.com (Ryan Dingman) Newsgroups: comp.sys.next.programmer Subject: Re: How to catch "message release sent to freed object=0xXXXXXX" Date: 10 Jun 1998 23:46:13 GMT Organization: Ball State University Message-ID: <6ln5s5$4rr@wizard.bsu.edu> References: <357E55A6.C248192F@Square.nl> Maurice le Rutte <MleRutte@Square.nl> wrote: >Is it possible to catch the "objc: FREED(id): message release sent to >freed object=0x492870" exception? On Windows-NT the user gets an awful >'0x0000DEAD' message on the screen. I'd like to catch it and present it >in a more 'humane' way. One thing that is kind of useful sometimes is to use NSZombie. In gdb type set environment NSZombieEnabled YES What this does is instead of dealloc'ing the object when it's retainCount is 0. It switches out it's isa pointer to be an NSZombie and now all further message get sent to NSZombie and it will report what type of object this is happening to. This works pretty well if you are double releaseing something like NSString. Hope this helps Ryan Dingman Tensor Information Systems, Inc. ryan@tensor.com
From: t=*ryan@tensor.com (Ryan Dingman) Newsgroups: comp.sys.next.programmer Subject: Re: How to catch "message release sent to freed object=0xXXXXXX" Date: 10 Jun 1998 23:55:17 GMT Organization: Ball State University Message-ID: <6ln6d5$4rr@wizard.bsu.edu> References: <357E55A6.C248192F@Square.nl> <6ln5s5$4rr@wizard.bsu.edu> >happening to. This works pretty well if you are double releaseing something >like NSString. Correction. This works pretty well if you AREN'T dealing with something that is common like NSString. Ryan Dingman Tensor Information Systems, Inc. ryan@tensor.com
From: Mark Bessey <"mark_bessey"@apple.com (no spam, please)> Newsgroups: comp.sys.next.programmer Subject: Re: Java Yucky Date: Thu, 11 Jun 1998 23:34:01 -0700 Organization: Apple Computer, Inc. Message-ID: <6lqi3n$d5a$1@news.apple.com> References: <357C2877.880F61D9@alum.mit.edu> <6lo5rt$vbg$1@news.apple.com> <3580706C.746C7AD7@alum.mit.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Eric Hermanson wrote: > I wish some compiler person would answer the question I've had for a long time: how > hard would it be to write either: > > 1. An Objective-C or WebScript virtual machine and associated compiler to compete > with Java? Is it possible to make Obj-C or WebScript truly garbage collected? Writing an interpreted Objective-C is pretty easy, in principle. WebScript is almost exactly that, though it lacks a type system, and support for other "C" features of Objective-C (like the standard C library). There are several variants of Objective-C available that provide some form or other of garbage collection. > 2. An Obj-C/WebScript to Java bytecode compiler. Is it possible to compile Obj-C or > WebScript code into Java bytecode and have it be garbage collected, and also have it > take into account the Obj-C categories and posing that you have going on in your code? An Objective-C to Java bytecode compiler is a little trickier. Since the Java runtime is not feature-complete for Objective-C objects, you'd either need to write an Objective-C runtime in Java (likely pretty slow, in an interpretive environment), or ditch the features that Java doesn't support (in which case you're not compiling Objective-C anymore). And there's the question of all the ANSI C features in Objective-C... > If either 1 or 2 is relatively easy, then why in the world doesn't Apple do it? > Coupled with their development tools, they could take over the market overnight. Even if Apple could do either/both of these things, I don't think they make much sense. Aside from technical considerations, I don't think Apple is really interested in trying to outdo Java's hype, and win developers (and browser and OS vendors) over to supporting this new interpreted environment. And the Rhapsody toolset isn't all that great. Sure, if the only alternative you've ever known is "vi" and "make" in a shell, ProjectBuilder/InterfaceBuilder are pretty nice. But there are a lot of Java development environments out there that are at least equivalent. -Mark
From: everhart@chronographer.com Newsgroups: comp.sys.next.programmer Subject: Re: [Q] NSSplitView in a NSSplitView? Date: Fri, 12 Jun 1998 04:28:06 GMT Organization: Calle Mayo Software Message-ID: <6lqaom$q1g$1@nnrp1.dejanews.com> References: <6lo3f8$hn2$1@news.apple.com> <_yXf1.26$cn.1326368@news.ipass.net> <6lpia9$6rh$1@news.digifix.com> In article <6lpia9$6rh$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: > Gee, its a darn shame that everyone is re-writing this > themselves. > > Anyone thought about releasing it? I haven't written an IB inspector for NSSplitView, but the new version of MiscSplitView comes with an inspector which lets you set the orientation of the split view, as well as other attributes. Its editor isn't as sophisticated as NSSplitView's, but it gets the job done. Another reason to use MiscSplitView is that it can automatically save the divider positions in the defaults database. You can also customize its appearance without subclassing it. MiscSplitView isn't in the current MiscKit release, but it can be found at http://www.chronographer.com/MiscKit/. -- Dwight Everhart "remember the way life's supposed to be Starving Programmer a frank and honest face Bella Vista, Arkansas could well destroy society" everhart@chronographer.com -- Mark Heard, "Hold Back Your Tears" -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: Mark Bessey <"mark_bessey"@apple.com (no spam, please)> Newsgroups: comp.sys.next.programmer Subject: Re: Java Yucky Date: Thu, 11 Jun 1998 23:14:34 -0700 Organization: Apple Computer, Inc. Message-ID: <6lqgvb$d16$1@news.apple.com> References: <357C2877.880F61D9@alum.mit.edu> <6lo5rt$vbg$1@news.apple.com> <3580706C.746C7AD7@alum.mit.edu> <6lq7i2$dmi$1@news.digifix.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Scott Anguish <sanguish@digifix.com> Scott Anguish wrote: > > Mark isn't an engineer. As I recall, he's in Product > Marketing or quality assurance... > > Mark?? > Oh, that's painful! For the record: My business card says "Software Engineer", and that's what I do - engineer software. Check out the credits in NetInfoManager.app for Rhapsody 1.0 when it comes out. I did formerly work in Quality Assurance, but contrary to popular belief, some QA personnel do know how to program. I worked for about 5 years as a C/C++/Visual Basic developer, writing everything from hardware diagnostics and drivers to inventory management systems, before coming to work at NeXT in the QA department. Right now, I'm working on something so cool, it'll make your jaw drop when you see it. But if I told you, then I'd have to kill you... I have a tendency to value correct behavior and reliability above performance on some arbitrary benchmark for most code. I expect that anybody else that's worked on control software probably shares the same bias. Ask me about the guy I know who lost a finger due to someone else's programming error... I've written, debugged, and helped others debug more C and Objective-C code than I like to think about in the last 10 years, and I've come to the conclusion that it's pretty difficult to write really reliable code in C-based languages. A (very) large number of the errors that have been found and fixed in OPENSTEP & Rhapsody over the last few years were exactly the kind of errors in memory allocation and pointer arithmetic that Java simply makes impossible. I'm not saying that Objective-C doesn't have advantages over Java in some situations. However, the exception handling mechanism and memory management in Java force you to write code that's in general more robust than Objective-C. You certainly *can* write reliable and resilient code in Objective-C. You're just *more likely* to do so in Java - and that's a win for you, and your customer (whoever that ultimately is). -Mark
From: henry@yusei.boeblingen.de.ibm.com (Henry Koplien) Newsgroups: comp.sys.next.programmer Subject: X3270 Date: 12 Jun 1998 06:25:49 GMT Organization: IBM Zurich Research Laboratory Distribution: world Message-ID: <6lqhld$fn4$1@grimsel.zurich.ibm.com> Where can I find a X3270 application, capable of doing all the neat stuff like graphics, etc. I was told that there is/was an application for a x3270, which is not the tn3270... Henry
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: Java Yucky Date: 11 Jun 1998 11:43:22 GMT Organization: All USENET -- http://www.Supernews.com Message-ID: <6lofsq$ldi$1@supernews.com> References: <357C2877.880F61D9@alum.mit.edu> <6lo5rt$vbg$1@news.apple.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: "mark_bessey"@apple.com (no spam, please) Mark Bessey may or may not have said: [snip] -> I never thought I'd say this but: Objective-C is a dead end, Java is the future. The future isn't here yet, (by definition.) If, in the future, Java's crippling shortcomings are addressed (not the least of which is that goddamned brain-dead C++ syntax) then I might reconsider it. In the meantime, Objective-C is a fine language with a number of good compilers, and an excellent set of libraries (Foundation and AppKit.) If I leave Obj-C, it will have to be for something *better*. A better syntax, a better object model (like prototypes instead of classes,) a better set of libraries. Java ain't it, and I don't think that Sun's moving it in the right direction. I can't tell you how much I wish Self had been given the same marketing push that Java did. -jcr
Subject: Re: Java Yucky From: "Robert A. Decker" <comrade@umich.edu> Newsgroups: comp.sys.next.programmer References: <B1A2C7DF-3C612@141.214.128.36> MIME-Version: 1.0 Message-ID: <B1A5AD1E-464319@141.214.128.36> Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Date: Thu, 11 Jun 1998 19:45:31 GMT NNTP-Posting-Date: Thu, 11 Jun 1998 15:45:31 EDT Someone made a statement about Java being the future and that they would give up on Objective-C. In a way you may be correct, but I really see Java taking the place of Visual BASIC. There will be quite a few projects out there that use it, but the scope of the projects will be smaller than Objective-C/C++ projects. You'll always be able to find Java work, but you'll be competing against hordes of crappy low-paid Java programmers... Then again, it could actually mature and be usable and take the place of C++. A time machine would be nice. rob -- <mailto: "Robert A. Decker" comrade@umich.edu> <http://hmrl.cancer.med.umich.edu/Rob/index.ssi> Programmer Analyst - Health Media Research Lab University of Michigan Comprehensive Cancer Center "Get A Life" quote #11: "Oh yeah right honey, and I'm the Goodyear blimp. Look! There are letters running across my rear!" -Chris Elliott
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: Java Yucky Date: 12 Jun 1998 09:11:00 GMT Organization: The World's Usenet -- http://www.Supernews.com Message-ID: <6lqrb4$8e3$1@supernews.com> References: <357C2877.880F61D9@alum.mit.edu> <6lo5rt$vbg$1@news.apple.com> <3580706C.746C7AD7@alum.mit.edu> <6lq7i2$dmi$1@news.digifix.com> <6lqgvb$d16$1@news.apple.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: "mark_bessey"@apple.com (no spam, please) Mark Bessey may or may not have said: [snip] -> A (very) large number of the errors that have been found and fixed in -> OPENSTEP & Rhapsody over the last few years were exactly the kind of errors in -> memory allocation and pointer arithmetic that Java simply makes impossible. Obj-C lacks smalltalk-style GC. Jumping to Java to solve that shortcoming is throwing out the baby with the bathwater, IMNSHO. If I were to switch languages to gain features, why not jump to Squeak, instead of putting up with Java's design flaws and deviant implementations? Does the phrase "Which Java?" ring a bell? -> I'm not saying that Objective-C doesn't have advantages over Java in some -> situations. However, the exception handling mechanism and memory management in -> Java force you to write code that's in general more robust than Objective-C. -> You certainly *can* write reliable and resilient code in Objective-C. You're -> just *more likely* to do so in Java - and that's a win for you, and your -> customer (whoever that ultimately is). You're also likely to take longer to write the same app in Java, due to the brain-dead syntax that inhibits the writing of legible code. If Java had a better method-call syntax, I'd look on it much more favorably. Consider it a "quality of life" issue. The worst of it though, is that once you accept all of its limitations in the *name* of security, you still don't get the security! Sure, Java isn't susceptible to buffer-overflow attacks like C, but neither is any other language that truly implements strings and arrays with proper bounds checking. I can get the same benefit in Obj-C by using NSStrings instead of char* arrays. Perl and Python aren't susceptible to buffer-overflow attacks, either. I've never enabled Java in a browser, and I *won't* until and unless the Java crowd comes up with a way to ensure trustable code. Crippling the language is a poor substitute. -jcr
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: Java Yucky Date: 12 Jun 1998 09:16:34 GMT Organization: The World's Usenet -- http://www.Supernews.com Message-ID: <6lqrli$8e3$2@supernews.com> References: <B1A2C7DF-3C612@141.214.128.36> <B1A5AD1E-464319@141.214.128.36> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: comrade@umich.edu "Robert A. Decker" may or may not have said: -> Someone made a statement about Java being the future and that they would -> give up on Objective-C. In a way you may be correct, but I really see Java -> taking the place of Visual BASIC. Exactly: Java is destined to become the language to use for Windoze-only apps. MIcrosquish has already succeeded in eviscerating Java's cross-platform potential. -> There will be quite a few projects out there that use it, but the scope of -> the projects will be smaller than Objective-C/C++ projects. You'll always -> be able to find Java work, but you'll be competing against hordes of crappy -> low-paid Java programmers... Good point. If I wanted to settle for $50/hr on projects that are so boring that I'd be tempted to go postal, I could specialize in VB. As it is, by refusing to revert when I've made a swtich to a better set of tools, I'm one of a fairly small pool of NEXTSTEP hackers, and I don't have to settle for low rates and lousy work. -> Then again, it could actually mature and be usable and take the place of -> C++. A time machine would be nice. Well, to be fair, Java is a Hell of a lot better than C++. It suffers from a bit of pandering compromise to the C++ crowd, but at least it has a true runtime-binding capability. -jcr
From: Eric Hermanson <eric@alum.mit.edu> Newsgroups: comp.sys.next.programmer Subject: Re: Java Yucky Date: Thu, 11 Jun 1998 17:03:57 -0700 Organization: Digital Universe Corporation Message-ID: <3580706C.746C7AD7@alum.mit.edu> References: <357C2877.880F61D9@alum.mit.edu> <6lo5rt$vbg$1@news.apple.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NNTP-Posting-Date: 12 Jun 1998 00:04:57 GMT Mark Bessey wrote: > I never thought I'd say this but: > Objective-C is a dead end, Java is the future. Once I get a Java to native > code compiler for Rhapsody/MacOS X (something like SuperCede for Windows), > I'll probably never use Objective-C again. Oh god. This, coming from an Apple engineer is truly scary! I wish some compiler person would answer the question I've had for a long time: how hard would it be to write either: 1. An Objective-C or WebScript virtual machine and associated compiler to compete with Java? Is it possible to make Obj-C or WebScript truly garbage collected? 2. An Obj-C/WebScript to Java bytecode compiler. Is it possible to compile Obj-C or WebScript code into Java bytecode and have it be garbage collected, and also have it take into account the Obj-C categories and posing that you have going on in your code? If either 1 or 2 is relatively easy, then why in the world doesn't Apple do it? Coupled with their development tools, they could take over the market overnight. Eric
From: mezzino@gauss.cl.uh.edu (Michael Mezzino) Newsgroups: comp.sys.next.programmer Subject: EOF2.1 EOQualifier question Date: 12 Jun 1998 14:19:45 GMT Organization: University of Houston Message-ID: <6lrde1$qed$1@Masala.CC.UH.EDU> I have now converted an old (non-trivial) NEXTSTEP3.3/DBKIT1.0 app to OPENSTEP4.2/EOF2.1. If you know anyone who is about to begin such a task, please add them to your prayer list. They will need it. In the process, I observed that EOQualifier converts the expression [EOQualifier qualifierWithQualifierFormat: @"dname IN %@", @"(SALES,RESEARCH)"] to (dname like '(SALES,RESEARCH)') and [EOQualifier qualifierWithQualifierFormat: @"dname IN %@", @"('SALES','RESEARCH')"] to (dname like '(\'SALES\',\'RESEARCH\')') Neither is correct. In particular, IN should remain IN and not be changed to LIKE and the quotes are all wrong. Am I missing something? Thanks. Mike Mezzino mezzino@gauss.cl.uh.edu
From: John Hornkvist <sorry@no.more.spams> Newsgroups: comp.sys.next.programmer Subject: Re: Java Yucky Date: Fri, 12 Jun 1998 12:18:35 GMT Organization: Chalmers Tekniska Högskola Sender: john@haddock.cd.chalmers.se (John Hprnkvist) Message-ID: <EuFu6z.Bws@haddock.cd.chalmers.se> References: <357C2877.880F61D9@alum.mit.edu> <6lo5rt$vbg$1@news.apple.com> <3580706C.746C7AD7@alum.mit.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: eric@alum.mit.edu In <3580706C.746C7AD7@alum.mit.edu> Eric Hermanson wrote: > 1. An Objective-C or WebScript virtual machine and associated compiler to compete > with Java? Is it possible to make Obj-C or WebScript truly garbage collected? IIRC garbage collection has very little to do with the language that is used. Knowing what type every object is and not having pointers should facilitate the job; then you know that a number that happens to "point" to some data is not a reference to that data, so there is no need for conservative collection. Garbage collection for Objective-C shouldn't be more difficult than garbage collection for most other languages, in fact, I think there is an experimental garbage collector for GNUStep. Garbage collection is interesting, since it is often quite difficult to assess how it affects the efficiency of a system. Some code gets faster with garbage collection than with raw malloc/free, because the garbage collector can do some tricky optimizations. I think that many garbage collection algorithms have quite appalling worst cases, though, which means that the programmer has to worry about the memory system despite using gc. Regards, John Hornkvist Address:cd.chalmers.se Name:nhoj
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer Subject: Re: Java Yucky Date: 12 Jun 1998 15:05:42 GMT Organization: Technical University of Berlin, Germany Message-ID: <6lrg46$5ut$1@news.cs.tu-berlin.de> References: <357C2877.880F61D9@alum.mit.edu> <6lo5rt$vbg$1@news.apple.com> <3580706C.746C7AD7@alum.mit.edu> <6lq7i2$dmi$1@news.digifix.com> <6lqgvb$d16$1@news.apple.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Mark Bessey <"mark_bessey"@apple.com (no spam, please)> writes: [..] >A (very) large number of the errors that have been found and fixed in >OPENSTEP & Rhapsody over the last few years were exactly the kind of errors in >memory allocation and pointer arithmetic that Java simply makes impossible. >I'm not saying that Objective-C doesn't have advantages over Java in some >situations. However, the exception handling mechanism and memory management in >Java force you to write code that's in general more robust than Objective-C. >You certainly *can* write reliable and resilient code in Objective-C. You're >just *more likely* to do so in Java - and that's a win for you, and your >customer (whoever that ultimately is). If you're extremely safety conscious, I can see an advantage of Java over Objective-C. For small projects. From all I've seen so far, Java is simply not suitable for creating the kinds of abstractions that really powerful frameworks need. With Java, you are tied to a certain level of abstraction, whereas Objective-C seems to enable the construction of new abstractions, as well as the flexibility to integrate various components, without access to the language specification. Marcel
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer Subject: Re: Java Yucky Date: 12 Jun 1998 15:48:45 GMT Organization: Technical University of Berlin, Germany Message-ID: <6lrikt$7rf$1@news.cs.tu-berlin.de> References: <357C2877.880F61D9@alum.mit.edu> <6lo5rt$vbg$1@news.apple.com> <3580706C.746C7AD7@alum.mit.edu> <6lqi3n$d5a$1@news.apple.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Mark Bessey <"mark_bessey"@apple.com (no spam, please)> writes: >An Objective-C to Java bytecode compiler is a little trickier. Since the Java >runtime is not feature-complete for Objective-C objects, you'd either need to >write an Objective-C runtime in Java (likely pretty slow, in an interpretive >environment), or ditch the features that Java doesn't support (in which case >you're not compiling Objective-C anymore). And there's the question of all the >ANSI C features in Objective-C... This is not really a problem: just use the Squeak approach! To target the JVM, you have to use a restricted form of Objective-C that doesn't, for example, use pointer arithmetic. (Squeak is a Smallltalk implementation where the VM is itself written in Smalltalk. To pull this off, the portions of Squeak that implement the VM were restricted to a subset of Smalltalk that could be translated with little pain to plain C). WebScript comes pretty close to this already, the two languages need to be clearly defined in their relation to one another. >Even if Apple could do either/both of these things, I don't think they make >much sense. Aside from technical considerations, I don't think Apple is really >interested in trying to outdo Java's hype, and win developers (and browser and >OS vendors) over to supporting this new interpreted environment. Who wants a *new* interpreted environemnt? Just give me one language that allows me to target all of YB development (Java can't do that) AND all of JVM (Objective-C currently cannot do that). With such a system, the decision wether to target JVM or native code can come at any point in the development cycle, on an individual basis, without getting stuck early-on with a language decision that significantly raise the FUD-factor and present a major obstacle to YB adoption. Marcel
From: "Dirk P. Fromhein" <Dirk.Fromhein@watershed.com> Newsgroups: comp.sys.next.programmer Subject: Re: Java Yucky (Not at all) Date: Fri, 12 Jun 1998 10:32:33 -0400 Organization: Watershed Technologies, Inc. Message-ID: <35813C01.AD7FE794@watershed.com> References: <B1A2C7DF-3C612@141.214.128.36> <B1A5AD1E-464319@141.214.128.36> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="------------519D3CB1F60DD945A5EDFEE5" This is a multi-part message in MIME format. --------------519D3CB1F60DD945A5EDFEE5 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit > Someone made a statement about Java being the future and that they would > give up on Objective-C. In a way you may be correct, but I really see Java > taking the place of Visual BASIC. > Objective-C is dead and has been for years, SmallTalk is also dead (where are the two leading Smalltalk vendors pinning their future? (yeah... Java)). Java is by no means perfect, but to say that it's not for enterprise level applications is just not true. It combines decent technical ability with killer marketing... and we all know the only thing that matters is Marketing. In the last two years we have "ported" (re-architected) some major NeXTSTEP applications to Java. It was quite easy and our clients (large Financial Houses) loved the result, we converted them from two tier to three tier and greatly improved salability, deploy ability, and maintenance. I think the largest app was about 180K lines of code (yeah I know... bad measure). In the process we created a Java object persistence layer for RDBMS that was of course influenced by our NeXSTEP work (ROF JavaBeans Edition and ROF Enterprise JavaBeans Edition). The Java IDE's are still very fragile and there are some things that need to be worked through, but Java is very much real and will become more so in the very near future. > There will be quite a few projects out there that use it, but the scope of > the projects will be smaller than Objective-C/C++ projects. You'll always > be able to find Java work, but you'll be competing against hordes of crappy > low-paid Java programmers... We only do Java, and we charge much more than when we did NeXTSTEP stuff. There is a world of difference between people who can spin heads in an applet and those that can architect and deliver enterprise level applications based on say CORBA or EJB. > Then again, it could actually mature and be usable and take the place of > C++. A time machine would be nice. Given how Java has matured in the last year I think that is a given. One year from today there will be little to no speed differences between Java and C++. Where is Objective-C going to be in one year under Apple's guidance? Good luck, Dirk Fromhein http://www.watershed.com --------------519D3CB1F60DD945A5EDFEE5 Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Dirk P. Fromhein Content-Disposition: attachment; filename="vcard.vcf" begin: vcard fn: Dirk P. Fromhein n: Fromhein;Dirk P. org: Watershed Technologies, Inc. adr: ;;;Framingham;MA;01701;USA email;internet: Dirk.Fromhein@watershed.com title: CTO/President note: http://www.watershed.com info@watershed.com x-mozilla-cpt: ;0 x-mozilla-html: FALSE version: 2.1 end: vcard --------------519D3CB1F60DD945A5EDFEE5--
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <nXag1.25139$xE3.8083324@axe.netdoor.com> Control: cancel <nXag1.25139$xE3.8083324@axe.netdoor.com> Date: 12 Jun 1998 14:29:40 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.nXag1.25139$xE3.8083324@axe.netdoor.com> Sender: vcvpdonb@aol.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <mXag1.25138$xE3.8083324@axe.netdoor.com> Control: cancel <mXag1.25138$xE3.8083324@axe.netdoor.com> Date: 12 Jun 1998 14:29:39 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.mXag1.25138$xE3.8083324@axe.netdoor.com> Sender: gwxiwygk@aol.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: Java Yucky Date: 12 Jun 1998 21:49:54 GMT Organization: Digital Fix Development Message-ID: <6ls7q2$5k5$1@news.digifix.com> References: <357C2877.880F61D9@alum.mit.edu> <6lo5rt$vbg$1@news.apple.com> <3580706C.746C7AD7@alum.mit.edu> <6lq7i2$dmi$1@news.digifix.com> <6lqgvb$d16$1@news.apple.com> In-Reply-To: <6lqgvb$d16$1@news.apple.com> On 06/11/98, Mark Bessey wrote: >Scott Anguish wrote: >> >> Mark isn't an engineer. As I recall, he's in Product >> Marketing or quality assurance... >> >> Mark?? >> > >Oh, that's painful! For the record: >My business card says "Software Engineer", and that's what I do - engineer >software. Check out the credits in NetInfoManager.app for Rhapsody 1.0 when it >comes out. I have already done this in email, but I wanted to do this here as well.. I've gotten my Mark's (Mark Dadgar is in product marketing... Mark Bessey WAS in QA but is now an engineer) mixed up, and I'm sorry to Mark Bessey for the screw-up. To avoid future mistakes like this Mark BESSEY was kind enough to forward me the Apple Mark Compatibility Chart version 1.0. -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6lve56$q04$5904@news0-alterdial.uu.net> Control: cancel <6lve56$q04$5904@news0-alterdial.uu.net> Date: 14 Jun 1998 02:56:48 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6lve56$q04$5904@news0-alterdial.uu.net> Sender: <kevin@vc1.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <3379897192022@digifix.com> Date: 14 Jun 1998 03:49:15 GMT Organization: Digital Fix Development Message-ID: <28217897796825@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1993. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <130698182206@sexy.com> Control: cancel <130698182206@sexy.com> Date: 14 Jun 1998 12:01:51 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.130698182206@sexy.com> Sender: LickIt@sexy.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <140698085517@sex.com> Control: cancel <140698085517@sex.com> Date: 14 Jun 1998 13:21:14 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.140698085517@sex.com> Sender: DeamOn@sex.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <140698092835@sexy.com> Control: cancel <140698092835@sexy.com> Date: 14 Jun 1998 13:28:40 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.140698092835@sexy.com> Sender: LickIt@sexy.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: quinlan@intergate.bc.ca Newsgroups: comp.sys.next.programmer Subject: OpenStep for Windows Date: Mon, 15 Jun 1998 08:51:08 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6m2n9s$o1j$1@nnrp1.dejanews.com> It's been a while since I've done OpenStep programming and I was wondering how little I would have to pay to get all the OpenStep development tools for Window 95. I would be using it for non-commercial reasons. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: szallies@energotec.de (Constantin Szallies) Newsgroups: comp.sys.next.programmer Subject: OS 4.2/Mach PDO bycopy objects leak? Date: 15 Jun 1998 09:38:10 GMT Organization: Customer of Technet Message-ID: <6m2q22$m87$1@oxygen.technet.net> I have the following method in one of our DO server protocols: - (bycopy NSMutableArray *)remoteObjectListForKey:(NSString *)aKey inTable:(NSString *)table keynr:(int)keynr dbErrNr:(out int *)err; When I invoke this method form a client process, an array and its content is copied into the client's process memory. The problem is that the returned array is never released. Right after the invocation, the returned array has a retain count of one. [NSAutoreleasePool showPools] reveals that the array is NOT autoreleased!!! I just can't believe that NeXT/Apple introduced such a bug into the PDO. Any idea? Greetings -- # Constantin Szallies, Energotec GmbH # szallies@energotec.de # http://www.energotec.de/~szallies/ # 49211-9144018
From: davendontlikespam@ldr.com (Dave Neuer) Newsgroups: comp.sys.next.programmer Subject: Help, I need headers! Message-ID: <davendontlikespam-1506981319240001@208.8.190.31> Organization: Litho Development & Research Date: Mon, 15 Jun 1998 20:19:45 GMT NNTP-Posting-Date: Mon, 15 Jun 1998 13:19:45 PDT I'm trying to do a PPP port for Rhapsody DR2, and it appears to be missing some necessary BSD headers. Specifically, I need: <net/ppp_defs.h> <net/vjcompress.h> <net/if_ppp.h> <net/ppp-comp.h> If anyone with a machine running NeXTSTEP or OPENSTEP has these files and can send them to me as attachments, I'd much appreciate it! Thanks, Dave Neuer
From: zegelin@actonline.com.au (Peter) Newsgroups: comp.sys.next.programmer Subject: Cost of CR1? Date: 15 Jun 1998 23:03:00 GMT Organization: Totally Disorganized Message-ID: <zegelin-1606980905060001@dialup120.canb.ispsys.net> Cache-Post-Path: genie.cyberone.com.au!unknown@dialup120.canb.ispsys.net Hi! Does anyone know the likely cost of CR1 when it finally released? Also, is it going to include development tools like Interface Builder or will they be extra? What about the compiler? Will it be possible to do small scale YB development without joining a developer program? Does anyone have an idea what the minimum cost to write programs with CR1 and YB (excluding CPU of course) on a PPC will be? Also, where does Metrowerks fit into all this?? Any help with these questions would be appreciated! thanks! Peter
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: compiler bug? Date: 16 Jun 1998 17:45:23 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6odbpo.458.rog@talisker.ohm.york.ac.uk> i've come across a problem with the following code on an intel box running openstep 4.2 - it coredumps with a bus error after printing "got here". it doesn't do it on black hardware, or under openstep for NT, neither does it do it if optimisation is turned on, but i'm a bit concerned that this problem might come back to bite me. (the problem also exhibits itself with a different gcc compiler that we've recompiled seperately, on the same hardware) has anyone out there encountered such a problem before? could someone please try it and tell me you get the same problem?! thanks, rog. //////////////////// snip here /////////////////// struct x { int x,y,z,w; // problem doesn't occur with < 3 members in the struct }; struct x fn(void) { struct x ret; // problem goes away if i remove either of the next two lines... double scale = 1; printf("got here\n"); return ret; } int main(void) { struct x rr; rr = fn(); printf("got there\n"); }
From: gamedev44@aol.com (Gamedev44) Newsgroups: comp.sys.next.programmer Subject: HEY ALL YOU INSPIRING GAME DEVELOPERS!!! Message-ID: <1998061700075100.UAA29704@ladder01.news.aol.com> Date: 17 Jun 1998 00:07:51 GMT Organization: AOL http://www.aol.com Check out this site if you have the talent in programming, 2D,3D art, music, etc. http://members.aol.com/gamedev44
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <1998061700075100.UAA29704@ladder01.news.aol.com> Control: cancel <1998061700075100.UAA29704@ladder01.news.aol.com> Date: 17 Jun 1998 00:17:16 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.1998061700075100.UAA29704@ladder01.news.aol.com> Sender: gamedev44@aol.com (Gamedev44) Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: henry@yusei.boeblingen.de.ibm.com (Henry Koplien) Newsgroups: comp.sys.next.programmer Subject: Re: compiler bug? Date: 17 Jun 1998 15:28:51 GMT Organization: IBM Zurich Research Laboratory Message-ID: <6m8nbj$1472$1@grimsel.zurich.ibm.com> References: <slrn6odbpo.458.rog@talisker.ohm.york.ac.uk> In article <slrn6odbpo.458.rog@talisker.ohm.york.ac.uk> rog@ohm.york.ac.uk (Roger Peppe) writes: > i've come across a problem with the following code on an intel box > running openstep 4.2 - it coredumps with a bus error after printing > "got here". it doesn't do it on black hardware, or under openstep for > NT, neither does it do it if optimisation is turned on, but i'm a bit > concerned that this problem might come back to bite me. (the problem > also exhibits itself with a different gcc compiler that we've > recompiled seperately, on the same hardware) > > has anyone out there encountered such a problem before? could > someone please try it and tell me you get the same problem?! > > thanks, > rog. > > //////////////////// snip here /////////////////// > > struct x { > int x,y,z,w; // problem doesn't occur with < 3 members in the struct > }; > > struct x fn(void) > { > struct x ret; > // problem goes away if i remove either of the next two lines... > double scale = 1; > printf("got here\n"); > return ret; > } > > int main(void) > { > struct x rr; > rr = fn(); > printf("got there\n"); > } I checked Your source and got no bug with a gcc 2.8.0 which I use under NS3.3 on Intel. It worked either with -O2 or without any optimization. Henry
From: lkwilson@teleport.com (larry wilson) Newsgroups: comp.sys.next.programmer Subject: IOStreams available on OpenStep 4.2? Message-ID: <lkwilson-ya023180001706982053170001@news.teleport.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-transfer-encoding: 8bit Date: Thu, 18 Jun 1998 03:51:10 GMT NNTP-Posting-Date: Wed, 17 Jun 1998 20:51:10 PDT I've got OpenStep 4.2 (the version that was distributed last year at WWDC) installed on an intel-based box. I've been trying to find the headers for iostream so I can do the following... cout << "hello world" << endl; So far, I've come up empty handed. Does 4.2 include the headers and libraries for iostreams? Thanks. -Larry -- lkwilson@teleport.com vancouver, wa
From: FreeXX@nomail.com Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.oric Subject: !!!FREE!!! (Only adults please) Date: 18 Jun 1998 07:13:04 GMT Organization: University of Rostock Message-ID: <6maem0$6bq$181@taiwan.informatik.unirostock.de> ------------------------------------------------------------------------------------------------------------- !!!! If you don't want to see any adult sites, exit this message now!!!!!! ------------------------------------------------------------------------------------------------------------- Look here for FREE SEX!!!! http://www.crosshairs.com/users/xpix/ This message was posted using Version: 3.5 of E-Mail Magnet's TV Broadcaster Software, World's Leading TV Software. Get our FREE news posting software, TV Broadcaster/News Blaster(tm), Call 1-561-470-7696
From: Car Subject: Drive any new car for $100 p/month. Message-ID: <A5MFkyom9GA.120@cidintnews.infosel.com.mx> Newsgroups: comp.sys.next.programmer Date: Thu, 18 Jun 1998 02:42:27 -0500 Drive any new car for $100 p/month. Any Make Any Model Any Price No Down Payment! No Bank Fees! No Security Deposits! Home Based Biz. Opportunity-NOT MLM! For more information, call 716-389-9669 and a live operator will take down your Name, Address, Phone, Fax (optional), and email (optional) so that we can send you an information package. It is important that you give this information, since we will not send any info without it.
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.oric Subject: cmsg cancel <6maem0$6bq$181@taiwan.informatik.unirostock.de> Control: cancel <6maem0$6bq$181@taiwan.informatik.unirostock.de> Date: 18 Jun 1998 07:19:53 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6maem0$6bq$181@taiwan.informatik.unirostock.de> Sender: FreeXX@nomail.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <A5MFkyom9GA.120@cidintnews.infosel.com.mx> Control: cancel <A5MFkyom9GA.120@cidintnews.infosel.com.mx> Date: 18 Jun 1998 07:57:29 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.A5MFkyom9GA.120@cidintnews.infosel.com.mx> Sender: Car Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Car Subject: Drive any new car for $100 p/month. Message-ID: <R4XfZ#pm9GA.120@cidintnews.infosel.com.mx> Newsgroups: comp.sys.next.programmer Date: Thu, 18 Jun 1998 04:58:10 -0500 Drive any new car for $100 p/month. Any Make Any Model Any Price No Down Payment! No Bank Fees! No Security Deposits! Home Based Biz. Opportunity-NOT MLM! For more information, call 716-389-9669 and a live operator will take down your Name, Address, Phone, Fax (optional), and email (optional) so that we can send you an information package. It is important that you give this information, since we will not send any info without it.
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <R4XfZ#pm9GA.120@cidintnews.infosel.com.mx> Control: cancel <R4XfZ#pm9GA.120@cidintnews.infosel.com.mx> Date: 18 Jun 1998 09:10:19 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.R4XfZ#pm9GA.120@cidintnews.infosel.com.mx> Sender: Car Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer,comp.sys.next.sysadmin Subject: Re: Connecting Palm III and NEXTSTEP. Has it been done? Date: 18 Jun 1998 10:33:31 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6ohr7u.5db.rog@talisker.ohm.york.ac.uk> References: <EupsrL.CIt@RnA.NL> On Wed, 17 Jun 1998 21:23:45 GMT, Gerben_Wierda@RnA.nl <Gerben_Wierda@RnA.nl> wrote: > Simple question, does anyone know if there is software to synchronize a Palm > III PDA (or Palm Pilot) under NEXTSTEP? Especially, does anyone know someone > who has this actually working? i've got some of the palm pilot link applications working under nextstep - sadly there is no 3rd party HotSync application, so they aren't nearly as useful as the PC applications provided with the Pilot. however, you can upload and download stuff without much hassle. i'm still trying to get gcc and the resource compiler working (but i haven't put *that* much effort into it) the tools port across quite easily, as i remember. hope this helps, cheers, rog.
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: 'template' keyword in Objective-C? Date: Thu, 18 Jun 1998 13:54:54 +0200 Organization: Square B.V. Message-ID: <3589000E.22868181@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit When I name a variable 'template' the ProjectBuilder highlights it as though is a keyword. It doesn't matter when I compile, but I never heard or saw it as a keyword. Is template a reserved keyword for future use? Maurice le Rutte. -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto:mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+ Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" and "to trick into believing or accepting as genuine something false and often preposterous"
From: "Charles W. Swiger" <chuck@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: 'template' keyword in Objective-C? Date: 18 Jun 1998 14:42:05 GMT Organization: Spacelab.net Internet Access Message-ID: <6mb8vt$plp$2@news.spacelab.net> References: <3589000E.22868181@Square.nl> Maurice le Rutte <MleRutte@Square.nl> wrote: >When I name a variable 'template' the ProjectBuilder highlights it as >though is a keyword. It doesn't matter when I compile, but I never heard >or saw it as a keyword. > >Is template a reserved keyword for future use? "template" is not a keyword in C or Obj-C. However, there's this other language called C++ which it might have some connection with, but you're better off not knowing about it.... -Chuck Charles Swiger | chuck@codefab.com | standard disclaimer ---------------+-------------------+-------------------- "Microsoft: we make the easy almost impossible."
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6m7pkl$s39@everest.vol.it> Control: cancel <6m7pkl$s39@everest.vol.it> Date: 18 Jun 1998 14:18:41 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6m7pkl$s39@everest.vol.it> Sender: <ironbrand@iname.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: 'eXtreme programming'? Newsgroups: comp.sys.next.programmer Message-ID: <35896ca9.0@news.depaul.edu> Date: 18 Jun 98 19:38:17 GMT Has anyone used Kent Beck's SmallTalk-oriented 'eXtreme programming' method? Have any comments? They seem to have used it successfully on a project at Chrysler. My main concern was the non-use of documentation. It seems to me that it would be difficult to bring a new person on board, or transfer a system to a different development group. -- "... and subpoenas for all." - Ken Starr
From: nospamnospamnospamnospamnospamnospamnospamnospamnospamnospam@luomat.peak.org (TjL) Newsgroups: comp.sys.next.software,comp.sys.next.programmer,comp.sys.next.sysadmin Subject: Re: VIM Followup-To: comp.sys.next.software MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit References: <6mbpqq$ndt$1@pegasus.csx.cam.ac.uk> Message-ID: <Yaei1.34635$BE5.9007300@news.rdc1.nj.home.com> Date: Thu, 18 Jun 1998 19:49:12 GMT NNTP-Posting-Date: Thu, 18 Jun 1998 12:49:12 PDT Organization: @Home Network NOTE: FOLLOWUPS TO COMP.SYS.NEXT.SOFTWARE In <6mbpqq$ndt$1@pegasus.csx.cam.ac.uk> Thomas Harte wrote: > Does anyone know if there is a version of Vim for OPENSTEP? > (cf. http://www.vim.org) ftp://ftp.peanuts.org/pub/next/unix/editor/ -- This email address does work. If you tried it previously and it bounced, my apologies.
From: jsd@gnyg.DOTcom Newsgroups: comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.programmer Subject: Rhapsody & Color Printers? Followup-To: comp.sys.next.hardware Date: 18 Jun 1998 20:05:27 GMT Organization: Valley Business Equipment Message-ID: <6mbru7$ckq@news.vbe.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NOTE: FOLLOWUPS TO COMP.SYS.NEXT.HARDWARE Are there any exceptional quality color ink jet printers such as the Epson Color Stylus 3000 which work with Rhapsody? Direct E-mail responses are appreciated but please remove the Spam "DOT" Guard first. Many Thanks, J.D. - jsd@gnyg.DOTcom
From: theisen@akaMail.com (Dirk Theisen) Newsgroups: comp.sys.next.programmer Subject: mutable-types Date: Thu, 18 Jun 1998 18:53:22 +0200 Organization: University of Bonn, Germany Message-ID: <1dau126.17esbol1tbis74N@rhrz-isdn3-p37.rhrz.uni-bonn.de> Hello! I'm currently researching on encapsulation issues in OO programming. I saw the (NSString, NSMutableString), (NSArray, NSMutableArray), etc. classes in OpenStep. Now I' looking for a reference where these classes (or their predecessors) appeared first. A pointer to an old class reference or something like that would be nice... Regards, Dirk -- No RISC - No fun
From: Dunlop <dstanl3@po-box.mcgill.ca> Newsgroups: comp.sys.next.programmer Subject: Need Training Advice Date: Wed, 17 Jun 1998 20:15:32 -0400 Organization: McGill University Computing Centre Message-ID: <35885C24.850BB0FE@po-box.mcgill.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit he current university certificate that I've just begun only briefly touches C++ and VB 5 after 10 courses at night for 2 years and $2700. It covers software design and file structures and operating systems etc, but the program coordinator says, "it's not our responsibilty to teach languages. You do that on your own. We teach you programming concepts." I want to learn programming concepts with current languages and not just Turbo Pascal (that's what they use for the 2 intro courses). Is the Microsoft certificate good for getting a programming employment? How are Webmaster certificates? Should I find another program?
From: nospamnospamnospamnospamnospamnospamnospamnospamnospamnospam@luomat.peak.org (TjL) Newsgroups: comp.sys.next.software,comp.sys.next.programmer Subject: anyone have tcpdump more recent than ver 3.0.4 or pcap.h ? Followup-To: comp.sys.next.software MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <7Gni1.34661$BE5.9274224@news.rdc1.nj.home.com> Date: Fri, 19 Jun 1998 06:36:51 GMT NNTP-Posting-Date: Thu, 18 Jun 1998 23:36:51 PDT Organization: @Home Network NOTE: FOLLOWUPS TO COMP.SYS.NEXT.SOFTWARE I downloaded the source for ver 3.4a6 from ftp://ee.lbl.gov/ "./configure" works, but "make" fails after one line: $ make cc -O -DHAVE_FCNTL_H=1 -DHAVE_MEMORY_H=1 -DTIME_WITH_SYS_TIME=1 -DHAVE_VFPRINTF=1 -DHAVE_STRCASECMP=1 -DHAVE_ETHER_NTOA=1 -DHAVE_SETLINEBUF=1 -DRETSIGTYPE=void -DRETSIGVAL= -Dint32_t=int -Du_int32_t=u_int -DLBL_ALIGN=1 -DHAVE_FDDI -I. -I/usr/local/include -c ./tcpdump.c For architecture m68k: ./tcpdump.c:43: header file 'pcap.h' not found ./tcpdump.c:88: undefined type, found `pcap_handler' ./tcpdump.c:106: undefined type, found `pcap_handler' ./tcpdump.c:119: undefined type, found `pcap_t' ./tcpdump.c:129: undefined type, found `bpf_u_int32' ./tcpdump.c:131: undefined type, found `pcap_handler' ./tcpdump.c:338: undefined type, found `pcap_dumper_t' *** Exit 1 Stop. I was going to try and compile for NeXT and Intel at least (more if possible). The CHANGELOG says that version 3.4 came out Wed Sep 24 19:51:33 PDT 1997 so I'm thinking/hoping that someone somewhere might have had time to compile this puppy... TjL
From: colombet@arles.timone.univ-mrs.fr (Bruno Colombet) Newsgroups: comp.sys.next.programmer Subject: prof or lprod for NS3.3 ? Date: 19 Jun 1998 09:25:57 GMT Organization: Universites d' Aix en Provence Message-ID: <6mdar5$6mh$2@news.univ-aix.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Hi all, I'm looking for a good objective C profiler. One that could be used in a decent way. I've tried gdb and start profile option but it is horrible for me. I did not understand how it works. Thanks -- Bruno Colombet - Université de la Méditerranée Laboratoire de Neuropsychologie et Neurophysiologie +33 (0)4 91 32 46 13 (direct) +33 (0)4 91 32 43 69 (secr. + répondeur) colombet@arles.timone.univ-mrs.fr
From: maul@anke.imsd.uni-mainz.DE (Mathias Maul) Newsgroups: comp.sys.next.programmer Subject: NXWriteRootObjectToBuffer problems Date: 19 Jun 1998 10:43:09 GMT Organization: Johannes Gutenberg-Universitaet Mainz, Germany Message-ID: <6mdfbt$h2g$1@bambi.zdv.Uni-Mainz.DE> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi there, please consider the following problem: When I say char *buf; int len; buf = NXWriteRootObject(someObj, &len); to store and someObj = NXReadObjectFromBuffer(buf, len); to retrieve (a new instance of) someObj, a "wrong super class" exception is raised. However, writing the object to and re-reading it from a file-buffered stream (with NXWriteRootObject) works fine. [I©m running NS3.3 on black hardware.] Any hints will be greatly appreciated ... foo!, Matt.
From: Christian Neuss <neuss.@informatik.th-darmstadt.de.nos-pam> Newsgroups: comp.sys.next.programmer Subject: Documenting a Framework? Date: 19 Jun 1998 16:00:07 GMT Organization: Technische Universitaet Darmstadt Message-ID: <6me1u7$bra$2@sun27.hrz.tu-darmstadt.de> Hi everynody, are there any pointers on how to properly document a Framework? The examples from NeXT aren't exactly that hot, and I wondered whether there are any 'Patterns' that I could adopt. Thanks in advance, Chris -- // Christian Neuss "static typing? how quaint.." // http://www.nexttoyou.de/~neuss/ // fax: (+49) 6151 16 5472
Newsgroups: comp.sys.next.programmer Subject: Re: OS 4.2/Mach PDO bycopy objects leak? Message-ID: <prrEz0O$D+3n@cc.usu.edu> From: root@127.0.0.1 Date: 15 Jun 98 14:54:27 MDT References: <6m2q22$m87$1@oxygen.technet.net> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit In <6m2q22$m87$1@oxygen.technet.net> Constantin Szallies wrote: > I have the following method in one of our DO server protocols: > > - (bycopy NSMutableArray *)remoteObjectListForKey:(NSString *)aKey > inTable:(NSString *)table keynr:(int)keynr dbErrNr:(out int *)err; > > When I invoke this method form a client process, an array and its content is > copied into the client's process memory. > > The problem is that the returned array is never released. > Right after the invocation, the returned array has a retain count of one. > [NSAutoreleasePool showPools] reveals that the array is NOT autoreleased!!! > > I just can't believe that NeXT/Apple introduced such a bug into the PDO. > Any idea? Bug? In the summary section of the "Object Ownership" chapter of Programming Topics, it says this: If you allocate, copy, or retain an object, you are responsible for releasing the newly created object with release or autorelease. So if an object is returned "bycopy", it is a -copy- and you are responsible for its release.
From: Drive Subject: Drive any new car for $100 p/month. Message-ID: <yY9tSF1m9GA.159@cidintnews.infosel.com.mx> Newsgroups: comp.sys.next.programmer Date: Fri, 19 Jun 1998 02:10:21 -0500 Drive any new car for $100 p/month. Any Make Any Model Any Price No Down Payment! No Bank Fees! No Security Deposits! Home Based Biz. Opportunity-NOT MLM! For more information, call 716-389-9669 and a live operator will take down your Name, Address, Phone, Fax (optional), and email (optional) so that we can send you an information package. It is important that you give this information, since we will not send any info without it.
From: holger@_REMOVE_THIS_.wizards.de (Holger Hoffstaette) Newsgroups: comp.sys.next.programmer Subject: Re: Documenting a Framework? Date: 20 Jun 1998 16:10:09 GMT Organization: the unstoppable code machine Message-ID: <6mgmt1$3pq@ragnarok.en.uunet.de> References: <6me1u7$bra$2@sun27.hrz.tu-darmstadt.de> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Christian Neuss wrote: > are there any pointers on how to properly document a Framework? > The examples from NeXT aren't exactly that hot, and I wondered > whether there are any 'Patterns' that I could adopt. You might want to start by asking your self what kind of framework you've built :-) and what kind of audience you are trying to reach. Writing isn't easy. Writing good documentation is hard. Writing *good* framework documentation is *really* hard. The following might be helpful: Pattern Language for Framework Construction Shai Ben-Yehuda http://st-www.cs.uiuc.edu/users/hanmer/PLoP-97/Proceedings/shai.pdf Ralph Johnson: documenting frameworks using patterns ftp://st.cs.uiuc.edu/pub/patterns/papers/documenting-frameworks.ps.gz (bad PostScript, use Tailor to fix) Holger
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <28217897796825@digifix.com> Date: 21 Jun 1998 03:49:16 GMT Organization: Digital Fix Development Message-ID: <12114898401620@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1993. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <JJMYRjan9GA.113@cidintnews.infosel.com.mx> Control: cancel <JJMYRjan9GA.113@cidintnews.infosel.com.mx> Date: 22 Jun 1998 06:44:48 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.JJMYRjan9GA.113@cidintnews.infosel.com.mx> Sender: Car Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: "Michelle L. Buck" <buck.erik@mcleod.net> Newsgroups: comp.sys.next.programmer Subject: Carbon and Yellow can NOT be mixed Date: Mon, 22 Jun 1998 01:43:22 -0500 Organization: McleodUSA - http://www.mcleodusa.net Message-ID: <6mkuh8$lk62@odie.mcleod.net> Several people in comp.sys.next.advocacy have claimed that one way developers will adopt Yellow Box APIs is by mixing them with Carbon gradually. This is nonsense! The two APIs can not be mixed in any rational way. Which system will handle events ? Which system will handle drawing ? If you know anything about either set of APIs, it should be obvious what a disaster this would be. An Apple employee at one of the MacOS-X WWDC sessions called this idea a "path to destruction".
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: Waiting for Window & Pasteboard Server before running other NSApp on NT Date: Mon, 22 Jun 1998 09:38:16 +0200 Organization: Square B.V. Message-ID: <358E09E8.1E1F9466@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I want an application to run when the user logs in on a Windows-NT machine. If I put it in the Startup folder sometimes it gets runned before the Window and Pasteboard server are up and running. The result is an ugly 0x0000DEAD (unknown exception) message and a screaming customer. Is there a way to check if the required servers are available before I run a progam. I tried it with a simple tool that creates an application and catches the exception, but that does not seem to work. Maurice le Rutte. -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto:mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+ Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" and "to trick into believing or accepting as genuine something false and often preposterous"
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: Carbon and Yellow can NOT be mixed Date: 22 Jun 1998 09:18:29 GMT Organization: The World's Usenet -- http://www.Supernews.com Message-ID: <6ml7h5$2do$1@supernews.com> References: <6mkuh8$lk62@odie.mcleod.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: buck.erik@mcleod.net "Michelle L. Buck" may or may not have said: -> Several people in comp.sys.next.advocacy have claimed that one way -> developers will adopt Yellow Box APIs is by mixing them with Carbon -> gradually. -> -> This is nonsense! -> -> The two APIs can not be mixed in any rational way. I don't see why not. -> Which system will handle events ? Well, if it were up to me, and the task was to port a Mac OS app to MacOS X, I think my order of battle would be: 1) Break it loose from the toolbox calls that Carbon can't handle. 2) Reimplement the GUI, using OpenStep and the Interface Builder (this is the easy part.) 3) Reimplement the file I/O. 4) Finally, reimplement the internal data representation, using the Foundation Kit collection classes and new Obj-C classes. Or, to put it another way, re-write the View and Controller, and then fix the Model when you can. Speaking as one who has, (among other things) ported a few megs of ancient FORTRAN code the the Mac, wrapping it up in a GUI along the way, and who has also made objects under NEXTSTEP that were simultaneously members of Objective-C and C++ classes, I'll say it: code in libraries can be mixed and matched, even if the developers of the libraries never had playing nicely with others in mind. BTW, my FORTRAN porting hack (QuickTOP) was an ancient NASA hack for finding optimal trajectories for ion-driven spacecraft. It's I/O was all through hard-coded device numbers which corresponded to specific plotters, hollerith card readers, and printers at NASA Ames. When I was done, the user saw a GUI with a bunch of forms to fill out to specify the mission parameters, but the QuickTOP code thought it was still running in batch mode on a VAX under VMS. The point being: You can fake code out if it sees the I/O it's expecting. -> Which system will handle drawing ? The new Quickdraw, of course. -> If you know anything about either set of APIs, it should be obvious what a -> disaster this would be. Sure, it's a lousy way to go about it. One would have far cleaner code if the parameters of the task are: "Write an OpenStep program that does everything this existing Mac program does." That would probably be faster in most cases than porting piecemeal, but you saw how the conservatives squealed when Apple merely presented them with the best dev tools they'd ever seen. -> An Apple employee at one of the MacOS-X WWDC sessions called this idea a -> "path to destruction". Clinging to the Carbon API is the path to destruction. Boil it down, it's a tool for porting legacy code. -jcr
From: rdieter@math.unl.edu (Rex Dieter) Newsgroups: comp.sys.next.programmer Subject: Re: IOStreams available on OpenStep 4.2? Date: 22 Jun 1998 13:25:02 GMT Organization: University of Nebraska-Lincoln Message-ID: <6mllve$cib$1@unlnews.unl.edu> References: <6mginn$3ge@ragnarok.en.uunet.de> In article <6mginn$3ge@ragnarok.en.uunet.de> holger@_REMOVE_THIS_.wizards.de (Holger Hoffstaette) writes: > larry wilson wrote: > > I've got OpenStep 4.2 (the version that was distributed last year at WWDC) > > installed on an intel-based box. I've been trying to find the headers for > > iostream so I can do the following... > > > > cout << "hello world" << endl; > > > > So far, I've come up empty handed. Does 4.2 include the headers and > > libraries for iostreams? > > No. They were not distributed with 4.2 because of an apparent compiler > bug. I just took everything from a 4.1 system and - wouldn't you know > it! - everything works fine. As far as 'working fine' can be applied > to C++, of course. I've uploaded C++ headers/libraries (including dylibs!) to peak. Look for libg++.2.7.2.3 packages at http://www.peak.org/openstep/ -- Rex A. Dieter rdieter@math.unl.edu (NeXT/MIME OK) Computer System Manager http://www.math.unl.edu/~rdieter/ Mathematics and Statistics University of Nebraska-Lincoln
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer,comp.next.announce Subject: Re: Conix OpenGL for Rhapsody DR2 Date: 22 Jun 1998 10:18:02 GMT Organization: I used to be organized, then they canceled the Newton. Message-ID: <6mlb0q$2hl$4@ns3.vrx.net> References: <358AFB71.1DC0@callamer.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: beretta@callamer.com In <358AFB71.1DC0@callamer.com> Robert Beretta claimed: > Announcing Conix OpenGL for Rhapsody DR2. > > Conix OpenGL for Rhapsody is available now for alpha testing. > Interested developers please read on. I'd like to thank you for taking the time to make this product available. I think it could prove to be a valuable tool in making YB interesting to specific developers, notably game authors. > * OpenGL is integrated with the YellowBox environment with a > new objective C framework. Can you describe this a bit? I think I speak for all YB developers here in saying that good integration is a wonderful thing, but bad integration can kill any product. Does your package include specific YB objects for manipulation? Do these objects take advantage of Coders, PPL's and Data's, Copying and the like? Maury
From: "Charles W. Swiger" <chuck@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: mutable-types Date: 22 Jun 1998 16:08:02 GMT Organization: Spacelab.net Internet Access Message-ID: <6mlvh2$kav$3@news.spacelab.net> References: <1dau126.17esbol1tbis74N@rhrz-isdn3-p37.rhrz.uni-bonn.de> <6mgir2$3ge@ragnarok.en.uunet.de> holger@_REMOVE_THIS_.wizards.de (Holger Hoffstaette) wrote: >They didn't. The OpenStep specification (1994?) was the first time that >NeXT publicly made the distinction between mutable and immutable objects. >As usual, they were once gain *way* ahead of the pack.. Sure, although anyone doing LISP/CLOS/Scheme with OO extensions (or another language based off of [ewww!, evil memories] lamba calculus) dealt with the same general ideas... -Chuck Charles Swiger | chuck@codefab.com | standard disclaimer ---------------+-------------------+-------------------- "Microsoft: we make the easy almost impossible."
From: ajmas@bigfoot.com Newsgroups: comp.sys.next.programmer Subject: [Q] Icon representation and information Date: Mon, 22 Jun 1998 20:36:11 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6mmf7r$5jo$1@nnrp1.dejanews.com> I am currently in the process of comparing how icons are represented on different platforms. For this I need to know the different types of information that an icon stores on each platform. So: - What is the color range of an OpenStep icon - What is the size range of an OpenStep icon - How is the mask represented - Does the mask store more information than simply transparent or not transparent (ie opaque) - Does the icon contain include a color palette or does it make reference to a system color table? - Is there any other information the icon stores that is worth noting - Where can get hold of the specification This information is need to gauge the requirements of a cross-platform icon format. Thanks AJ -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: "Michelle L. Buck" <buck.erik@mcleod.net> Newsgroups: comp.sys.next.programmer Subject: Re: [Q] Icon representation and information Date: Mon, 22 Jun 1998 21:49:13 -0500 Organization: McleodUSA - http://www.mcleodusa.net Message-ID: <6mn569$71a2@odie.mcleod.net> References: <6mmf7r$5jo$1@nnrp1.dejanews.com> ajmas@bigfoot.com wrote in message <6mmf7r$5jo$1@nnrp1.dejanews.com>... > > - What is the color range of an OpenStep icon Anything from 2 bit grey scale to 24 bit true color. Multiple representations can be stored in one tiff file. > - What is the size range of an OpenStep icon 48 by 48 > - How is the mask represented Alpha channel in tiff images. Anything from 1 bit alpha to 8 bit alpha allowing 256 diferent levels of transpearency > - Does the mask store more information than simply > transparent or not transparent (ie opaque) YES: anything in-between > - Does the icon contain include a color palette or > does it make reference to a system color table? Icons are diplayed with DisplayPostscript just like everything else. This is a device independent (printable) imaging model that does not use/expose color palettes (LUTs) to users or developers. Icons do not use the palette features of tiff. > - Is there any other information the icon stores > that is worth noting NO, but the operating system stores a database of file type to icon relationships. Plus every directory can have two optional icons to represent open and closed varients. > - Where can get hold of the specification Look into TIFF and Display Postscript. See "Encyclopedia of Graphics File Formats" by Murray & VanRyper & O'Reilly & Associates. > >This information is need to gauge the requirements of a cross-platform >icon format. > You will need device independance at the very least.
From: Car Subject: Drive any new car for $100 p/month. Message-ID: <OoOqbtmn9GA.119@cidintnews.infosel.com.mx> Newsgroups: comp.sys.next.programmer Date: Tue, 23 Jun 1998 00:54:17 -0500 Drive any new car for $100 p/month. Any Make Any Model Any Price No Down Payment! No Bank Fees! No Security Deposits! Home Based Biz. Opportunity-NOT MLM! For more information, call 716-389-9669 and a live operator will take down your Name, Address, Phone, Fax (optional), and email (optional) so that we can send you an information package. It is important that you give this information, since we will not send any info without it.
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <OoOqbtmn9GA.119@cidintnews.infosel.com.mx> Control: cancel <OoOqbtmn9GA.119@cidintnews.infosel.com.mx> Date: 23 Jun 1998 06:33:34 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.OoOqbtmn9GA.119@cidintnews.infosel.com.mx> Sender: Car Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: Advice for frameworks builders Date: Tue, 23 Jun 1998 13:43:58 +0200 Organization: Square B.V. Message-ID: <358F94FE.2D5DCA4C@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I don't know if any of you are currently building a framework on a non Windows platform, but if you want it to run on a Windows platform please study the __declspec(dllimport) and __declspec(dllexport) definitions. NeXT uses them in their AppKit. I spent an entire day not knowing I had to declare my extern variables with these modifiers. I hope I can save your day! Maurice. -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto:mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+ Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" and "to trick into believing or accepting as genuine something false and often preposterous"
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.pen,comp.sys.powerpc.advocacy,comp.sys.powerpc.misc Subject: cmsg cancel <8c81.b59d.0@ping-pong_1> Control: cancel <8c81.b59d.0@ping-pong_1> Date: 23 Jun 1998 12:00:33 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.8c81.b59d.0@ping-pong_1> Sender: The Ping Pong Team <pingpong@pemail.net> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.pen,comp.sys.powerpc.advocacy,comp.sys.powerpc.misc Subject: cmsg cancel <8c81.be88.3c0@ping-pong_1> Control: cancel <8c81.be88.3c0@ping-pong_1> Date: 23 Jun 1998 12:37:54 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.8c81.be88.3c0@ping-pong_1> Sender: The Ping Pong Team <pingpong@pemail.net> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: alexnik@tomcat.ru Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell Subject: Welcome to our Web-Site ! www.chat.ru/~andy_new Date: 23 Jun 1998 13:06:26 GMT Organization: CD-Shop Message-ID: <6mo98i$pc2$14@news.metrocom.ru> On our WWW Page you can choose and order many popular graphics programs, add-ons and many plugins.All programs are on CD.All are _FULL_ OEM version. 1. 3DS MAX 2.5 + MANY plugins (it is new CD)Price: 500$ 2. 3DS MAX 2.0 + standart plugins (include Character Studio 1.5)Price: 350$ 3. SoftImage Extreme 3.7 (Include Mental Ray and manuals)Price: 350$ 4. LightWave 5.5 + plugins Price: 300$ 5. PhotoShop 4.0 + many filters (Include After Effects)Price: 150$ 6. Adobe Illustrator 7.0 professional $150 7. Adobe Pagemaker 6.5 $150 8. Adobe Photoshop 4.01 $150 9. Adobe Premiere 4.2 $100 10. Delphi 3.0 Client/Server suite $300 11. Borland C++ Builder Client/Serv suite $100 12. Microsoft Developer Visual Studio 97 (4CD) $500 13. QNX v.4.24 $500 And more. Visited our www page and confirm. www.chat.ru/~andy_new For more information e-mail only on alexnik@tomcat.ru Thank You.
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell Subject: cmsg cancel <6mo98i$pc2$14@news.metrocom.ru> Control: cancel <6mo98i$pc2$14@news.metrocom.ru> Date: 23 Jun 1998 14:44:31 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6mo98i$pc2$14@news.metrocom.ru> Sender: alexnik@tomcat.ru Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
Message-ID: <358FE73A.C922590F@mediaone.com> From: Ahmad Dakkak <adakkak@mediaone.com> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: test Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Tue, 23 Jun 1998 17:35:07 GMT NNTP-Posting-Date: Tue, 23 Jun 1998 10:35:07 PDT Organization: MediaOne Express, Western Region test
From: Ahmad Dakkak <adakkak@mediaone.com> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: cancel <358FE73A.C922590F@mediaone.com> References: <358FE73A.C922590F@mediaone.com> Control: cancel <358FE73A.C922590F@mediaone.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <7IRj1.433$0N.4322356@cynws01.we.mediaone.net> Date: Tue, 23 Jun 1998 17:36:03 GMT NNTP-Posting-Date: Tue, 23 Jun 1998 10:36:03 PDT Organization: MediaOne Express, Western Region This message was cancelled from within Mozilla.
From: ajmas@bigfoot.com Newsgroups: comp.sys.next.programmer Subject: Getting screen refresh rate? Date: Tue, 23 Jun 1998 19:41:19 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6mp0cv$evg$1@nnrp1.dejanews.com> I am looking at creating a simple ligh pen project. In order to be able to make it to work I need to be able to get hold of the screen refresh and timing info. Does anyone have any idea how I could do this under OpenStep? Has anyone already put together such a porject? AJ -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: "Michelle L. Buck" <buck.erik@mcleod.net> Newsgroups: comp.sys.next.programmer Subject: Re: Carbon and Yellow can NOT be mixed Date: Tue, 23 Jun 1998 15:03:18 -0500 Organization: McleodUSA - http://www.mcleodusa.net Message-ID: <6mp1p3$egc1@odie.mcleod.net> References: <6mkuh8$lk62@odie.mcleod.net> Michelle L. Buck wrote in message <6mkuh8$lk62@odie.mcleod.net>... I over reacted in saying that Carbon and YellowBox can not be mixed. I was only considering the AppKit v. ToolBox issues at the time. Carbon Applications could benefit from FoundationKit up to a point and a C or Pascal model could certainly benefit from an AppKit GUI. The incompatible union would be ToolBox GUI and AppKit GUI mixed. That is just irrational.
From: Tristan Austin <tristan.austin@dsto.defence.gov.au> Newsgroups: comp.sys.next.programmer Subject: WebObjects resources Date: Wed, 24 Jun 1998 15:20:01 +1000 Organization: Defence Science & Technology Organisation Message-ID: <35908C81.E8975649@dsto.defence.gov.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I'm looking for some high level details on WebObjects so I can convince my boss to train me in it in a few weeks. If anyone knows where I can get some information which "sells" WebObjects on its functionality and features, I'd appreciate a pointer to it. Thanks. -- Tristan Austin tristan.austin@dsto.defence.gov.au ***************************************** "One World, One Web, One Program" - Microsoft Promotional Ad "Ein Volk, Ein Reich, Ein Fuhrer" - Adolf Hitler (02) 62658828 DSTO C3 Research Centre Fernhill Park Canberra ACT 2600 Department of Defence *****************************************
From: "John F. Kennedy" <jfk@rlw.com> Newsgroups: comp.sys.mac.programmer.misc,comp.sys.next.programmer Subject: Re: Where can YellowBox for Windows be obtained. Date: Wed, 24 Jun 1998 09:11:01 -0700 Message-ID: <6mr8b5$jud$1@news0-alterdial.uu.net> At 06:55 AM 6/24/98 , you wrote: >In article <6l9j58$s08$1@news0-alterdial.uu.net> you write: >>I am a Windows programmer, and I am investigating the potential to do >>crossplatform programming with a Mac. I am evaluating Create which has been >>compiled for the YellowBox, but I can not find the necessary files. >> >>Where can I find YellowBox for Windows? > >I am in exactly the same situation! Did you get any helpful answers so >far? > >Thanks in advance, > >Marcus > No, I have not. You would think it was in Stone's and Apple's interest to allow developers to test their technology, allowing YB for windows to be distributed with programs like Create, before making an investment. Apple appears to be dragging their feet on this. Everyday that they delay is another programmer who abandons the Apple platform to develop for Windows. Windows development environments, VB5 and Visual Interdev, are enticing programmers to develop for Windows based systems. Apple has offered nothing that compares to these development environments on ease of use grounds or their inexpensive deployment costs. Nothing. Good luck finding help on the yellowbox. I feel Apple will abandon this technology in the near term, and I am not going to invest a dime in any code dependent on it with such a silent response. They had the same attitude with OpenDoc and the Newton. Silence means abandonment.
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: WebObjects resources Date: 24 Jun 1998 06:18:00 GMT Organization: Digital Fix Development Message-ID: <6mq5mo$m08$1@news.digifix.com> References: <35908C81.E8975649@dsto.defence.gov.au> In-Reply-To: <35908C81.E8975649@dsto.defence.gov.au> On 06/23/98, Tristan Austin wrote: >I'm looking for some high level details on WebObjects so I can convince >my boss to train me in it in a few weeks. > >If anyone knows where I can get some information which "sells" >WebObjects on its functionality and features, I'd appreciate a pointer >to it. > >Thanks. http://www.wosource.com http://enterprise.apple.com -- Scott Anguish <sanguish@digifix.com> NEXTSTEP/OpenStep Information <URL:http://www.stepwise.com>
From: gerriet@hazel.north.de (Gerriet M. Denkmann) Newsgroups: comp.sys.next.programmer Subject: serial line without posix Date: 24 Jun 1998 08:54:42 GMT Organization: Oldenburger Informations-Systeme, FRG Distribution: world Message-ID: <6mqesi$hcq@hazel.north.de> As everybody knows posix and Next do not mix very well. I have a small program which works in the posix version, but I would love to have a non-posix version as well. But I am stuck and need your help. Here is the program - what did I do wrong in the non-posix version? #ifdef USE_POSIX struct termios tty_thing; #else struct sgttyb tty_thing ; #endif char some_data[8] ; char a_char ; int fileDescr = open( "/dev/ttyb", O_RDWR) ; // We need the device to be raw. 8 bits even parity on 9600 baud. #ifdef USE_POSIX // this works tcgetattr(fileDescr, &tty_thing); cfmakeraw(&tty_thing); tty_thing.c_oflag &= ~CSTOPB; tty_thing.c_cflag |= PARENB; tty_thing.c_cflag &= ~PARODD; tty_thing.c_cc[VMIN] = 0; tty_thing.c_cc[VTIME] = 50; cfsetospeed(&tty_thing, B9600); cfsetispeed(&tty_thing, B9600); tcsetattr(fileDescr, TCSANOW, &tty_thing); #else // this does not work ioctl( fileDescr, TIOCGETP, &tty_thing ) ; tty_thing.sg_ispeed = B9600 ; tty_thing.sg_ospeed = B9600 ; tty_thing.sg_flags = EVENP | RAW ; ioctl( fileDescr, TIOCSETP, &tty_thing ) ; #endif write( fileDescr, some_data, 8) ; // this works for both versions read( fileDescr, &a_char, 1 ); // here the posix version gets a character back (as it should) // but the non-posix version does get nothing - maybe the device // did not get the right data and so refuses to respond Any help would be very much appreciated! Gerriet.
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: Setting user's preferred language Date: Fri, 26 Jun 1998 15:08:18 +0200 Organization: Square B.V. Message-ID: <35939D42.500D5A5@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I'd like to be able to influence the user's preferred language programatically. For instance, I use an English version of Windows NT but my native language is Dutch. OpenStep does not recognize Dutch (as far as I know) and always uses the English.lproj bundles. Is there a user default that I can change so that the system will use other language projects? Most favourable would be that it uses ISO codes, like 'en-UK','en-US' or 'nl' and 'nl-BE'. Maurice le Rutte. -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto:mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+ Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" and "to trick into believing or accepting as genuine something false and often preposterous"
From: rdieter@math.unl.edu (Rex Dieter) Newsgroups: comp.sys.next.programmer Subject: Re: Shared memory mapping Date: 26 Jun 1998 14:52:22 GMT Organization: University of Nebraska-Lincoln Message-ID: <6n0cj6$qsj$1@unlnews.unl.edu> References: <6mv4fm$104@sjx-ixn2.ix.netcom.com> In article <6mv4fm$104@sjx-ixn2.ix.netcom.com> no.spam.please@no.spam.period writes: > > What function calls does NeXT provide > to map shared memory segments backed by regular files, and/or > backed by anonymous swap regions? > > db2.4.14 uses mmap(2) and shmget(2) which are missing from NS/OS. That's exactly the reason why db v2.x won't work reliably under NS/OS, and never will, unless db v2.x internals are rewritten to not rely upon mmap and/or shmget so heavily. Rhapsody, on the other hand, should be better, since it does (should?) have a fully functional mmap. -- Rex A. Dieter rdieter@math.unl.edu (NeXT/MIME OK) Computer System Manager http://www.math.unl.edu/~rdieter/ Mathematics and Statistics University of Nebraska-Lincoln
From: "Ziya Oz" <ziyaoz@earthlink.net> Newsgroups: comp.sys.next.programmer Subject: Tell Apple to make Rhapsody compatible with PBG3s Date: Wed, 24 Jun 1998 17:49:57 -0400 Organization: BM Message-ID: <6mrs1j$2si$1@oak.prod.itd.earthlink.net> Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit Here's what Apple says about Rhapsody and PowerBook G3 compatibility: "Engineering is currently working on support for the new PowerBook G3 systems, however we don't know if that will be in Rhapsody 1.0 or not. If Rhapsody 1.0 does include support for PowerBooks, it will be limited support and will not include power management." So if you wanted to try out Rhapsody or WebObjects on your PBG3 and take it out to show them to colleagues and clients, as of now, you're out of luck. If you believe that Rhapsody CR1 should be compatible with PBG3s, won't you take a minute a send an email (to the address below) to remind Apple that it is a good idea: dts@apple.com
Newsgroups: comp.sys.next.programmer Subject: Re: serial line without posix Message-ID: <yUFVJXhLFlAm@cc.usu.edu> From: root@127.0.0.1 Date: 24 Jun 98 10:31:52 MDT References: <6mqesi$hcq@hazel.north.de> Distribution: world MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit In <6mqesi$hcq@hazel.north.de> Gerriet M. Denkmann wrote: > As everybody knows posix and Next do not mix very well. > > I have a small program which works in the posix version, but I would love > to have a non-posix version as well. But I am stuck and need your help. > > Here is the program - what did I do wrong in the non-posix version? > > #ifdef USE_POSIX > struct termios tty_thing; > #else > struct sgttyb tty_thing ; > #endif > char some_data[8] ; > char a_char ; > int fileDescr = open( "/dev/ttyb", O_RDWR) ; > > // We need the device to be raw. 8 bits even parity on 9600 baud. > #ifdef USE_POSIX // this works > tcgetattr(fileDescr, &tty_thing); > cfmakeraw(&tty_thing); > tty_thing.c_oflag &= ~CSTOPB; > tty_thing.c_cflag |= PARENB; > tty_thing.c_cflag &= ~PARODD; > tty_thing.c_cc[VMIN] = 0; > tty_thing.c_cc[VTIME] = 50; > cfsetospeed(&tty_thing, B9600); > cfsetispeed(&tty_thing, B9600); > tcsetattr(fileDescr, TCSANOW, &tty_thing); > #else // this does not work > ioctl( fileDescr, TIOCGETP, &tty_thing ) ; > tty_thing.sg_ispeed = B9600 ; > tty_thing.sg_ospeed = B9600 ; > tty_thing.sg_flags = EVENP | RAW ; > ioctl( fileDescr, TIOCSETP, &tty_thing ) ; > #endif > > write( fileDescr, some_data, 8) ; > // this works for both versions > > read( fileDescr, &a_char, 1 ); > // here the posix version gets a character back (as it should) > // but the non-posix version does get nothing - maybe the device > // did not get the right data and so refuses to respond > > Set the CBREAK flag to make the device return a character as soon as it is typed. tty_thing.sg_flags = EVENP | RAW | CBREAK;
From: "Charles W. Swiger" <chuck@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: Shared memory mapping Date: 26 Jun 1998 18:31:35 GMT Organization: Spacelab.net Internet Access Message-ID: <6n0pe7$dfm$2@news.spacelab.net> References: <6mv4fm$104@sjx-ixn2.ix.netcom.com> <6n0cj6$qsj$1@unlnews.unl.edu> rdieter@math.unl.edu (Rex Dieter) wrote: >In article <6mv4fm$104@sjx-ixn2.ix.netcom.com> >no.spam.please@no.spam.period writes: >> >> What function calls does NeXT provide >> to map shared memory segments backed by regular files, and/or >> backed by anonymous swap regions? >> >> db2.4.14 uses mmap(2) and shmget(2) which are missing from NS/OS. > >That's exactly the reason why db v2.x won't work reliably under NS/OS, and >never will, unless db v2.x internals are rewritten to not rely upon mmap >and/or shmget so heavily. With the map.c from Fabien Roy, you can get the mmap/munmap sections working pretty well, and DB2 (at least 2.4.10) does not depend on the SysV shared memory and can be configured not to use it. I've gotten DB working fairly well, although there appears to be some bug closing files cleanly after map'ing them, and anything explicitly trying to use the shared memory functionality loses: % grep -A3 -B3 FAIL ALL.OUT Unlinking log: error message OK Archive.a: Writing log records; checkpoint every 500 records FAIL:10:21:11 (00:00:00) nlogfiles: expected 4, got 8 Byteorder: DB_HASH 1000 Test001: DB_HASH (-order 1234 {}) 1000 equal key/data pairs Test001.a: put/get loop -- Test003: DB_HASH (-order 1234 {}) filename=key filecontents=data pairs Test003.a: put/get loop Test003.b: dump file FAIL:10:21:49 (00:00:00) Test003:datamismatch(./../os/os_rpath.c,./TESTDIR/d1): expected 0, got 1 Lock001: open/close test dbtest: ./TESTDIR/__db_lock.share: No such file or directory FAIL:10:21:50 (00:00:00) lock_unlink: expected 0, got -1 Unlinking log: error message OK Log001: Open/Close/Create/Unlink test dbtest: ./TESTDIR/__db_log.share: No such file or directory -- Memp003.f: Get more pages for the read-only file Memp003.g: Sync from the read-only file dbtest: mpool.db: Invalid argument FAIL:10:33:05 (00:00:36) db_close: expected 0, got db_close:invalid argument -shmem anon not supported; test skipped -shmem named not supported; test skipped Mutex001: Basic functionality -- dbtest: ./TESTDIR/__db_txn.share: No such file or directory Txn002: Basic begin, commit, abort Txn002.a: Beginning/Committing Transactions FAIL:11:02:11 (00:00:00) txn_begin: expected 1, got 0 Dead001: Deadlock detector tests Dead001.a: creating environment Dead001: 2 procs of test ring -- Test003.a: put/get loop Test003.b: dump file Test003.c: close, open, and dump file FAIL:11:39:45 (00:00:40) Test003:datamismatch(./../btree/bt_split.c,./TESTDIR/d1): expected 0, got 1 run_method: DB_RBTREE 4 4 11:39:46 (00:00:00) Test004: DB_BTREE (-flags 0x0010) 10000 delete small key; medium data pairs -- Test008.b: Delete re-add loop Test008.c: Dump contents forward Test008.d: Dump contents backward FAIL:11:41:21 (00:00:51) diff(./../test/tcl_mpool.c,./TESTDIR/d1): expected 0, got 1 run_method: DB_RBTREE 9 9 11:41:21 (00:00:00) Test009: DB_BTREE filename=key filecontents=data pairs(with close) -- Test003: DB_HASH () filename=key filecontents=data pairs Test003.a: put/get loop Test003.b: dump file FAIL:12:10:21 (00:00:27) Test003:datamismatch(./../os/os_config.c,./TESTDIR/d1): expected 0, got 1 run_method: DB_HASH 4 4 12:10:22 (00:00:00) Test004: DB_HASH () 10000 delete small key; medium data pairs -- Test018.b: dump file Test018.c: close, open, and dump file dbtest: <db: Invalid argument FAIL:13:46:50 (00:00:03) db_close: expected 0, got db_close:invalid argument run_method: DB_RECNO 19 19 13:46:50 (00:00:00) Test019: DB_RECNO () 10000 partial get test -- Test003.b: dump file Test003.c: close, open, and dump file Test003.d: close, open, and dump file in reverse direction FAIL:13:52:10 (00:00:46) Test003:datamismatch(./../mp/mp_region.c,./TESTDIR/d1): expected 0, got 1 run_method: DB_RRECNO 4 4 13:52:10 (00:00:00) Test004: DB_RECNO (-flags 0x0020) 10000 delete small key; medium data pairs -- Test018.b: dump file Test018.c: close, open, and dump file dbtest: <db: Invalid argument FAIL:14:08:42 (00:00:28) db_close: expected 0, got db_close:invalid argument run_method: DB_RRECNO 19 19 14:08:42 (00:00:00) Test019: DB_RECNO (-flags 0x0020) 10000 partial get test -- Recd005.5.b: Populating databases File recd005.1.db executed and aborted. File recd005.2.db executed and aborted. FAIL:14:17:50 (00:00:00) diff(initial,post-abort):diff(./TESTDIR/recd005.2.db.t1,./TESTDIR/recd005.2.d b.t3): expected 0, got 1 Recd001: DB_BTREE operation/transaction tests Recd001.a: creating environment Recd001.b: put abort -- Command executed and aborted. About to run recovery ... complete dbtest: <: Invalid argument FAIL:14:18:10 (00:00:00) db_close: expected 0, got db_close:invalid argument Recd003: DB_BTREE duplicate recovery tests Recd003.a: creating environment Recd003.b: add dups abort -- Command executed and aborted. About to run recovery ... complete dbtest: lt: Invalid argument FAIL:14:20:50 (00:00:00) db_close: expected 0, got db_close:invalid argument Recd005: DB_BTREE catastropic recovery Recd005.1: 1000 10 abort abort Recd005.1.a: creating environment -- Recd001.i: overwrite commit Command executed and committed. About to run recovery ... complete FAIL:14:22:47 (00:00:00) diff(pre-commit,post-commit):diff(./TESTDIR/t2,./TESTDIR/t3): expected 0, got 1 Recd002: DB_RECNO split recovery tests Recd002.a: creating environment Recd002.b: splits abort Command executed and aborted. About to run recovery ... complete dbtest: <: Invalid argument FAIL:14:22:50 (00:00:00) db_close: expected 0, got db_close:invalid argument Recd003 skipping for method RECNO Recd003 skipping for method DB_RECNO Recd005: DB_RECNO catastropic recovery -- Command executed and aborted. About to run recovery ... complete dbtest: <: Invalid argument FAIL:14:25:26 (00:00:00) db_close: expected 0, got db_close:invalid argument Recd003 skipping for method RECNO Recd003 skipping for method DB_RECNO Recd005: DB_RECNO catastropic recovery -Chuck Charles Swiger | chuck@codefab.com | standard disclaimer ---------------+-------------------+-------------------- "Microsoft: we make the easy almost impossible."
From: macghod@concentric.net (Steve Sullivan) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Confused about Apple Date: Fri, 26 Jun 1998 15:05:52 -0700 Organization: Great till Apple got rid of the newton Message-ID: <macghod-2606981505530001@sdn-ar-001casbarp301.dialsprint.net> I am quite confused about something.... As far as I can tell, Apple has now more or less explicitly stated "we are fully concentrating on the macos consumer market", and basically NOT focusing on the non consumer market at this time. I believe one can infer this from what they have said, they have left room to further focus on the non consumer market after macos x comes out, but untill then they are focusing on the consumer. What the f is up with this? They bought the best enterprise os in the world, and instead they focus wholely on the consumer??!??! Macos x is good, but why not also at the same time also focus on the enterprise market, besides focusing on macos x also have a "macos enterprise". -- So many pedestrians, so little time.
From: Pascal Bourguignon <pbourgui@afaa.asso.fr> Newsgroups: comp.sys.next.programmer Subject: Re: Setting user's preferred language Date: 26 Jun 1998 20:32:00 GMT Organization: ImagiNET Message-ID: <6n10g0$1ng$1@news.imaginet.fr> References: <35939D42.500D5A5@Square.nl> Maurice le Rutte <MleRutte@Square.nl> wrote: >I'd like to be able to influence the user's preferred language >programatically. For instance, I use an English version of Windows NT >but my native language is Dutch. OpenStep does not recognize Dutch (as >far as I know) and always uses the English.lproj bundles. Is there a >user default that I can change so that the system will use other >language projects? Most favourable would be that it uses ISO codes, like >'en-UK','en-US' or 'nl' and 'nl-BE'. % dread -l | grep 'System Language' System Language English;French;German;Spanish;Italian;Swedish I think you could try: % dwrite System Language 'Dutch;English;French;German;Spanish;Italian;Swedish' This would probably work. __Pascal Bourguignon__
From: rmcassid@uci.edu (Robert Cassidy) Newsgroups: comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Confused about Apple Date: Fri, 26 Jun 1998 16:04:47 -0700 Organization: UC Irvine, School of Engineering Message-ID: <rmcassid-2606981604470001@dante.eng.uci.edu> References: <macghod-2606981505530001@sdn-ar-001casbarp301.dialsprint.net> In article <macghod-2606981505530001@sdn-ar-001casbarp301.dialsprint.net>, macghod@concentric.net (Steve Sullivan) wrote: >What the f is up with this? They bought the best enterprise os in the >world, and instead they focus wholely on the consumer??!??! Because there is a difference between what a product is, and what consumers will buy. The MacOS is what people will buy right now. Some of us want the enterprise stuff as well, but it's not all together enough right now to be thrust out into comparisons with either NT or Solaris. So they will sell MacOS now and develop the enterprise stuff in the mean time, but not make a big deal about it. At some point, 'milking the MacOS for all it is worth' will have gone about as far as it can, the enterprise components will be much better, and the focus will start to shift. IMO. -Bob Cassidy
From: leigh@NOSPAMcs.uwa.edu.au (Leigh Smith) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Rhapsody DriverKit? Date: 26 Jun 1998 07:37:21 GMT Organization: The University of Western Australia Distribution: world Message-ID: <6mvj3h$ofk$1@enyo.uwa.edu.au> References: <7iCszC6Jo1$V@cc.usu.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: root@127.0.0.1 In <7iCszC6Jo1$V@cc.usu.edu> root@127.0.0.1 wrote: > Does anyone know if Rhapsody/Intel will be released with > a DriverKit? Since it is clear there will be no future > releases of Rhapsody/Intel post 1.0, one might assume there > will be little incentive for Apple to spend time writing drivers > for a large variety of hardware. At least with a DriverKit we > would be able to write our own. > > So what is it folks? DriverKit or no? Remember the DriverKit arose because NeXT needed a means to handle Intel hardware. The DriverKit docs for Rhapsody are the same as the DriverKit docs for 3.3 and the *.config files on Rhapsody are the same as 3.3 (except with a recompile for BSD4.4) so it is fair bet RDR2/Intel will have the development kit. My reading of the situation is that Apple are not announcing a Mac OS X port for Intel but are keeping all the development they already have on Rhapsody up to date (at relatively little cost). This allows them to commit an Intel version at the drop of a hat and it acts as a potential threat to NT - especially with new architectures expected. -- Leigh Computer Music Lab, Computer Science Dept, Smith University of Western Australia +61-8-9380-2279 leigh@cs.uwa.edu.au (NeXTMail/MIME) C++ is to C, as Lung Cancer is to Lung - John C. Randolph
From: Thomas Harte <tph1001@CL.cam.ac.uk> Newsgroups: comp.sys.next.software,comp.sys.next.programmer,comp.sys.next.sysadmin Subject: Emacs viper-mode Date: 27 Jun 1998 11:36:11 GMT Organization: Cambridge University Message-ID: <6n2lfb$7m6$2@pegasus.csx.cam.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I installed version 19.34.1 of Emacs for NEXTSTEP the other day. viper-mode was working just fine. Today when I do M-x viper-mode RET I get: Symbol's function definition is void: x-color-defined-p and viper-mode does not load. Any idea what I have done to warrant this warning? When I do emacs --no-init from the terminal shell viper-mode works fine. The problem seems to exist only when emacs is fired as the Emacs app. Any ideas what the above error is indicating? c-h f x-color-defined-p draws a blank. --- thomas harte @ computer laboratory, cambridge university; tph1001@CL.cam.ac.uk; phone: +44 1223 335089; fax: 334678; ¬http://www.CL.cam.ac.uk/users/tph1001 MIME & NeXT Mail OK.
From: andrewp@interSPAMport.net (Andrew Pontious) Newsgroups: comp.sys.next.programmer Subject: Getting Started with Rhapsody Date: Sat, 27 Jun 1998 14:27:01 -0500 Organization: Interport Communications Corp. Message-ID: <andrewp-2706981427020001@ts5port13.port.net> I've toyed with Mac programming on and off and want to get into serious Rhapsody development at home without paying a huge amount of money. First question is whether to wait for Rhapsody 1.0 or actually sign up for the $500 Applet Developer program now. That one's pretty easy--I'll probably wait. I have a PowerMac 7200/75, souped up to 64 Mb. RAM and 4.0 Gb. HD. But the 601 chip still won't run Rhapsody anyway, right? So I'm going to have to buy *some* new computer at *some* point. What question is *what* and *when*. The *what* is probably going to be a PowerMac G3, since I want to be able to run MacOS X when it comes out. If I get it now, I'll be paying mid-range prices for an officially non-upgradeable computer, when a *new* set of machines, possibly upgradeable, is set to debut in November. The bad news is that those new machines are probably going to be high-end, out of my budget, but their introduction might lower the mid-range prices so getting a computer that will only last 2-3 years as a cutting-edge machine for programming might be a good idea. (For example, ideally I'd be able to play around with and compile Netscape for Mac/Rhapsody, which yes, I know, would also involve lots of memory.) On the other hand, I could get a cheap PC to run Rhapsody, forget my dreams of MacNetscape for now, but still be able to code Mac programs on my old Mac, and then get a brand-new G3 Mac 1 1/2 years from now when Mac OS X comes out. As yet another option, I could try to install Rhapsody in some sort of dual-boot setup on my WinNT box at work. Does Rhapsody coexist well with WinNT and fit comfortably on 1-2 gigs? If anybody has any advice, let me know. Andrew Pontious email: take out SPAM from address above
From: Tristan Austin <tristan.austin@dsto.defence.gov.au> Newsgroups: comp.sys.next.programmer Subject: Re: Getting Started with Rhapsody Date: Mon, 29 Jun 1998 09:41:39 +1000 Organization: Defence Science & Technology Organisation Message-ID: <3596D4B3.57DA83E4@dsto.defence.gov.au> References: <andrewp-2706981427020001@ts5port13.port.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Andrew Pontious <andrewp@interSPAMport.net> In a previous posting from someone elseon this newsgroup, I noticed someone say that Rhapsody 1.0 would only run on NT. I asked the guy who runs softpac, which handles the Intel side of things for Apple here in Australia about it. He said: >Tristan, > >>release 1.0 of Rhapsody will only run on Intel machines >this is OpenStep 4.2 with a few Apple cosmetics, so will only run on NT > [...] > >Rgds dave t Obviously the developer releases run on the PowerPC so who knows what's going to come out the other end. Andrew Pontious wrote: > I've toyed with Mac programming on and off and want to get into serious > Rhapsody development at home without paying a huge amount of money. > > First question is whether to wait for Rhapsody 1.0 or actually sign up for > the $500 Applet Developer program now. That one's pretty easy--I'll > probably wait. > > I have a PowerMac 7200/75, souped up to 64 Mb. RAM and 4.0 Gb. HD. But the > 601 chip still won't run Rhapsody anyway, right? So I'm going to have to > buy *some* new computer at *some* point. What question is *what* and > *when*. > > The *what* is probably going to be a PowerMac G3, since I want to be able > to run MacOS X when it comes out. > > If I get it now, I'll be paying mid-range prices for an officially > non-upgradeable computer, when a *new* set of machines, possibly > upgradeable, is set to debut in November. The bad news is that those new > machines are probably going to be high-end, out of my budget, but their > introduction might lower the mid-range prices so getting a computer that > will only last 2-3 years as a cutting-edge machine for programming might > be a good idea. (For example, ideally I'd be able to play around with and > compile Netscape for Mac/Rhapsody, which yes, I know, would also involve > lots of memory.) > > On the other hand, I could get a cheap PC to run Rhapsody, forget my > dreams of MacNetscape for now, but still be able to code Mac programs on > my old Mac, and then get a brand-new G3 Mac 1 1/2 years from now when Mac > OS X comes out. > > As yet another option, I could try to install Rhapsody in some sort of > dual-boot setup on my WinNT box at work. Does Rhapsody coexist well with > WinNT and fit comfortably on 1-2 gigs? > > If anybody has any advice, let me know. > > Andrew Pontious > email: take out SPAM from address above -- Tristan Austin tristan.austin@dsto.defence.gov.au ***************************************** "One World, One Web, One Program" - Microsoft Promotional Ad "Ein Volk, Ein Reich, Ein Fuhrer" - Adolf Hitler (02) 62658828 DSTO C3 Research Centre Fernhill Park Canberra ACT 2600 Department of Defence *****************************************
From: nurban@crib.corepower.com (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: Getting Started with Rhapsody Date: 28 Jun 1998 20:55:51 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6n6omn$9ts$1@crib.corepower.com> References: <andrewp-2706981427020001@ts5port13.port.net> <3596D4B3.57DA83E4@dsto.defence.gov.au> In article <3596D4B3.57DA83E4@dsto.defence.gov.au>, Tristan Austin <tristan.austin@dsto.defence.gov.au> wrote: > In a previous posting from someone elseon this newsgroup, I noticed someone > say that Rhapsody 1.0 would only run on NT. Rhapsody doesn't run on top of NT, it is an operating system. Yellow Box for Windows provides the same APIs as Rhapsody (Yellow Box / OpenStep) for NT. That is a separate product from Rhapsody. > I asked the guy who runs softpac, which handles the Intel side of things for > Apple here in Australia about it. He said: > >>release 1.0 of Rhapsody will only run on Intel machines He is very confused. Rhapsody will very definitely run on PowerMacs (at least some of them). Perhaps Apple is not doing a good job of educating people about this stuff, but I really can't see how _anyone_ could have gotten that impression. > >this is OpenStep 4.2 with a few Apple cosmetics, so will only run on NT Quite false. OPENSTEP 4.2 didn't "run on NT" either, since it was also a separate operating system. It did run on Intel and did not run on Macs, but Rhapsody does run on Macs. > Obviously the developer releases run on the PowerPC so who knows what's going > to come out the other end. Right. If anything, it is the _Intel_ OS offering that is in jeopardy; it's amusing to think that anyone would consider it to be "only Intel".
From: uli@tallowcross.uni-frankfurt.de (Uli Zappe) Newsgroups: comp.sys.next.programmer Subject: Can't open HELP anymore in my app Date: 30 Jun 1998 03:14:20 GMT Organization: NEXTTOYOU Message-ID: <6n9l6c$8vr$1@tallowcross.uni-frankfurt.de> NNTP-Posting-Date: 30 Jun 1998 03:14:20 GMT Hi, I just recompiled an app (NS 3.3) that worked without any problems so far. While I'm not aware that I made any relevant changes, suddenly I can't open the NEXTSTEP help anymore. If I click on the Help menu, an alert panel says: MyApp must be linked with libIndexing in order to load help. and the console says: objc: class `IXBTree' not linked into application The only thing I could think of is that I linked a new version of an external library into my app. Any idea what happened and what to do? Thanks for any hint! Bye Uli -- _____________________________________________________________________ Uli Zappe E-Mail: uli@tallowcross.uni-frankfurt.de (NeXTMail,Mime,ASCII) PGP on request Lorscher Strasse 5 WWW: - D-60489 Frankfurt Fon: +49 (69) 9784 0007 Germany Fax: +49 (69) 9784 0042 staff member of NEXTTOYOU - the German NEXTSTEP/OPENSTEP magazine _____________________________________________________________________
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: YB/Win license fee suggestion Newsgroups: comp.sys.next.programmer Message-ID: <3599945f.0@news.depaul.edu> Date: 1 Jul 98 01:43:59 GMT I posted this to rhapsody-talk, but thought I'd post it here, too. (I'm just not getting enough spam lately.) Seems to me one solution to the $20 question would be for Apple to allow developers to license JUST the free bits. No AppKit, just Foundation, EOF, AIAT, etc. Pretty much like WebObjects, without WOF. Developers could then implement core functionality in a framework that would sit on top of the Apple frameworks. The developer would include a Win-compliant procedural API. A GUI would be implemented using some native Windows tool like Delphi or Visual C++. It would use the YB framework through the provided procedural API. (Naturally, this wouldn't work for AppKit-heavy applications. For less-visual data-oriented apps, this would work well.) When the DPS-free YellowBox ships, the developer switches to that. A YellowBox GUI would have been developed for in-house use and Rhapsody users. Windows users would get a much nicer version 2.0. This wouldn't be as good as shipping license-free 100% Yellow apps, but it does make Windows a more viable platform for ISV's, and it could allow SOME use of YellowBox, and a lot of code could be shared. Quite a lot, actually, if non-pure Java is used instead of Delphi or C++. As things stand now, developers are faced with "$20 or nothing", which is silly when Apple practically already has the answer. Thoughts? -- "... and subpoenas for all." - Ken Starr
From: boracay@hotmail.com Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Programming and the colossal failure to advance Date: Thu, 02 Jul 1998 13:48:58 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6ng34a$j2g$1@nnrp1.dejanews.com> Today an individual seeking to develop an application or have a computer fill a need which involves a programmed algorithm faces a most difficult task. Years ago programming was difficult but not extremely so. Programming complexity has now reached a level of complexity that is far beyond this and perhaps beyond where it should be. Programming today is little more than back breaking mind concentrating labor. In fact, computer programming takes longer, is harder and requires more attention than it ever has. More labor is being used to create the same functionality than ever before. The average individual may be able to define what they need their application or algorithm to do, or even be able to pseudocode or code the framework of the algorithm in matters of days. However programming into todays environments, including the popular OS’s and virtual machines (VM), may take months or years. While the GUI made using the computer easy, and a web browser made communicating with a computer easy, there has been little or no progress in making programming easy over the years. What is needed is to leap forward so that functional application requirements are important and here is where I spend my time, NOT the writing of tons or months of code to acheive it. The few tools developed to bridge this gap have been slow or were just fundamentally assistants to traditional languages. In order to use them we are often still forced to understand large portions of how the OS or VM is implemented and especially how the user interface is made order to program with them. Until some company is truly committed to helping the average computer owner to program, we will unfortunately continue with labor intensive programming. This has not changed with recent environments like JAVA. To this end here are some very simple guidelines for programming environment progress: 1)We should not need to understand details of the OS or VM design to write an application. We should know how to use the OS, and some simple ways we can hook our application into using it with ease (even with graphics), but we should not need to know details as we are forced today. We should be left to deal with our needs and functional requirements, not those of the computer, OS, user interface, virtual machine, and so on. 2)We should not be forced into understanding abstract objects and programming concepts which don’t relate to our functional needs. If we want to be abstract with our programming needs then that is our optional choice only. It shouldn’t be required to write a fast, efficient algorithm which solves our needs using a full GUI. Certainly by far, we should not need the abstraction that is required to simply use todays programming environments. 3)If only a computer scientist can program it or understand all the complexities surrounding development with it, and not your average scientist, engineer or other individual with basic computer training, then it is too complicated and is NOT an advancement. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 2 Jul 1998 14:31:58 GMT Organization: The World's Usenet -- http://www.Supernews.com Message-ID: <6ng5ku$nub$1@supernews.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: boracay@hotmail.com boracay@hotmail.com may or may not have said: -> Today an individual seeking to develop an application or have a computer fill -> a need which involves a programmed algorithm faces a most difficult task. -> Years ago programming was difficult but not extremely so. Programming -> complexity has now reached a level of complexity that is far beyond this and -> perhaps beyond where it should be. ??? we're not working with IBM 370 assembler anymore. -> Programming today is little more than back breaking mind concentrating -> labor. Sorry you're not enjoying it. I rather *like* the coding I do. ->In fact, computer programming takes -> longer, is harder and requires more attention than it ever has. Then you're using the wrong tools. Try using a modern OO development system, with a decent class library. OpenStep, Smalltalk, NewtonScript and Smalltalk are all pretty good. -> More labor is being used to create the same functionality than ever -> before. Not so, unless you're trying to get a job done in a broken environment, with a crippled language like C++. ->The average -> individual may be able to define what they need their application or -> algorithm to do, or even be able to pseudocode or code the framework of the -> algorithm in matters of days. And in NeXTSTEP, I can typically limit my work to just writing that core algorithm. In most cases, the GUI, the printing code, the file I/O, etc, I can use without modification. -> However programming into todays environments, including the popular OS’s -> and virtual machines (VM), may take months or years. Yeah, well: most tools suck, and so do many developers. -jcr
From: torbenm@diku.dk (Torben AEgidius Mogensen) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 2 Jul 1998 17:36:54 +0200 Organization: Department of Computer Science, U of Copenhagen Sender: torbenm@tyr.diku.dk Message-ID: <6ng9em$h77@tyr.diku.dk> References: <6ng34a$j2g$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit boracay@hotmail.com writes: >Today an individual seeking to develop an application or have a computer fill >a need which involves a programmed algorithm faces a most difficult task. >Years ago programming was difficult but not extremely so. Programming >complexity has now reached a level of complexity that is far beyond this and >perhaps beyond where it should be. Programming today is little more than back >breaking mind concentrating labor. In fact, computer programming takes >longer, is harder and requires more attention than it ever has. More labor is >being used to create the same functionality than ever before. I disagree. The problem is that costumers expect _more_ functionality, in terms of user interface and interoperability with other programs than they used to do. > The average >individual may be able to define what they need their application or >algorithm to do, or even be able to pseudocode or code the framework of the >algorithm in matters of days. However programming into todays environments, >including the popular OS’s and virtual machines (VM), may take months or >years. Which is mainly due to the added functionality requirement. >While the GUI made using the computer easy, and a web browser made >communicating with a computer easy, there has been little or no progress in >making programming easy over the years. That is not entirely true. However, the problem is that the research in this area (programming langauges) hasn't moved into the main stream of programming. And the main reason for this is that to do so, new languages have to interface to the complex GUI's etc. thata re around. > What is needed is to leap forward so >that functional application requirements are important and here is where I >spend my time, NOT the writing of tons or months of code to acheive it. The >few tools developed to bridge this gap have been slow or were just >fundamentally assistants to traditional languages. In order to use them we >are often still forced to understand large portions of how the OS or VM is >implemented and especially how the user interface is made order to program >with them. Until some company is truly committed to helping the average >computer owner to program, we will unfortunately continue with labor >intensive programming. This has not changed with recent environments like >JAVA. To this end here are some very simple guidelines for programming >environment progress: >1)We should not need to understand details of the OS or VM design to write an >application. We should know how to use the OS, and some simple ways we can >hook our application into using it with ease (even with graphics), but we >should not need to know details as we are forced today. We should be left to >deal with our needs and functional requirements, not those of the computer, >OS, user interface, virtual machine, and so on. I agree. The main reason for this not alrady having happened is that 90% of computers use a single OS and a single GUI. Hence, there has been little incentive for application companies to move to an OS-independent methodology. And the OS vendor (no plural here) has little incentive to make the API easy to move to other OS's, as this may threathen his monopoly. >2)We should not be forced into understanding abstract objects and programming >concepts which don’t relate to our functional needs. If we want to be >abstract with our programming needs then that is our optional choice only. It >shouldn’t be required to write a fast, efficient algorithm which solves our >needs using a full GUI. Certainly by far, we should not need the abstraction >that is required to simply use todays programming environments. What we need is probably to separate the user interface from the functionality. At the moment, this is hard to do. But you should ideally be able in your program to say "I present a binary choice to the user" without having to specify widgets, polling loops etc. The actual appearance of the choice should be up to the GUI itself to decide, possibly guided by preferences or GUI designer programs that given the set of abstract interface operations specified by the your program allows you to edit a "look-and-feel". >3)If only a computer scientist can program it or understand all the >complexities surrounding development with it, and not your average scientist, >engineer or other individual with basic computer training, then it is too >complicated and is NOT an advancement. Actually, I would rather put it this way: If only nerds that enjoy reading 300+ page manuals can use the environment, then it is too complicated. The problem isn't that the GUI libraries etc. are conceptually difficult, it is the amount of detail that is the problem. Torben Mogensen (torbenm@diku.dk)
From: Alan Gauld <alan.gauld@gssec.bt.co.uk> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Thu, 02 Jul 1998 17:12:19 +0100 Organization: BT Message-ID: <359BB163.7A97DEC6@gssec.bt.co.uk> References: <6ng34a$j2g$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit boracay@hotmail.com wrote: > Today an individual seeking to develop an application or have a computer fill > a need which involves a programmed algorithm faces a most difficult task. Not so. if he doesn't mind a teletype interface then its no more difficult(indeed much easier!) than it was 10 years ago - and light years easier than it was 20 or 30 years ago. - No punched cards, no single line paper teletypes, no line oriented editors etc etc. Even if it requires a GUI environments like VB and Delphi enable even fairly complex poroblems to be solved with realative ease 0- much more than I could do on my Commodore PET in the early '80s. > Years ago programming was difficult but not extremely so. Years ago, people worked mainly in assembler - in the very early ywars they typed in binary(or octal usually) codes by hand - not even having assemblers! On PCs this was still the norm into the mid '80's. > Programming today is little more than back > breaking mind concentrating labor. I would have agreed with this between 1990 and 1995 but the tools available in recent years have changed my mind - its now only as hard as it was 20 years ago.... > longer, is harder and requires more attention than it ever has. Simply not true - see above. > being used to create the same functionality than ever before. Not true - the functionality is orders of magnitude greater. Users expectations have risen dramatically. The core functions may not have changed but the peripheral functionality is enormous - compare MS Word 97 with Wordstar V3 say. The output looks the same the functionality in the programs is worlds apart. > While the GUI made using the computer easy, and a web > browser made communicating with a computer easy, > there has been little or no progress in making programming easy I doubt it will ever be *easy*. We are persuading a machine which can add 1+1 and copy data from A to B to do all manner of tasks. Programming environmentsd hide that fact from us but ultimately codifying human actions will allways be hard. It can become easier but never easy! IMO. > 1)We should not need to understand details of the OS or VM design The bulk of MIS programs writtenm today in VB fit this criteria. > 2)We should not be forced into understanding abstract objects Why not? If that what makes programming easier? An average person can build a bookcase from a kit but still needs to understand how screwdrivers and screws interact. The issue is with the complexity of the models not the need to use a model. > 3)If only a computer scientist can program it or understand all the > complexities surrounding development with it, and not your average > scientist, engineer The average computer scientist can't design computer hardware, why would the average electrical engineer be expected to design a modern software application? Far less someone whose field is in a radically different area like genetics. You seem, to assume that computers shouyld somehow be easy - why? What they do is intrinsically complex. If it were easy, true Turing machines would be widely available. Alan G.
From: "Kelvin Khong" <wendyk@pl.jaring.my> Newsgroups: comp.sys.next.programmer Subject: yellow box runtime Date: Wed, 1 Jul 1998 22:56:26 +0800 Organization: Unconfigured Message-ID: <6ndiiu$172@news2.jaring.my> i'm not a registered apple developer, but i'd like to obtain the yellow box runtime environment for windows 95 so i can run some yellow box applications. anyone knows where i can find it ?? thanks !
From: sdm7g@elvis.med.Virginia.EDU (Steven D. Majewski) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 3 Jul 1998 00:31:48 GMT Organization: University of Virginia Message-ID: <6nh8pk$82k$1@murdoch.acc.Virginia.EDU> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6ng8c3$vb52@onews.collins.rockwell.com> In article <6ng8c3$vb52@onews.collins.rockwell.com>, Erik M. Buck <embuck@palmer.cca.rockwell.com> wrote: >This is just silly. > >A computer scientist does not expect to understand geology, genetics, >surgery, particle physics, systems engineering, or even ditch digging to any >greater degree than the average population and certainly not to any degree of >professional skill. Why would a geologist, surgeon, physicist, systems >engineer, or ditch digger expect to know enough about computer science to >exhibit professional skill in computer science ? > >People are tantalized by how easy and powerful programming seems at first. >It is not a problem that computers are too hard; it is that they are >approachable. Almost anyone can learn intro programming. Why would that >person expect to be any more capable as a computer scientist than a computer >scientist would be at geology after an intro geology course. > >It is condescending in the extreme to claim the computer science should be >easy enough for anyone. This is as large and hard a proffesion as any other >and people do not expect geology to be instantly understood by every novice >that comes along. > But there's a difference between computer science and programming. One of the goals of computer science is to make programming easier. If it is exactly as difficult to teach programming to geologists or teach geology to programmers, then we will never get many good geology application programs. One of them has got to become easier. I expect the number of facts and techniques needed to master both geology and computer science to grow. However, if computer scientists do their job, programming should get easier and simpler. And in fact, it you hold constant the complexity of the job you're trying to accomplish, and you avoid poorly designed systems ( like C++ and MFC ), programming has in fact gotten easier. ---| Steven D. Majewski (804-982-0831) <sdm7g@Virginia.EDU> |--- ---| Department of Molecular Physiology and Biological Physics |--- ---| University of Virginia Health Sciences Center |--- ---| P.O. Box 10011 Charlottesville, VA 22906-0011 |--- "I'm not as big a fool as I used to be, I'm a smaller fool." - Jack Kerouac Some of the Dharma <http://members.aol.com/kerouacsis/SomeDharma.html>
From: sdm7g@elvis.med.Virginia.EDU (Steven D. Majewski) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 3 Jul 1998 00:18:46 GMT Organization: University of Virginia Message-ID: <6nh816$7r2$1@murdoch.acc.Virginia.EDU> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359BB163.7A97DEC6@gssec.bt.co.uk> <6nh07l$hl$1@nnrp1.dejanews.com> In article <6nh07l$hl$1@nnrp1.dejanews.com>, <hmmueller@my-dejanews.com> wrote: >In article <359BB163.7A97DEC6@gssec.bt.co.uk>, > alan.gauld@gssec.bt.co.uk wrote: >> >> The average computer scientist can't design computer hardware, >> why would the average electrical engineer be expected to >> design a modern software application? Far less someone whose >> field is in a radically different area like genetics. You seem, >> to assume that computers shouyld somehow be easy - why? >> What they do is intrinsically complex. > >Thanks for saying that! I dont know why, but there is still this idea around >that computers should be programmed by laymen - electrical engineers, >astronomers, mechanical engineers etc. Nobody would allow me to design >the smallest pedestrians' bridge over a small river - but somehow people (and >companies!) believe that everybody who USED a computer should program it. > That expectation may not be reasonable, but there is a reasonable reason why people want that: Because, whatever way you look at it, the choices come down to (to pick one example) either teaching biology to computer programmers or teaching programming to biologists. The more biologists or physicists learn about their domain, the more has to be learned to master that subject. I think there is also a lot more "Computer Science" to digest that their was 20 years ago. But, since this is an engineering and design discipline, there is the hope that we can construct artifacts ( tools and programming languages, object libraries ) that hide some of that complexity and appear to make programming simple. So far, we haven't done a very good job of that. But of the two choices, I'ld bet we have a better chance of making programming simple enough for a biologist or physicist to master, than we do of teaching a useful amount of biology or physics to programmers. [ One way around that is to build a team of programmers and domain experts, but then that require that you have a manager that really knows what he's doing, and teaching someone to be a *good* manager if probably even more difficult that teaching someone to program. ] ---| Steven D. Majewski (804-982-0831) <sdm7g@Virginia.EDU> |--- ---| Department of Molecular Physiology and Biological Physics |--- ---| University of Virginia Health Sciences Center |--- ---| P.O. Box 10011 Charlottesville, VA 22906-0011 |--- "I'm not as big a fool as I used to be, I'm a smaller fool." - Jack Kerouac Some of the Dharma <http://members.aol.com/kerouacsis/SomeDharma.html>
Message-ID: <359C50BC.770F2812@home.com> From: John Mott <johnmott@home.com> Organization: @Home Network MIME-Version: 1.0 Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming To: "Steven D. Majewski" <sdm7g@elvis.med.Virginia.EDU> Subject: Re: Programming and the colossal failure to advance References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359BB163.7A97DEC6@gssec.bt.co.uk> <6nh07l$hl$1@nnrp1.dejanews.com> <6nh816$7r2$1@murdoch.acc.Virginia.EDU> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Fri, 03 Jul 1998 03:38:43 GMT NNTP-Posting-Date: Thu, 02 Jul 1998 20:38:43 PDT Steven D. Majewski wrote: > Because, whatever way you look at it, > the choices come down to (to pick one example) either teaching biology > to computer programmers or teaching programming to biologists. > > The more biologists or physicists learn about their domain, the more > has to be learned to master that subject. > For certain "degrees of freedom" I agree with that. I first learned computers as a physics major, because it was a tool I had to learn. I did *not* learn programming as a craft there but I did learn how to convert the problems into FORTRAN to derive answers to problems. It is very reasonable to have, as a goal, the ability for someone in their field be able to sit down and utilize the computer to solve problems, and its reasonable to assume that, depending on the field, more or less individual attention has to be paid to the development of the logic involved in answering questions. I doubt any biologist would be interested in taking 6 months to develop the graphics engine used in visualization of molecules, but I'm certain they would be willing to spend some hours learning to specify, in some meta-langauge, enough to accomplish some specific goal which involved visualization of molecules. (forgive the hand waving, I don't know how biologists use computers :-) A good analogy would be a 'cosmo quiz' kind of test to take to see if you had a genetic predisposition to something based on the specification of the conditions of antecedents. I could check boxes in such a thing, in essence 'programming it', and arrive at a result which was useful within that limited context. I could not color outside the lines by using that same test for a condition which might be *almost identical* in inheritance probabilities because the test I'm taking is constrained to a particular problem. Its not reasonable to expect that genetics as a discipline be "made simpler" so that I can find out if I have the such and such gene, but it is possible to present a properly and reasonably constrained "program" which would guide me to that sort of result. > I think there is also a lot more "Computer Science" to digest that > their was 20 years ago. But, since this is an engineering and design > discipline, there is the hope that we can construct artifacts ( tools > and programming languages, object libraries ) that hide some of that > complexity and appear to make programming simple. > I agree, but only for specific named tasks with strictly defined constraints *and* a recognition that there are lines there that cannot be crossed without devolving into the generalizations which making the specific tool no longer relavent to the task. > So far, we haven't done a very good job of that. > But of the two choices, I'ld bet we have a better chance of making > programming simple enough for a biologist or physicist to master, than > we do of teaching a useful amount of biology or physics to programmers. > I would agree with that with the condition that the 'programming' required be constrained within a (hopefully soft) box. You are simply not going to interest any non-computer professional in resolving the problems associated with corrupt B-Trees or memory leaks or spurious hardware interrupts or uninitialized stack variables or fencepost logic conditions or the format of a page table entry or resolving file system corruptions or semaphore locking, etc. etc. etc. John Mott john@sandh.com http://www.iplsystem.com/ ; A nice soft box to write web programs in!
From: ppith@my-dejanews.com Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Thu, 02 Jul 1998 22:57:08 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6nh384$4mj$1@nnrp1.dejanews.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> > breaking mind concentrating labor. In fact, computer programming takes > longer, is harder and requires more attention than it ever has. More labor is > being used to create the same functionality than ever before. The average > individual may be able to define what they need their application or > algorithm to do, or even be able to pseudocode or code the framework of the > algorithm in matters of days. However programming into todays environments, > including the popular OS s and virtual machines (VM), may take months or > years. Actually, I think it depends on how you're developing your project. I think people who want more ease in developing code should try some black box program like Matlab since "the days of writing your own C++ string class are over." The popularity of GUI systems has forced a lot of us to consider visual versions of whatever language we're comfortable developing in...to stay competitive. > making programming easy over the years. What is needed is to leap forward so > that functional application requirements are important and here is where I > spend my time, NOT the writing of tons or months of code to acheive it. The > few tools developed to bridge this gap have been slow or were just > fundamentally assistants to traditional languages. In order to use them we Search for programs like "Purify" for any programming language that is difficult to develop in. I'd also recommend reading Dr. Dobb's Journal. > are often still forced to understand large portions of how the OS or VM is > implemented and especially how the user interface is made order to program > with them. Until some company is truly committed to helping the average gah. Why not learn multiple OS systems? Sure 80-90% of the world uses Windows desktops, but why not learn UNIX too? Besides, you can write more powerful applications once you start using system commands to assist you. <snip> Okay, now I'm addressing the fact you didn't like reading everything you could about a programming language. It's your choice how much you want to learn about a language. You can learn barebones C++ and write small, non-interesting programs. Or you can (this applies to any language) go all out and learn how to use socket classes, system commands, reuseable code, multi threaded apps... Depending on how much you're willing to learn(for fun, money, or challenge) you can't expect to live in a world where programming is easy enough for any engineer to simply "pick up" if they're bored of their job or laid off. I think it's a constant and lifelong learning process, and keeping up with new languages (or learning ones you don't know) is part of the fun. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <iXIL5vjp9GA.152@cidintnews.infosel.com.mx> Control: cancel <iXIL5vjp9GA.152@cidintnews.infosel.com.mx> Date: 03 Jul 1998 05:21:08 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.iXIL5vjp9GA.152@cidintnews.infosel.com.mx> Sender: PC Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 2 Jul 98 21:29:35 Organization: Is a sign of weakness Message-ID: <SCOTT.98Jul2212935@slave.doubleu.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6ng8c3$vb52@onews.collins.rockwell.com> In-reply-to: embuck@palmer.cca.rockwell.com's message of 2 Jul 1998 15:18:27 GMT In article <6ng8c3$vb52@onews.collins.rockwell.com>, embuck@palmer.cca.rockwell.com (Erik M. Buck) writes: This is just silly. <...> People are tantalized by how easy and powerful programming seems at first. It is not a problem that computers are too hard; it is that they are approachable. You've hit the nail on the head - so far as I'm concerned, that should be the end of the discussion! In a project I've been involved with for much of the past five years, we repeatedly have this problem. Every new person who comes onboard _immediately_ sees to the core simplicity of the problem we're solving, and has all sorts of wonderous ideas. Depending on their character, after a couple months they either decide that we're all idiots (because we willfully ignore the simplicity of what they think we should add), or they decide that we're all geniusses (because they not only see the sometimes inspired work we've done - they understand why anything less than "inspired" wouldn't have cut it). I suspect that part of the problem is that humans can be so flexibly programmed. When you say "turn left at the McDonalds" to a human, you can trust them to notice that they're in a residential neighborhood and if they see a mailbox with "McDonalds" on it, turn there. You don't have to tell them to close the paint can before putting it in the trunk. You don't have to tell them to close the freezer door after putting the ice cream away. Nine times out of ten, humans can be treated as Do What I Mean devices. Computers are _always_ Do What I Say devices. Even then, artificial intelligence isn't the answer, because you probably won't stand for it if a computer misunderstood you as often as humans do. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
Message-ID: <359C16EA.5AF08B84@home.com> From: John Mott <johnmott@home.com> Organization: @Home Network MIME-Version: 1.0 Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming Subject: Re: Programming and the colossal failure to advance References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6nh384$4mj$1@nnrp1.dejanews.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Thu, 02 Jul 1998 23:32:02 GMT NNTP-Posting-Date: Thu, 02 Jul 1998 16:32:02 PDT ppith@my-dejanews.com wrote > > Okay, now I'm addressing the fact you didn't like reading everything you > could about a programming language. It's your choice how much you want to > learn about a language. You can learn barebones C++ and write small, > non-interesting programs. Or you can (this applies to any language) go all > out and learn how to use socket classes, system commands, reuseable code, > multi threaded apps... > > Depending on how much you're willing to learn(for fun, money, or challenge) > you can't expect to live in a world where programming is easy enough for any > engineer to simply "pick up" if they're bored of their job or laid off. I > think it's a constant and lifelong learning process, and keeping up with new > languages (or learning ones you don't know) is part of the fun. > There are a few hard facts about programming as a trade/profession. Its a science to the degree that its a study of nature; logic and problem solving all within the constraints of nature. To that end its simply as hard as it is. I have observed that there have been many tools developed and attempts made to 'eliminate the programmer'. Cobol was supposed to eliminate the programmer because it was so wordy that it was presumed that secretaries could learn to do it. There is an inevitable quality to all of these attempts, an attempt to hide the 'heavy lifting' that cannot be avoided for completely general solutions. Show someone a self-contained perfect solution for some reasonably narrow problem and the first thing you know you hear "but I need for it to do *this*" and it can't, or an unforseen sequence or combination occurs. There is a simple reason for this: the universe is infinitely complex. Its nature. Its inevitable. It can't be worked around. Programs save people time who need to accomplish a task of some sort, but they will never be able to predict what people really want but cannot express aside from very restricted environments with limited outcomes and choices. Embrace the mystery! Bow to the absurd! I can feel a new religion in here somewhere... :-) John Mott Chief Magician, S&H john@sandh.com http://www.iplsystem.com/ ; Magical web programming...
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 2 Jul 1998 19:03:21 GMT Organization: Technical University of Berlin, Germany Message-ID: <6nglhp$c5o$1@news.cs.tu-berlin.de> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6ng9em$h77@tyr.diku.dk> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit torbenm@diku.dk (Torben AEgidius Mogensen) writes: >boracay@hotmail.com writes: [...] >>While the GUI made using the computer easy, and a web browser made >>communicating with a computer easy, there has been little or no progress in >>making programming easy over the years. >That is not entirely true. However, the problem is that the research >in this area (programming langauges) hasn't moved into the main stream >of programming. Yes, it is dizzying to see the Java hype when *better* systems existed 15-20 years ago. >And the main reason for this is that to do so, new >languages have to interface to the complex GUI's etc. thata re around. I think it has to do with the acceptance of mediocrity, the unwillingness to learn and to a good extent with the debilitating effect MS has had on innovation in the industry. [...] >What we need is probably to separate the user interface from the >functionality. Actually: seperate the interfaces from the functionality. Functionality can have multiple interfaces: a C-level API, a scripting interface, pipe-filter interface, GUI and Web- frontend. Functionality and interfaces should be orthogonal, meaning that if I add functionality, it should be automagically available in all interfaces, and if I add an interface, it should easily encompass all functionality. >At the moment, this is hard to do. Understatement... :-) [GUI-specific ideas snipped] Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: hmmueller@my-dejanews.com Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Thu, 02 Jul 1998 22:05:41 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6nh07l$hl$1@nnrp1.dejanews.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359BB163.7A97DEC6@gssec.bt.co.uk> In article <359BB163.7A97DEC6@gssec.bt.co.uk>, alan.gauld@gssec.bt.co.uk wrote: > > 3)If only a computer scientist can program it or understand all the > > complexities surrounding development with it, and not your average > > scientist, engineer > > The average computer scientist can't design computer hardware, > why would the average electrical engineer be expected to > design a modern software application? Far less someone whose > field is in a radically different area like genetics. You seem, > to assume that computers shouyld somehow be easy - why? > What they do is intrinsically complex. Thanks for saying that! I dont know why, but there is still this idea around that computers should be programmed by laymen - electrical engineers, astronomers, mechanical engineers etc. Nobody would allow me to design the smallest pedestrians' bridge over a small river - but somehow people (and companies!) believe that everybody who USED a computer should program it. I've seen many non-computer engineering engineers doing program - and I know now that they usually can't do it - neither the distinction between a good and a bad hash function nor between a good and a bad UI nor between a good and a bad test case or test plan can usually be learned "on the job". You need some background. - ok, that was harsh, maybe. So let me add: Everybody can learn new things, also computer programming. However, it will take a few years; and if you do it on your own, you'll like not exceed mediocre status. Harald M. Mueller -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: Matt Atterbury <matt@qbd.com.au> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Fri, 03 Jul 1998 18:49:43 +1000 Organization: Quality By Design Message-ID: <359C9B27.5819@qbd.com.au> References: <6ng34a$j2g$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Cache-Post-Path: kermit.netlink.com.au!unknown@h081.mel.netlink.com.au boracay@hotmail.com wrote: > = > Today an individual seeking to develop an application or have a compute= r fill > a need which involves a programmed algorithm faces a most difficult tas= k. > Years ago programming was difficult but not extremely so. Programming > complexity has now reached a level of complexity that is far beyond thi= s and > perhaps beyond where it should be. Programming today is little more tha= n back > breaking mind concentrating labor. In fact, computer programming takes > longer, is harder and requires more attention than it ever has. More la= bor is > being used to create the same functionality than ever before. The avera= ge ^^^^^^^^^^^^^^^^^^ I don't know what you do or how old you are, but this is just plain WRONG. Systems today are MUCH more complex/sophisticated than ever before, mainly due to customers' increasing demands for functionality, interoperability, etc. Twenty years ago, the complex systems were accounting packages, compilers, operating systems, airline reservation systems, banking systems, and the like - today, most of these are either off-the-shelf or buildable from a large suite of components. These systems were often batch oriented. They rarely communicated in real-time with other systems, rather they usually dumped files and loaded dump files. The kind of inter-system linkage we see these days (even on something as "simple" as a PC, via OLE etc) was simply unheard of; now however, if your system can't talk to other systems, it's dead in the water. Users have become more demanding (unfortunately, often for things that they do not really need and/or are not worth the cost of development), and the systems have become correspondingly more complex. This will NOT change, even if your magic bullet comes along - as soon as some bright spark develops something whizzy with it, a user will nice, "That's nice, but why can't it do this, and this, and ...". > individual may be able to define what they need their application or > algorithm to do, or even be able to pseudocode or code the framework of= the > algorithm in matters of days. However programming into todays environme= nts, > including the popular OS=92s and virtual machines (VM), may take months= or > years. > = > [...] -- = -------------------------------------------------------------------------= ---- matt@qbd.com.au (ignore all other addresses :-) "klaatu barada nikto"
From: "resslere" <resslere@erols.com> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 3 Jul 1998 04:51:47 GMT Organization: Ressler Family Message-ID: <01bda63e$276733e0$507aaccf@zen> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6ng8c3$vb52@onews.collins.rockwell.com> <6nh8pk$82k$1@murdoch.acc.Virginia.EDU> Steven D. Majewski <sdm7g@elvis.med.Virginia.EDU> wrote in article > But there's a difference between computer science and programming. > One of the goals of computer science is to make programming easier. You are confused. Computer science has done a lot to make programming easier. But there is a big difference between computer science research results and programming systems the market/industry will support. Get VB and be happy... Gene
From: "resslere" <resslere@erols.com> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 3 Jul 1998 05:02:18 GMT Organization: Ressler Family Message-ID: <01bda63f$a0056320$507aaccf@zen> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359BB163.7A97DEC6@gssec.bt.co.uk> <6nh07l$hl$1@nnrp1.dejanews.com> <359C25D4.7B2A@computer.org> This is a fair point, but in fact there are more than a few brilliant computer scientists including some "founding fathers" are unexceptional programmers and system designers, just as there are brilliant metalurgists who can't design or build a bridge. Gene B. Thomas <b.thomas@computer.org> wrote in article <359C25D4.7B2A@computer.org>... > > - ok, that was harsh, maybe. So let me add: Everybody can learn new things, > > also computer programming. However, it will take a few years; and if you do > > it on your own, you'll like not exceed mediocre status. > > Since when was computer science taught in universities? > > Most of the founding fathers of computer science came from other fields > - mathematics, electrical engineering etc. Do you imply that they were > mediocre too, since they were also self-taught (though most of that > self-teaching process was part of research). > > Biju Thomas >
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 2 Jul 98 21:02:20 Organization: Is a sign of weakness Message-ID: <SCOTT.98Jul2210220@slave.doubleu.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> In-reply-to: boracay@hotmail.com's message of Thu, 02 Jul 1998 13:48:58 GMT In article <6ng34a$j2g$1@nnrp1.dejanews.com>, boracay@hotmail.com writes: 1)We should not need to understand details of the OS or VM design to write an application. Agreed. 2)We should not be forced into understanding abstract objects and programming concepts which don’t relate to our functional needs. If we want to be abstract with our programming needs then that is our optional choice only. It shouldn’t be required to write a fast, efficient algorithm which solves our needs using a full GUI. Certainly by far, we should not need the abstraction that is required to simply use todays programming environments. I have _no_ idea where you're going, here. You've got the cart before the horse. The abstract objects and programming concepts have evolved as an attempt to model functional needs. It wasn't like some egghead invented object oriented programming and design and then everyone in the world decided that we should try to cram every project into that mold. [In fact, it's much the opposite, OOP&D's ability to simplify projects was pretty much ignored until recently. <saracasm type="heavy">Now, of course, we're _all_ object oriented programmers</saracasm>] Do you think that programmers just sit around inventing bizarre new concepts and then fit your functional needs to the concepts because we _enjoy_ it? Well, ok, do you think we _all_ enjoy it? Do you think that multi million dollar projects with the cream of the crop programming staff routinely fail because the functional needs are too _simple_? In general, projects fail because people _vastly_ underestimate the complexity of the "simple functional spec". I don't mean that they forget to specify the exact color they want. I mean that the parts they don't describe up front are generally five to ten times more complex than the parts they do describe. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 2 Jul 98 21:16:41 Organization: Is a sign of weakness Message-ID: <SCOTT.98Jul2211641@slave.doubleu.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6ng8c3$vb52@onews.collins.rockwell.com> <6nh8pk$82k$1@murdoch.acc.Virginia.EDU> In-reply-to: sdm7g@elvis.med.Virginia.EDU's message of 3 Jul 1998 00:31:48 GMT In article <6nh8pk$82k$1@murdoch.acc.Virginia.EDU>, sdm7g@elvis.med.Virginia.EDU (Steven D. Majewski) writes: In article <6ng8c3$vb52@onews.collins.rockwell.com>, Erik M. Buck <embuck@palmer.cca.rockwell.com> wrote: >This is just silly. > >A computer scientist does not expect to understand geology, <...> >It is condescending in the extreme to claim the computer science >should be easy enough for anyone. This is as large and hard a >proffesion as any other and people do not expect geology to be >instantly understood by every novice that comes along. But there's a difference between computer science and programming. One of the goals of computer science is to make programming easier. Not necessarily. Computer science is just as successful when it allows you to tackle larger problems with similar expenditures. So, you might not be able to write a word processor in a day - but you can at least conceive of writing real time derivatives trading system or a grandmaster chess program. That's progress, too. If it is exactly as difficult to teach programming to geologists or teach geology to programmers, then we will never get many good geology application programs. One of them has got to become easier. Why is that? Perhaps 80% of my worthwhile programming skills have been learned over a relatively long period of time. I could have easily fit in a couple more B.S. level degrees, and perhaps a M.S. along the way. [Sorry, hard sciences and match only, no humanities or "social sciences" for me, thank you very much :-).] I expect the number of facts and techniques needed to master both geology and computer science to grow. However, if computer scientists do their job, programming should get easier and simpler. You might as easily expect geology to get easier and simpler, too. Does research in Neural Networks make people developing banking applications more productive? Why can't computer science also have the goal of expanding the horizons it applies to? Actually, today's geology will get easier, because over time more and more of the grunt-work will be pushed down into the tools. This doesn't necessarily take an expert. Say you needed a tool to measure the gravitational gradient in an area. These days that hardly takes an expert in geology or physics to design. And it doesn't necessarily take a top-10% programmer to program it's microcontroller. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
Message-ID: <359CA7B3.41C6@iris.rz.uni-konstanz.de> Date: Fri, 03 Jul 1998 11:43:15 +0200 From: ldbuchho <ldbuchho@iris.rz.uni-konstanz.de> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Bakoma-Fonts Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Organization: University of Constance, Germany Hi, does anyone know how to convert the Bakoma-TeX-Font selection into normal PS-Fonts, so that I can see them in the Fontpanel and use the special "math"-symbols in an ordinary text dokument? I am grateful for any suggestion -- Lars Dierk Buchholz
From: embuck@palmer.cca.rockwell.com (Erik M. Buck) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 2 Jul 1998 15:18:27 GMT Organization: Rockwell International Message-ID: <6ng8c3$vb52@onews.collins.rockwell.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: boracay@hotmail.com This is just silly. A computer scientist does not expect to understand geology, genetics, surgery, particle physics, systems engineering, or even ditch digging to any greater degree than the average population and certainly not to any degree of professional skill. Why would a geologist, surgeon, physicist, systems engineer, or ditch digger expect to know enough about computer science to exhibit professional skill in computer science ? People are tantalized by how easy and powerful programming seems at first. It is not a problem that computers are too hard; it is that they are approachable. Almost anyone can learn intro programming. Why would that person expect to be any more capable as a computer scientist than a computer scientist would be at geology after an intro geology course. It is condescending in the extreme to claim the computer science should be easy enough for anyone. This is as large and hard a proffesion as any other and people do not expect geology to be instantly understood by every novice that comes along. The argument seems to be: Premise 1: Software development tasks are too difficult for a novice to be proficient Premise 2: There is no need for it to be so difficult Premise 3: It is desirable for novices to be proficient at software development -------------------------------------------- Conclusion: Software development should be made easier Premise 2 and premise 3 are false Consider: Premise 1: Geology tasks are too difficult for a novice to be proficient Premise 2: There is no need for it to be so difficult Premise 3: It is desirable for novices to be proficient at geology -------------------------------------------- Conclusion: Geology should be made easier
From: "Scott P. Duncan" <softqual@mindspring.com> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Fri, 03 Jul 1998 09:40:00 -0400 Organization: SoftQual (Quality That Lasts) Message-ID: <359CDF2B.D3BA8A8@mindspring.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6ng8c3$vb52@onews.collins.rockwell.com> <SCOTT.98Jul2212935@slave.doubleu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; x-mac-type="54455854"; x-mac-creator="4D4F5353" Content-Transfer-Encoding: 7bit Scott Hess wrote: > You > don't have to tell them to close the paint can before putting it in > the trunk. You don't have to tell them to close the freezer door > after putting the ice cream away. I guess you don't have any children! :-)
From: "Judson McClendon" <judmc123@bellsouth.net> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> Subject: Re: Programming and the colossal failure to advance Message-ID: <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> Date: Fri, 03 Jul 1998 13:08:52 GMT NNTP-Posting-Date: Fri, 03 Jul 1998 09:08:52 EST Scott Hess wrote: > > It wasn't like some egghead invented object oriented programming and > design and then everyone in the world decided that we should try to > cram every project into that mold. Actually, that's a pretty good description of what DID happen. Had OO been invented by the folks in the trenches, not only would it have been far easier to learn and use than it is, but it would be time and cost effective, which it has certainly not proven to be, in real world practice. The movement for OO came from the universities, not the field. Any time you see all the universities trumpeting for something which has never been proven practical in practice, you can *know* it comes from academia. :-) -- Judson McClendon This is a faithful saying and worthy of all Sun Valley Systems acceptance, that Christ Jesus came into the judmc123@bellsouth.net world to save sinners (1 Timothy 1:15) (please remove numbers from email id to respond)
Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng From: Mathias Holmgren <mathias.holmgren@ecs.se> Subject: Re: Programming and the colossal failure to advance Message-ID: <359CF6CB.68EB68E3@ecs.se> Date: Fri, 03 Jul 1998 17:20:43 +0200 References: <6ng34a$j2g$1@nnrp1.dejanews.com> Organization: Entra Capital Solutions MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit boracay@hotmail.com wrote: > > In fact, computer programming takes > longer, is harder and requires more attention than it ever has. More labor is > being used to create the same functionality than ever before. I'd say it it very easy and quick nowadays to wite, for instance, a program that acts as a calculator compared to when you had to do it by punching cards. Is that not progress? > While the GUI made using the computer easy, and a web browser made > communicating with a computer easy, there has been little or no progress in > making programming easy over the years. What is needed is to leap forward so > that functional application requirements are important and here is where I > spend my time, NOT the writing of tons or months of code to acheive it. The > few tools developed to bridge this gap have been slow or were just > fundamentally assistants to traditional languages. What do you expect? The HAL computer of 2001, a space oddyssey? Programming is difficult by it's nature, period. > Until some company is truly committed to helping the average > computer owner to program, we will unfortunately continue with labor > intensive programming. This is just plain out rediculous. Consider this line of argument: "Until some company is truly committed to helping the average car owner to design and build a car from scratch, we will unfortunatelly continue being dependent on labor intensive car manufacturing." > 2)We should not be forced into understanding abstract objects and programming > concepts which don’t relate to our functional needs. If we want to be > abstract with our programming needs then that is our optional choice only. It > shouldn’t be required to write a fast, efficient algorithm which solves our > needs using a full GUI. Certainly by far, we should not need the abstraction > that is required to simply use todays programming environments. You just don't get it, do you? The abstract objects and concepts you refere to very much relates to your functional needs, namely by making possible to forfill your functional needs _at all_. First you say it is to much labor to get into the details of programming, then you dislike abstract thinking and programming concepts. I think I know why. Because they are both needed to produce programs, and neither can be mastered by someone not willing to learn. > 3)If only a computer scientist can program it or understand all the > complexities surrounding development with it, and not your average scientist, > engineer or other individual with basic computer training, then it is too > complicated and is NOT an advancement. You are pathetic, do you really think that someone with "basic computer training" will _ever_ be competent to produce anything that looks like a program? That's like saying that someone who knows how to start a car should be able to design and build a racecar. What do you think a computer is, a simple pocket calculator? BTW: Your definition of advancement: Something that can be mastered by anyone with nothing more than basic training. Wow, I guess you are thinking of the evolution of the bathroom. I assume that you have had your basic training... The reason I even bothered to answer this post is the implicit insult to us highly educated people who actually know how to do this. This moron obviously views us the lowest scum on the earth and probably thinks that what he wants has been possible for years, but somehow withheld by us. </FLAME> All programms should be written with the user, and his/her needs as the most important factor at hand. Everything in the program should be made as simple as possible for the user. This is the way all users expect it to be, including myself. -- Mathias Holmgren Entra Capital Solutions +46 70 979 08 83 http://www.ecs.se/
From: Alan Gauld <alan.gauld@gssec.bt.co.uk> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Fri, 03 Jul 1998 16:44:00 +0100 Organization: BT Message-ID: <359CFC40.E80D1E55@gssec.bt.co.uk> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Judson McClendon wrote: > Actually, that's a pretty good description of what DID happen. Had OO > ..... The movement for OO came from the universities, not the field. Could you justify that statement? Simula 67, Smalltalk and C++ were all invented and popularized outside of academia. Only the LISP variants of OO could be described as coming from universities and even there the case isn't strong. All the design/analysis methods came from industry practitioners - Booch, Coad/Yourdon, Schlaer-Mellor, Rumbaugh - all were working in the trenches. In the early/mid '80's nobody at university told me anything about OO but immediately on leaving I found myself having to learn it fast! Just where does the alleged university influence come from?! Alan G.
From: sdm7g@elvis.med.Virginia.EDU (Steven D. Majewski) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 3 Jul 1998 16:03:15 GMT Organization: University of Virginia Message-ID: <6nivc3$ret$1@murdoch.acc.Virginia.EDU> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> In article <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net>, Judson McClendon <judmc123@bellsouth.net> wrote: >Scott Hess wrote: >> >> It wasn't like some egghead invented object oriented programming and >> design and then everyone in the world decided that we should try to >> cram every project into that mold. > >Actually, that's a pretty good description of what DID happen. Had OO >been invented by the folks in the trenches, not only would it have been >far easier to learn and use than it is, but it would be time and cost >effective, which it has certainly not proven to be, in real world >practice. The movement for OO came from the universities, not the field. >Any time you see all the universities trumpeting for something which has >never been proven practical in practice, you can *know* it comes from >academia. :-) >-- You've got it exactly wrong. OO was invented by the folks in the trenches. I don't know if Simula was developed in a university or not, but it was certainly developed to solve practical problems. Smalltalk was developed at Xerox -- a research group, but not a university. Real Alan Kay's HOPL History of Smalltalk paper to see how practical his concerns were. His inspiration came from very practical concerns, and things like the indirect jump tables in some device driver code he had fixed. C++ was deveopled at AT&T - like Unix and C more as a tool than as a theoretical research project. ( If it had been designed to be less practical, it would have been a cleaner language! ) If you read Brad Cox's book on Objective C, his orientation was also very practical and problem oriented. His concern was that programming had to become more like hardware design, with reusable "black-box" components before programming could advance. Before C and C++ were solidly entrenched, Apple, Borland and others developed their own O-O extensions to Pascal. (lack of any standardization here is what left room for the growth of C++) They were not interested in any theoretical problems, but they found that the O-O paridigm was a simper way to program GUI systems than procedural event loop programming. This, more than any academic research is what prompted the growth of OO programming. Mostly, with a few exceptions, the academic work followedthe tool builders. Nicholas Wirth and some of the folks that worked with him are one of the exceptions. I think he had a role in Apple's Object Pascal. The Scheme community was in contact with Alan Kay's Smalltalk group, and both groups were aware of the fact that closures+continuations were orthogonal but functionally equivalent to Objects. In fact, if you compare Scheme and Smalltalk, you'll see a good example of the way academics (Scheme) looked at the problem and the less theoretical, more practical view of OO. ---| Steven D. Majewski (804-982-0831) <sdm7g@Virginia.EDU> |--- ---| Department of Molecular Physiology and Biological Physics |--- ---| University of Virginia Health Sciences Center |--- ---| P.O. Box 10011 Charlottesville, VA 22906-0011 |--- "I'm not as big a fool as I used to be, I'm a smaller fool." - Jack Kerouac Some of the Dharma <http://members.aol.com/kerouacsis/SomeDharma.html>
From: Kris Sheglova <95155580@brookes.ac.uk> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Fri, 03 Jul 1998 12:01:48 +0100 Organization: Disorganised Message-ID: <359CBA1C.4E9F@brookes.ac.uk> References: <6ng34a$j2g$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit boracay@hotmail.com wrote: > 3)If only a computer scientist can program it or understand all the > complexities surrounding development with it, and not your average >scientist, > engineer or other individual with basic computer training, then it is >too > complicated and is NOT an advancement. Programming by its very nature *is* complicated. It is easy to make mistakes - this is not only a function of the programming languages used but by the structure that a program must take to work. Many systems are complex, not linear. These systems need to be programmed efficiently and reliably. This is why computer science should be a specialised field and not left to people with basic computer training. Would argue: "Well electric wiring is too complicated. We must make it simpler so that every one can do it " at the cost of people risking their lives and/or property for no reason. Your average jo bloggs can learn to program a computer using simple systems e.g. BASIC if they need to knock up a program or two every so often but these should not become programs relied on by the masses. Kris Sheglova 95155580@brookes.ac.uk
Message-ID: <359C25D4.7B2A@computer.org> Date: Thu, 02 Jul 1998 18:29:08 -0600 From: "B. Thomas" <b.thomas@computer.org> MIME-Version: 1.0 Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359BB163.7A97DEC6@gssec.bt.co.uk> <6nh07l$hl$1@nnrp1.dejanews.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Organization: IBM.NET hmmueller@my-dejanews.com wrote: > > Thanks for saying that! I dont know why, but there is still this idea around > that computers should be programmed by laymen - electrical engineers, > astronomers, mechanical engineers etc. Nobody would allow me to design > the smallest pedestrians' bridge over a small river - but somehow people (and > companies!) believe that everybody who USED a computer should program it. > > I've seen many non-computer engineering engineers doing program - and I know > now that they usually can't do it - neither the distinction between a good > and a bad hash function nor between a good and a bad UI nor between a good > and a bad test case or test plan can usually be learned "on the job". You > need some background. > > - ok, that was harsh, maybe. So let me add: Everybody can learn new things, > also computer programming. However, it will take a few years; and if you do > it on your own, you'll like not exceed mediocre status. Since when was computer science taught in universities? Most of the founding fathers of computer science came from other fields - mathematics, electrical engineering etc. Do you imply that they were mediocre too, since they were also self-taught (though most of that self-teaching process was part of research). Biju Thomas
From: sdm7g@elvis.med.Virginia.EDU (Steven D. Majewski) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 3 Jul 1998 00:39:14 GMT Organization: University of Virginia Message-ID: <6nh97i$86v$1@murdoch.acc.Virginia.EDU> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6ng9em$h77@tyr.diku.dk> <6nglhp$c5o$1@news.cs.tu-berlin.de> In article <6nglhp$c5o$1@news.cs.tu-berlin.de>, Marcel Weiher <marcel@cs.tu-berlin.de> wrote: > >>At the moment, this is hard to do. > >Understatement... :-) > >-- > >Java and C++ make you think that the new ideas are like the old ones. >Java is the most distressing thing to hit computing since MS-DOS. > - Alan Kay - Question: How did God create the world in 7 days ? Answer: He didn't have compatibility with an installed base to worry about! ---| Steven D. Majewski (804-982-0831) <sdm7g@Virginia.EDU> |--- ---| Department of Molecular Physiology and Biological Physics |--- ---| University of Virginia Health Sciences Center |--- ---| P.O. Box 10011 Charlottesville, VA 22906-0011 |--- "I'm not as big a fool as I used to be, I'm a smaller fool." - Jack Kerouac Some of the Dharma <http://members.aol.com/kerouacsis/SomeDharma.html>
From: JRStern@gte.net (JRStern) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Sat, 04 Jul 1998 00:39:49 GMT Organization: gte.net Message-ID: <6njtlt$gk2$2@news-1.news.gte.net> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6ng5ku$nub$1@supernews.com> On 2 Jul 1998 14:31:58 GMT, jcr.remove@this.phrase.idiom.com (John C. Randolph) wrote: >Yeah, well: most tools suck, and so do many developers. A more concise summary of the dark side, I've never heard! Joshua Stern JRStern@gte.net , then we will never get many good >geology application programs. One of them has got to become easier. Sez who? >I expect the number of facts and techniques needed to master both >geology and computer science to grow. However, if computer scientists >do their job, programming should get easier and simpler. And in fact, >it you hold constant the complexity of the job you're trying to >accomplish, and you avoid poorly designed systems ( like C++ and MFC ), >programming has in fact gotten easier. There's worse things than C++ and MFC -- for example, writing Windows apps in C and SDK. However, your point is good, and let me make it more explicit: we tend to want more out of our apps today, our expectations are rising at least as fast as the technology is improving. Joshua Stern JRStern@gte.net
From: eiffelpgmr@nni.com (Tom Morrisette) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Fri, 03 Jul 1998 17:48:58 GMT Message-ID: <359d16db.3073435@news.nni.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> On Fri, 03 Jul 1998 13:08:52 GMT, "Judson McClendon" <judmc123@bellsouth.net> wrote: >Scott Hess wrote: >> <SNIP> > Had OO >been invented by the folks in the trenches, not only would it have been >far easier to learn and use than it is, but it would be time and cost >effective, which it has certainly not proven to be, in real world >practice. The movement for OO came from the universities, not the field. >Any time you see all the universities trumpeting for something which has >never been proven practical in practice, you can *know* it comes from >academia. :-) >-- >Judson McClendon This is a faithful saying and worthy of all >Sun Valley Systems acceptance, that Christ Jesus came into the >judmc123@bellsouth.net world to save sinners (1 Timothy 1:15) >(please remove numbers from email id to respond) > > > I'd say that OO is hard to learn for two reasons: 1. Many people (present company excluded, of course) think C++ syntax is OOP, where C++ is actually the logical culmination of these driving design criteria. (A) C++ is as compatible with C as possible. (B) No C++ program pays any price (in size or efficiency) for any C++ feature that it doesn't use. It acheives these wonderfully, but the result isn't optimized for learning or straightforwardness. 2. Those of us who first learned in the procedural world have a lot of trouble changing our mind set. The years of practice distorted our concepts of "natural" and "intuitive". People who learn OO first, with a simpler language, find that it isn't so hard to learn. Imho, they can take those concepts into the C++ world, and with more difficulty, into FORTRAN and COBOL worlds.
From: PC Subject: Brand New 200MHZ PC's for $599. Includes Monitor. Limited Supply! Message-ID: <InQpTwxp9GA.163@cidintnews.infosel.com.mx> Newsgroups: comp.sys.next.programmer Date: Sat, 04 Jul 1998 02:19:26 -0500 Brand New 200MHZ PC's for $599. Includes Monitor. Limited Supply! Check it out at http://www.pc4lessonline.com/specialoffer
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <InQpTwxp9GA.163@cidintnews.infosel.com.mx> Control: cancel <InQpTwxp9GA.163@cidintnews.infosel.com.mx> Date: 04 Jul 1998 08:18:22 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.InQpTwxp9GA.163@cidintnews.infosel.com.mx> Sender: PC Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Arnaud <debayeux@easynet.fr> Newsgroups: comp.sys.next.programmer Subject: NS 3.2 or NS 3.3 libraries Date: Sat, 04 Jul 1998 11:49:31 +0200 Organization: [posted via] Easynet France Message-ID: <359DFAAB.6155AD3@easynet.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello, I have an OS 4.2 dev that I would like to complete with some static libraries from precedent versions. I would like (I guess the names are correct) : - 3Dkit (libMedia_s.a + include files) - Interceptor (libInterceptor_s.a + include files) (it has been documented for Rhapsody) - DriverKit full example source code Could anybody tell me if it is in NS 3.2, or NS 3.3 ? What are the differences between NS 3.2 dev and NS 3.3 dev ? Better than that, where could I find a content of the 3.2 and 3.3 developer CDs (search failed at NeXTAnswers) ? Or could someone send me a listing ? With many thanks, Arnaud
From: "Judson McClendon" <judmc123@bellsouth.net> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> Subject: Re: Programming and the colossal failure to advance Message-ID: <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> Date: Sat, 04 Jul 1998 16:16:42 GMT NNTP-Posting-Date: Sat, 04 Jul 1998 12:16:42 EST Judson McClendon wrote in message ... >Scott Hess wrote: >> It wasn't like some egghead invented object oriented programming and >> design and then everyone in the world decided that we should try to >> cram every project into that mold. > >Actually, that's a pretty good description of what DID happen. Had OO >been invented by the folks in the trenches, not only would it have been >far easier to learn and use than it is, but it would be time and cost >effective, which it has certainly not proven to be, in real world >practice. The movement for OO came from the universities, not the field. >Any time you see all the universities trumpeting for something which has >never been proven practical in practice, you can *know* it comes from >academia. :-) Since there were a number of replies to this post, I'll answer them all here. First, I wrote my post hastily and did not clearly articulate what I meant to say. I used the word 'university', when I should have said 'academia', meaning the broader base of research and theoretical folks, such as Xerox PARC, not just universities. Sorry. Here is an example of how C became popularized. In the 70's, DEC was seriously dominant in the university environment. Because Bell Labs had written C for the PDP, most of the system software for those computers was in C. Students who wanted to seriously get into the system software simply had to master C. As those students graduated and moved into industry, they knew C, and stumped for C. This is where the main impetus for C in the industry came from. Why else do you think companies like DEC spend so much money assisting universities? The fact that C was developed at a company became almost irrelevant, because it had saturated the universities. To say that a language developed by academia would be better, because they would prefer something elegant, is amusing. Nothing wrong with elegance, but unless it works in the real world, on real world applications, the point is irrelevant. Most academic folks have little or no experience in writing large, real world applications, and in maintaining them over time. You cannot learn to play a musical instrument, or ride a motorcycle, or program a computer by reading about it in a book. You must actually do it for a long time to learn to do it well. Not to say that study is of no or little benefit, only that it is insufficient of itself. It is one thing to read in a book, or be told by an instructor, that such and such is or is not, important. But it is like being burned by fire. You can be told a million times that fire will burn you, but when you stick your hand into fire for the first time, your knowledge suddenly takes on a whole new dimension! And the lasting burn gives you a respect, and a 'feel' for what a burn is, that you could *never* learn any other way. This experiential knowledge is essential in making 'judgment' calls and making decisions. In my opinion, the most serious problem we have today with development environments is too much complexity. I say that because the cost of new systems development and maintenance has skyrocketed. About 45% of new systems started are not completed because of cost or time overruns, and inability to meet design goals. Back in the early days of computing, hardware was the expensive item, and software was relatively cheap. But over the years, this has flip-flopped dramatically, and development costs continue to rise at an alarming rate. But more importantly, as the pace of change in business, and society in general, quickens, the life span of new systems dwindles. At the same time, the time it takes to develop new applications is increasing. When the time to develop a new application becomes as great or greater than the life span of the application, then you are in a "can't get there from here" scenario. Many of the failed systems development efforts today die from just this problem. It was clear by 1980 that this problem was becoming critical. One reason development time is increasing is that applications are becoming more and more complex. Some of this increased complexity is unavoidable. However, much of this added complexity is created by the move to 'gee-whiz' features like GUI, which may offer absolutely no benefit whatever, in certain applications. At the same time, software development tools and environments are becoming more and more complex, for similar reasons. On the shelves over my desk are my software reference manuals. There are about 2 feet (.6 m) of manuals for VC++, and about 1 foot (.3 m) of Windows reference manuals, plus a host of other manuals for VB, Office 97, Java, etc. That represents a truly *enormous* amount of information to assimilate and deal with on a daily basis. Part of the reason for the complexity is the Windows API, MFC, and C++. It is an undeniable fact that when people must deal with more complex environments, they will make more mistakes. A programming language environment should be *only* as complex as necessary, and definitely as simple as possible. It is fairly easy to tell when a new method is effective. Simply put it into practice and observe the results. As the saying goes "The proof of the pudding is in the eating." Hard, cold, objective results show that the result of these new development environments, including OO, are greatly increased costs and development time, and less reliability. After all, we must not forget that complexity increases geometrically. Unfortunately, some people seem attracted to these methodologies with little or no regard to the actual results from the field. Recently I encountered two prominent articles, one on the front page of the Wall Street Journal (April 30, 1998, "Pulling the Plug"), and one on the front page of Info World (June 29, 1998, "Return of the Mainframe"). The people on the board of large companies are interested in the bottom line, people. If these new technologies were delivering as they have been touted, companies *would not* be backing off their use. What kind of fool would consider abandoning technologies that were delivering the goods? Not the people who make seven figures sitting on the boards of large corporations, of that you can be sure. They just want what works. The problem, people, is that we need to make programming easier and simpler, not harder. If OO was easier, or more economical, we would have seen clear evidence by now in the numbers. What we do see is clear evidence to the contrary. If we want economical, reliable software, we will do this. We can't make people smarter. If we want more programmer productivity, and better reliability, we had better make our tools easier to use. We will never do it by ignoring the evidence that is clearly at hand. And we had better learn to apply complex technology where it is needed or useful, not spray it across everything, whether it is warranted or not. We are free to ignore the evidence. But we will have to deal with the consequences, and we are. -- Judson McClendon This is a faithful saying and worthy of all Sun Valley Systems acceptance, that Christ Jesus came into the judmc123@bellsouth.net world to save sinners (1 Timothy 1:15) (please remove numbers from email id to respond)
From: William Chesters <williamc@toy.william.bogus> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 03 Jul 1998 23:22:18 +0100 Organization: Edinburgh University Message-ID: <m2btr64lt1.fsf@toy.william.bogus> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> "Judson McClendon" <judmc123@bellsouth.net> writes: > Had OO been invented by the folks in the trenches, not only would it > have been far easier to learn and use than it is, but it would be > time and cost effective, which it has certainly not proven to be, in > real world practice. You must be joking. OO as we know it had practically no input from universities. In fact: Had C++ been invented by a proper egghead, rather than by someone with a frankly partial grip on CS, it would have been far cleaner and less difficult to learn and use. That's because university researchers value elegance in their languages above anything else---that combination of simplicity and expressiveness which makes them much less fiddly and intimidating than, say, C++.
From: james@unspam.edu (James McCartney) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Sat, 04 Jul 1998 12:26:41 -0700 Organization: The University of Texas at Austin, Austin, Texas Message-ID: <james-ya023180000407981226410001@news> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net>, "Judson McClendon" <judmc123@bellsouth.net> wrote: >the pudding is in the eating." Hard, cold, objective results show that >the result of these new development environments, including OO, are >greatly increased costs and development time, and less reliability. What results? Trials I've seen posted using good OO environments such as Smalltalk show reduced costs and development times. C++ is likely not the best test of the benefits of OO. (In fact a friend told me of his team of a few Smalltalk programmers who were so fast at getting their own work done that the larger team of C++ programmers weren't ready with their component to test on time so the Smalltalk team mocked up that part as well for testing and it worked so well that the C++ programmers were all let go.) >simpler, not harder. If OO was easier, or more economical, we would have >seen clear evidence by now in the numbers. What we do see is clear >evidence to the contrary. If we want economical, reliable software, we So you propose that OO makes things too complex and we should go back to procedural programming? If so, you're a nut case. OO is a way to deal with complexity. Perhaps OO IS a mistake and we should all be using FP, however the efficiency and expressability problems there have not yet been fully solved. (And if you thought that the paradigm shift to OO was hard for many programmers its nothing compared to the shift to FP.. but I digress..) But OO is certainly a reduction in complexity of what went before it. It is just now that we have OO we try to express much more complex problems because we see that we can. --- James McCartney
From: Larry Blische <larry@lkba.com> Newsgroups: comp.sys.next.programmer Subject: I need help building LazyScroll with the MiscTableScroll palette Date: Sat, 04 Jul 1998 16:16:45 +0000 Organization: LKB Associates, Inc. Message-ID: <359E475D.D6142BBA@lkba.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I'm trying to build the examples that come with the MiscTableScroll Palette which I've just picked up off of peak. I'm running NEXTSTEP 3.3 with the 3.2 Dev release. I don't have any more of the MiscKit (which I'll probably pick up too) but the Readme says I don't need it. When I build the LazyScroll example I get this error: cc -g -O -Wall -I./sym -arch m68k -I. -c DirArray.m -o ./m68k_obj/DirArray.o In file included from DirArray.m:39: /NextDeveloper/Headers/bsd/sys/types.h:88: redefinition of `u_long' /NextDeveloper/Headers/sys/types.h:88: `u_long' previously declared here /NextDeveloper/Headers/bsd/sys/types.h:118: redefinition of `dev_t' /NextDeveloper/Headers/sys/types.h:118: `dev_t' previously declared here /NextDeveloper/Headers/bsd/sys/types.h:119: redefinition of `gid_t' /NextDeveloper/Headers/sys/types.h:119: `gid_t' previously declared here /NextDeveloper/Headers/bsd/sys/types.h:120: redefinition of `uid_t' /NextDeveloper/Headers/sys/types.h:120: `uid_t' previously declared here /NextDeveloper/Headers/bsd/sys/types.h:122: redefinition of `ino_t' /NextDeveloper/Headers/sys/types.h:122: `ino_t' previously declared here /NextDeveloper/Headers/bsd/sys/types.h:131: redefinition of `off_t' /NextDeveloper/Headers/sys/types.h:131: `off_t' previously declared here /NextDeveloper/Headers/bsd/sys/stat.h:69: redefinition of `struct stat' *** Exit 1 Stop. *** Exit 1 Stop. Can anyone tell me what I'm doing wrong? Why would the compiler be including two types.h sources? Thanks, -- Larry Blische * Consultant/Programmer Desktop Apps : Client/Server : Embedded Systems : Device Drivers : Etc. 6195 Eagles Nest Drive * Jupiter, FL 33458 USA * 561.747.7844 mailto:larry@lkba.com * resume at http://www.charm.net/~lkb
From: Charles Hixson <charleshixsn@earthlink.net> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Sat, 04 Jul 1998 13:38:13 -0700 Organization: Mandala Fluteworks Message-ID: <359E92B5.5E4A5A33@earthlink.net> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359BB163.7A97DEC6@gssec.bt.co.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Actually, I think that programming is, in fact, getting harder. This is because the simple things that are worth doing tend to already be done. I don't currently see and reason to design a spread-sheet, it's been done. The current models even have tail-fins. Alan Gauld wrote: > > boracay@hotmail.com wrote: > > Today an individual seeking to develop an application or have a computer fill > > a need which involves a programmed algorithm faces a most difficult task. > > Not so. if he doesn't mind a teletype interface then its no > more difficult(indeed much easier!) than it was 10 years ago > - and light years easier than it was 20 or 30 years ago. > - No punched cards, no single line paper teletypes, no > line oriented editors etc etc. > > Even if it requires a GUI environments like VB and > Delphi enable even fairly complex poroblems to be > solved with realative ease 0- much more than I could > do on my Commodore PET in the early '80s. > > > Years ago programming was difficult but not extremely so. > > Years ago, people worked mainly in assembler - in the very > early ywars they typed in binary(or octal usually) codes > by hand - not even having assemblers! On PCs this was still > the norm into the mid '80's. > > > Programming today is little more than back > > breaking mind concentrating labor. > > I would have agreed with this between 1990 and 1995 but > the tools available in recent years have changed my mind > - its now only as hard as it was 20 years ago.... > > > longer, is harder and requires more attention than it ever has. > > Simply not true - see above. > > > being used to create the same functionality than ever before. > > Not true - the functionality is orders of magnitude greater. > Users expectations have risen dramatically. The core functions > may not have changed but the peripheral functionality is > enormous - compare MS Word 97 with Wordstar V3 say. > > The output looks the same the functionality in the programs > is worlds apart. > > > While the GUI made using the computer easy, and a web > > browser made communicating with a computer easy, > > there has been little or no progress in making programming easy > > I doubt it will ever be *easy*. We are persuading a machine > which can add 1+1 and copy data from A to B to do all manner > of tasks. Programming environmentsd hide that fact from us but > ultimately codifying human actions will allways be hard. > It can become easier but never easy! IMO. > > > 1)We should not need to understand details of the OS or VM design > > The bulk of MIS programs writtenm today in VB fit this criteria. > > > 2)We should not be forced into understanding abstract objects > > Why not? If that what makes programming easier? > An average person can build a bookcase from a kit but still > needs to understand how screwdrivers and screws interact. > The issue is with the complexity of the models not the > need to use a model. > > > 3)If only a computer scientist can program it or understand all the > > complexities surrounding development with it, and not your average > > scientist, engineer > > The average computer scientist can't design computer hardware, > why would the average electrical engineer be expected to > design a modern software application? Far less someone whose > field is in a radically different area like genetics. You seem, > to assume that computers shouyld somehow be easy - why? > What they do is intrinsically complex. > > If it were easy, true Turing machines would be widely available. > > Alan G.
From: jonas.v.andersson@seinf.mail.abb.com (Jonas Andersson) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Fri, 03 Jul 1998 03:55:56 GMT Organization: ABB Business Systems AB Message-ID: <359c550c.83892561@138.221.200.100> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6ng8c3$vb52@onews.collins.rockwell.com> <6nh8pk$82k$1@murdoch.acc.Virginia.EDU> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit On 3 Jul 1998 00:31:48 GMT, sdm7g@elvis.med.Virginia.EDU (Steven D. Majewski) wrote: >In article <6ng8c3$vb52@onews.collins.rockwell.com>, >Erik M. Buck <embuck@palmer.cca.rockwell.com> wrote: >>This is just silly. >> >>A computer scientist does not expect to understand geology, genetics, >>surgery, particle physics, systems engineering, or even ditch digging to any >>greater degree than the average population and certainly not to any degree of >>professional skill. Why would a geologist, surgeon, physicist, systems >>engineer, or ditch digger expect to know enough about computer science to >>exhibit professional skill in computer science ? >> >>People are tantalized by how easy and powerful programming seems at first. >>It is not a problem that computers are too hard; it is that they are >>approachable. Almost anyone can learn intro programming. Why would that >>person expect to be any more capable as a computer scientist than a computer >>scientist would be at geology after an intro geology course. >> >>It is condescending in the extreme to claim the computer science should be >>easy enough for anyone. This is as large and hard a proffesion as any other >>and people do not expect geology to be instantly understood by every novice >>that comes along. >> > >But there's a difference between computer science and programming. >One of the goals of computer science is to make programming easier. I wonder. Computer science makes some strange things usually not very related to the "reality" or production environment. I would say that corporations that makes the tools have to make programming easier. > >If it is exactly as difficult to teach programming to geologists >or teach geology to programmers, then we will never get many good >geology application programs. One of them has got to become easier. Yes. The teaching of geology to the programmers will probably always be easier. This is because you dont have to teach the whole field, just the subset that will be interesting for the application. > >I expect the number of facts and techniques needed to master both >geology and computer science to grow. However, if computer scientists >do their job, programming should get easier and simpler. And in fact, >it you hold constant the complexity of the job you're trying to >accomplish, and you avoid poorly designed systems ( like C++ and MFC ), >programming has in fact gotten easier. I think that C++ is a pretty good design but MFC isn't. The only major thing lacking in C++ is garbage collection, which probably can be bought. By the way I wonder if it is relevant to discuss language anymore, I think it is more relevant to discuss function-libraries. > > >---| Steven D. Majewski (804-982-0831) <sdm7g@Virginia.EDU> |--- >---| Department of Molecular Physiology and Biological Physics |--- >---| University of Virginia Health Sciences Center |--- >---| P.O. Box 10011 Charlottesville, VA 22906-0011 |--- > >"I'm not as big a fool as I used to be, I'm a smaller fool." - Jack Kerouac >Some of the Dharma <http://members.aol.com/kerouacsis/SomeDharma.html>
From: jonas.v.andersson@seinf.mail.abb.com (Jonas Andersson) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Fri, 03 Jul 1998 04:06:41 GMT Organization: ABB Business Systems AB Message-ID: <359c5749.84465264@138.221.200.100> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359BB163.7A97DEC6@gssec.bt.co.uk> <6nh07l$hl$1@nnrp1.dejanews.com> <359C25D4.7B2A@computer.org> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit On Thu, 02 Jul 1998 18:29:08 -0600, "B. Thomas" <b.thomas@computer.org> wrote: >hmmueller@my-dejanews.com wrote: > >> >> Thanks for saying that! I dont know why, but there is still this idea around >> that computers should be programmed by laymen - electrical engineers, >> astronomers, mechanical engineers etc. Nobody would allow me to design >> the smallest pedestrians' bridge over a small river - but somehow people (and >> companies!) believe that everybody who USED a computer should program it. >> >> I've seen many non-computer engineering engineers doing program - and I know >> now that they usually can't do it - neither the distinction between a good >> and a bad hash function nor between a good and a bad UI nor between a good >> and a bad test case or test plan can usually be learned "on the job". You >> need some background. Who writes hash-functions anymore? But I agree programming is easy at the beginning, >> >> - ok, that was harsh, maybe. So let me add: Everybody can learn new things, >> also computer programming. However, it will take a few years; and if you do >> it on your own, you'll like not exceed mediocre status. If you spend a few years you will probably exceed mediocre status (if you are fascinated about a subject enough to spend a couple of years you usually will). It takes about 10 years to be a master of any field (with some size). On the other hand, I havent seen anybody from university that writes good code at the start, there are just to many things not taught at the university. > >Since when was computer science taught in universities? > >Most of the founding fathers of computer science came from other fields >- mathematics, electrical engineering etc. Do you imply that they were >mediocre too, since they were also self-taught (though most of that >self-teaching process was part of research). And most program code that I have seen come out of computer science is still: a. Not written for understandability. b. Poorly documented. > >Biju Thomas
From: hmmueller@my-dejanews.com Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Sun, 05 Jul 1998 00:20:05 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6nmgrl$tfp$1@nnrp1.dejanews.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359BB163.7A97DEC6@gssec.bt.co.uk> <6nh07l$hl$1@nnrp1.dejanews.com> <359C25D4.7B2A@computer.org> In article <359C25D4.7B2A@computer.org>, b.thomas@computer.org wrote: > > hmmueller@my-dejanews.com wrote: > > > > > Thanks for saying that! I dont know why, but there is still this idea around > > that computers should be programmed by laymen - electrical engineers, > > astronomers, mechanical engineers etc. Nobody would allow me to design [...] > > Most of the founding fathers of computer science came from other fields > - mathematics, electrical engineering etc. Do you imply that they were > mediocre too, since they were also self-taught (though most of that > self-teaching process was part of research). No, I dont. My remark is a general one; this includes that there may be many intelligent non-formally-CS-trained people who are great with CS and computers and programming. If I say that physicians' knowledge of music shouldn't be the guideline for professional music, I still accept that Albert Schweitzer (a physician) was one of the greatest organists and Bach-specialists around. What I say is exactly what I say: That programming is wrongly assumed to be something everybody (or at least every engineer) should be able to do, whereas this does not hold for other engineering disciplines. Re the argument that it is easier to train a biologist in programming than a programmer in biology: In the one (university) biology lab where I had friends, they had a lab technician. If they needed a new cage; or a special machine to to do an experiment with chimpanzees (3 switches, a few food outlets and two lamps, in one case), it was clear that the design and building of the machine was NOT the biologist's job; he would put a sketch on paper and talk with the technician who would then design and build the machine; in the case of the cage, he would usually only do the design, and the cage would be built by some welding company or so. All that is obvious. So why was it assumed that the same biologist would write the program for controlling the lamps and food outlets and measuring the timing of the chimpanzees' responses himself (where he failed miserably)? Or that me (the programmer coming in afterwards) would have to understand biology to write a small real-time program? Neither has the biologist to learn programming, nor the programmer biology. They just have to talk to each other. It works; and is almost always much cheaper, and more more certainly gets the job done. And again: If someone has all the required expertise, from biology to programming to turning some part on a lathe, let him/her do it. But, please, dont assume that it is the fault of CS if he/she cant. It is simply unreasonable to require programming to be simple. Harald M. Mueller -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: gio+van+no+ni+8@tal+star+spam.com (Giovanni 8) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 5 Jul 1998 02:38:16 GMT Organization: TalStar Communications Message-ID: <gio+van+no+ni+8-0407982248370001@dialup12.tlh.talstar.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <359d16db.3073435@news.nni.com> Cache-Post-Path: flux!unknown@dialup12.tlh.talstar.com > Tom Morrisette wrote: >> On Fri, 1998-07-03 13:08:52 GMT, Judson McClendon wrote: >> Had OO been invented by the folks in the trenches, not only >> would it have been far easier to learn and use than it is, >> but it would be time and cost effective, which it has >> certainly not proven to be, in real world practice. The >> movement for OO came from the universities, not the field. >> Any time you see all the universities trumpeting for >> something which has never been proven practical in practice, >> you can *know* it comes from academia. :-) There's a big difference between object-oriented analysis or design, and programming. I was 1st introduced to OOA and OO Data Modelling in the corporate world in the early 1980s. The terminology used then certainly sounded "foreign" -- e.g. LOTs (lexical object types) & NOLOTS (non-lexical object types) -- and weirdly more abstract than seemed warranted. From what I can tell in my readings, most of the development was done in Scandihoovia during the 1960s & 1970s, before folks in the USA were aware of it, for the most part. There are some things it lends itself very well to. One that comes to mind is CAD/CAM in which you are literally modelling (what will be) physical objects, and can use the objects constructively to produce assemblies, etc. Most applications are more abstract, and thus lend themselves to different POVs, and resultant differences in design. Experience should give one the chance to learn which approaches will work best in the long run. > I'd say that OO is hard to learn for two reasons: > > 1. Many people (present company excluded, of course) think > C++ syntax is OOP, where C++ is actually the logical > culmination of these driving design criteria... It > acheives these wonderfully, but the result isn't > optimized for learning or straightforwardness. > > 2. Those of us who first learned in the procedural world > have a lot of trouble changing our mind set. The years of > practice distorted our concepts of "natural" and "intuitive". > > People who learn OO first, with a simpler language, find that > it isn't so hard to learn. Imho, they can take those concepts > into the C++ world, and with more difficulty, into FORTRAN and > [Yuck!!!] worlds. Most of C++ is simple. What complications there are come in with SOME of the OOP features such as class & template declarations (no minor matter). Another general problem is the proliferation & development of the attendant libraries (a problem I wouldn't have thought of as a problem 20 years ago). It seems, reading MS books, that they introduce another library with another couple acronyms, every few pages, none of them differentiating anything especially significant, but creating another "product" that must be purchased in order to put together a useful whole. I think the biggest problem is psychological, a hesitance because of fear that one may not be able to completely shift to the new paradigm, rather than any actual inability to do so. "In our experience, it takes only a few weeks for a professional developer to master the syntax & semantics of a new programming language. It may take several more weeks for the same developer to begin to appreciate the importance & power of classes & objects. Finally, it may take as many as 6 months of experience for that developer to mature into a competent class designer." --- Grady Booch 1991 _Object-Oriented Design with Applications_ pg 218 "[A]s we scale up to larger & larger applications, we soon reach a point where the lack of a module structure makes our design totally incomprehensible. Without a means of collecting abstractions into increasingly larger chunks, we soon reach the limits of the human ability to comprehend many things at once." --- Grady Booch 1991 _Object-Oriented Design with Applications_ pg 305 "If your language is confused, your intellect, if not your whole character, will almost certainly correspond." --- Arthur Quiller-Couch (quoted in Richard Lederer 1991 _The Miracle of Language_ pg 232) -- copyright 1998 by g8 all rights reserved, except that license is given freely to respond on usenet. other licenses are available at reasonable rates.
From: gio+van+no+ni+8@tal+star+spam.com (Giovanni 8) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 5 Jul 1998 03:06:17 GMT Organization: TalStar Communications Message-ID: <gio+van+no+ni+8-0407982316390001@dialup12.tlh.talstar.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359BB163.7A97DEC6@gssec.bt.co.uk> <359E92B5.5E4A5A33@earthlink.net> Cache-Post-Path: flux!unknown@dialup12.tlh.talstar.com > Charles Hixson wrote: >> Alan Gauld wrote: >>> boracay@hotmail.com wrote: >> I would have agreed with this between 1990 and 1995 but >> the tools available in recent years have changed my mind >> - its now only as hard as it was 20 years ago... > Actually, I think that programming is, in fact, getting harder. > This is because the simple things that are worth doing tend to > already be done. I don't currently see and reason to design a > spread-sheet, it's been done. The current models even have > tail-fins. Yah, I see a lot more organizations buying off the shelf and then adjusting how they do things to conform to what's available & easily customizable, rather than doing a lot of programming in order to customize. Let's see: spread-sheets, accounting, DBMS, "business graphics" & statistics are all things that we were generating custom code for back in the 1970s but that exist in shrink- wrap today. A lot of what I see being called "application development" now is really use of an application, e.g. a DBMS. So, in a sense, the programmer today is doing what the computer user of 25 years ago was doing. But, the main point I want to make is that the fancy GUIs and IDEs are actually making development more difficult in some ways. I mean, how complex is binary? It's just a few 1s and 0s, after all. You just line them up in the right clumps in the right order, and you've got a program. Now, some of the IDEs & code generators get in the way. A simple SQL query with a couple inner joins, gets turned into a nested mess by some of the "tools". And they hide things from you that make it more difficult to debug. Ideally, they'd reveal all, though in a conveniently organized manner, with e.g. forms, form elements, DB tables, code and all the attributes related to each all nicely inter- related graphically. Yes, we need abstraction, conceptualizing complex things in order to make them managable, but one must be able to check one's premises, to see the underlying entities subsumed in the concepts. "The span of absolute judgment & the span of immediate memory impose severe limitations on the amount of information that we are able to receive, process & remember. By organizing the stimulus input simultaneously into several dimensions & successively into a sequence of chunks, we manage to break... this informational bottle-neck." --- George Miller 1956-03-?? "The Magical #7, + or - 2" _The Psychological Review_ vol 63 #12 pg 95 (quoted in Grady Booch 1991 _Object-Oriented Design with Applications_ pg 17) -- copyright 1998 by g8 all rights reserved, except that license is given freely to respond on usenet. other licenses are available at reasonable rates.
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <3088899006434@digifix.com> Date: 5 Jul 1998 03:48:51 GMT Organization: Digital Fix Development Message-ID: <668899611234@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1993. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: gio+van+no+ni+8@tal+star+spam.com (Giovanni 8) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 5 Jul 1998 03:26:28 GMT Organization: TalStar Communications Message-ID: <gio+van+no+ni+8-0407982336510001@dialup88.tlh.talstar.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6ng8c3$vb52@onews.collins.rockwell.com> <6nh8pk$82k$1@murdoch.acc.Virginia.EDU> <359c550c.83892561@138.221.200.100> Cache-Post-Path: flux!unknown@dialup88.tlh.talstar.com >> On 1998-07-03 00:31:48 GMT Steven D. Majewski wrote: >>> Erik M. Buck wrote: >>> A computer scientist does not expect to understand geology, >>> genetics, surgery, particle physics, systems engineering, >>> or even ditch digging to any greater degree than the average >>> population and certainly not to any degree of professional >>> skill. Why would a geologist, surgeon, physicist, systems >>> engineer, or ditch digger expect to know enough about >>> computer science to exhibit professional skill in computer >>> science?... >> But there's a difference between computer science and >> programming... Exactly. A programmer does need to understand geology, if s/he's working with geologists; genetics if s/he's working with geneticists; surgery if s/he's working with surgeons; particles physics if s/he's working with particle physicists... & ditch digging if s/he's working with ditch diggers. Similarly, I would expect ditch diggers, mechanical engineers, etc., who are working with programmers to learn a little about programming. Now, in some cases, the programmer will delve deeply into the other field, and vice versa. I know chemists who have gotten deeply into programming and programmers who have gotten deeply into chemistry, electrical engineering, design of nuclear weapons systems, economics, architecture, telecommunications & automotive engineering. "Experience without theory teaches nothing. In fact, experience can not even be recorded unless there is some theory, however crude, that leads to an hypothesis & a system by which to catalog observations. Sometimes only a hunch, right or wrong, is sufficient theory to lead to useful observation." --- Clarence Irving Lewis & W. Edwards Deming (quoted in W. Edwards Deming _Out of the Crisis_) "One must never forget the importance of subject matter." --- W. Edwards Deming -- copyright 1998 by g8 all rights reserved, except that license is given freely to respond on usenet. other licenses are available at reasonable rates.
From: paipai@tin.it (Paolo Di Francesco) Newsgroups: comp.sys.next.programmer Subject: Re: Getting Started with Rhapsody Date: Sat, 27 Jun 1998 23:36:59 GMT Organization: Telecom Italia Net Message-ID: <35958171.1301902@news.tin.it> References: <andrewp-2706981427020001@ts5port13.port.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit On Sat, 27 Jun 1998 14:27:01 -0500, andrewp@interSPAMport.net (Andrew Pontious) wrote: > >I have a PowerMac 7200/75, souped up to 64 Mb. RAM and 4.0 Gb. HD. But the >601 chip still won't run Rhapsody anyway, right? So I'm going to have to >buy *some* new computer at *some* point. What question is *what* and >*when*. Oh, well, maybe you have to wait 1-2 years... > >The *what* is probably going to be a PowerMac G3, since I want to be able >to run MacOS X when it comes out. > Maybe G4... > >As yet another option, I could try to install Rhapsody in some sort of >dual-boot setup on my WinNT box at work. Does Rhapsody coexist well with >WinNT and fit comfortably on 1-2 gigs? > Yes, you can use dual boot and more than NT-Rhapsody solution. You can install anything from freebsd to linux + NT + DOS + Rhapsody... ;) And, yes the 2Gb partition is enough! ;) Visit www.rhapsody-it.com and enjoy!
From: dd@dandrake.com (Dan Drake) Newsgroups: comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming Subject: Re: Programming and the colossal failure to advance Date: 6 Jul 1998 22:27:53 GMT Message-ID: <vhIsdqY67dTD-pn2-YqJWEd2Zw85n@dynamic47.pm03.san-rafael.best.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 8bit On Thu, 2 Jul 1998 13:48:58, boracay@hotmail.com wrote: > Today an individual seeking to develop an application or have a computer fill > a need which involves a programmed algorithm faces a most difficult task. > Years ago programming was difficult but not extremely so. Programming > complexity has now reached a level of complexity that is far beyond this and > perhaps beyond where it should be.... > The average > individual may be able to define what they need their application or > algorithm to do, or even be able to pseudocode or code the framework of the > algorithm in matters of days. However programming into todays environments, > including the popular OS’s and virtual machines (VM), may take months or > years. > I have to disagree. Writing the equivalent of a program of 20-30 years ago is immensely, marvelously easier that it was then, but -- > While the GUI made using the computer easy, the GUI did indeed make writing the computer harder; also, of course, as soon as we mastered writing big, difficult 1970 programs, we took on much bigger and nastier problems. Here's how I learned the truth of the matter: I wrote a program in 1967 for an academic project, using a fast computer (CDC 6400), a modified Fortran IV, and so on. Hard work, took a long time to get working. About 25 years later, with too much time on my hands, I rewrote the program from scratch (really from scratch, with no copy of the old source and no notes) in C on a Sun workstation -- using a decent text editor, of course. After a modest amount of programming and testing, I looked up in surprise and said, "That's odd, I'm done already." (Being done in this case meant being ready to go on with the next step in research beyond my original project.) You have to do something like this to realize just how much difference the improvements in hardware and system software (and programming techniques, of course) have made. But of course, in real life no one would want a crummy old command-line program like that one. To make something minimally acceptable, your neat algorithms and minimal I/O code would be imbedded in 3 or 4 times as much GUI code -- which would be harder to write and harder to debug than the algorithms ever were. That's why it looks as if there has been no, or negative, progress in making programming more efficient. -- Dan Drake dd@dandrake.com http://www.dandrake.com/index.html
From: rwaltman@bellatlantic.net (Roberto Waltman) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Sun, 05 Jul 1998 13:29:29 GMT Organization: . Message-ID: <35a37df6.7660204@news5.bellatlantic.net> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6ng8c3$vb52@onews.collins.rockwell.com> <6nh8pk$82k$1@murdoch.acc.Virginia.EDU> <359c550c.83892561@138.221.200.100> <gio+van+no+ni+8-0407982336510001@dialup88.tlh.talstar.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit gio+van+no+ni+8@tal+star+spam.com (Giovanni 8) wrote: >A programmer does need to understand geology, if s/he's >working with geologists; genetics if s/he's working with >geneticists; surgery if s/he's working with surgeons; >particles physics if s/he's working with particle physicists... >& ditch digging if s/he's working with ditch diggers. >Similarly, I would expect ditch diggers, mechanical >engineers, etc., who are working with programmers to >learn a little about programming. Now, in some cases, >the programmer will delve deeply into the other field, >and vice versa. And that's a good thing ! I'm a programmer working today on your average industry environment. One of the things I badly miss from previous jobs in an academic environment, was the constant exposure to different fields of knowledge, together with the need to learn enough about each one as to be able to solve the programming tasks that were brought to me. This still occurs in my job today, but to a much lesser degree. Motion control yesterday, biomechanics today, image processing tomorrow, ultrasound scanning techniques last week... Those were the days ! RW.
From: dd@dandrake.com (Dan Drake) Newsgroups: comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming Subject: Re: Programming and the colossal failure to advance Date: 6 Jul 1998 23:13:39 GMT Message-ID: <vhIsdqY67dTD-pn2-dGSKqdjMrSfd@dynamic47.pm03.san-rafael.best.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 8bit On Sat, 4 Jul 1998 16:16:42, "Judson McClendon" <judmc123@bellsouth.net> wrote: > Judson McClendon wrote in message ... >... > > Here is an example of how C became popularized. In the 70's, DEC was > seriously dominant in the university environment. Because Bell Labs had > written C for the PDP, most of the system software for those computers > was in C. Students who wanted to seriously get into the system software > simply had to master C. As those students graduated and moved into > industry, they knew C, and stumped for C. This is where the main impetus > for C in the industry came from. Popularized schmopularized. I wrote (co-wrote) one of the first PC applications ever written in C, at the time that Lotus 1-2-3 was being done in assembler. Was that because I learned C in school? Not quite; they didn't teach C when I was in school: in 1967! By 1982 I had 15, count them, years of working experience in assembler, Fortran, and other things. (You know -- Cobol, Pascal, Lisp, Simula, and various things you've never heard of.) No C, except reading K&R and thinking at first that it looked awfully kludgy. The reason my colleagues and I decided on C was that no other available choice could possibly be adequate for the task (though PL/I, believe it or not, came close). So we learned C and used it. There were no freshly minted CS graduates in the company -- you'd laugh if you heard how many there were of any kind -- just programmers who needed to make a CAD program work and knew what they had to do. Somebody who was in school in the very early 80s (before which C had no real-life non-Unix presence whatever) could maybe tell us whether C was de rigueur then. My recollection is that Pascal owned the whole academic show except for the Unixoids (to whom praise and thanks). >... > To say that a language developed by academia would be better, because they > would prefer something elegant, is amusing. Nothing wrong with elegance, > but unless it works in the real world, on real world applications, the > point is irrelevant. Most academic folks have little or no experience in > writing large, real world applications, and in maintaining them over time. >... On this, I'll leave the dissent to someone else, since I mostly agree. My problem is just that C is a terrible example for your point; its strength was that it was _not_ developed for academic elegance or big-company marketing, but for the needs of some programmers who wanted to get a job done. To be sure, you could call Bell Labs an academic environment within a big company -- but as to the forces in the invention of C, its further development, and its early commercial success, that's just misleading. (Please don't lecture me on how bad C is. A lot of time has passed since 1982. ) >... > > The problem, people, is that we need to make programming easier and > simpler, not harder. If OO was easier, or more economical, we would have > seen clear evidence by now in the numbers. Some of us, including some so old as to remember a time before C, have seen the clear evidence in their own work and that of colleagues. That the technology is insanely over-hyped doesn't invalidate the ideas. Or surprise anyone. >... -- Dan Drake dd@dandrake.com http://www.dandrake.com/index.html
Newsgroups: comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming From: bks@netcom.com (Bradley K. Sherman) Subject: Re: Programming and the colossal failure to advance Message-ID: <bksEvp5Jy.945@netcom.com> Organization: DNA + Sunlight References: <6ng34a$j2g$1@nnrp1.dejanews.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> <vhIsdqY67dTD-pn2-dGSKqdjMrSfd@dynamic47.pm03.san-rafael.best.com> Date: Mon, 6 Jul 1998 23:35:58 GMT Sender: bks@netcom19.netcom.com In article <vhIsdqY67dTD-pn2-dGSKqdjMrSfd@dynamic47.pm03.san-rafael.best.com>, Dan Drake <dd@dandrake.com> wrote: > ... >Somebody who was in school in the very early 80s (before which C had no >real-life non-Unix presence whatever) could maybe tell us whether C was de >rigueur then. My recollection is that Pascal owned the whole academic >show except for the Unixoids (to whom praise and thanks). > ... Even at Berkeley in that period Pascal was the choice for undergraduate work. If you look at the popular computer magazines around 1981 you will see Pascal, Modula, Ada and Assembler, but no C. C took off because it was cross-platform, small and fast. I still have a functioning Kaypro 10 (CP/M) with an Aztec Manx C compiler that had almost a complete implementation of the standard library. There was a BDS C compiler that sacrificed some parts of the language but was really, really fast under CP/M. C took over because you could get work done and you had real reuse of the code when porting. In all honesty I would say that C is still, in some sense, more portable than Java. --bks
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming Subject: Re: Programming and the colossal failure to advance Date: 6 Jul 1998 16:21:53 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6q1udv.bko.rog@talisker.ohm.york.ac.uk> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6nh384$4mj$1@nnrp1.dejanews.com> <359C16EA.5AF08B84@home.com> On Thu, 02 Jul 1998 23:32:02 GMT, John Mott <johnmott@home.com> wrote: > There are a few hard facts about programming as a trade/profession. Its a > science to the degree that its a study of nature; logic and problem solving > all within the constraints of nature. To that end its simply as hard as > it is. [...] > There is a simple reason for this: the universe is infinitely complex. > Its nature. Its inevitable. It can't be worked around. much of the complexity encountered in computer programming is unnecessary - a consequence of lazy programmers that understand a task just well enough for the moment, but not well enough for the general case. we've all encountered this - it's easier to make the program more complex (one more feature == one more piece of code) than to simplify back to a system which is more than powerful enough to meet most of our demands. as with all good engineering solutions, the *right* solution seems tailor-made for the job in hand - it shows us how to proceed, and there's no fuzziness around the edges where the vision of the software engineer has become ropy. that's why unix became so popular - its fundamental concepts were as simple as possible, but no simpler. simple enough to encompass an enormous range of tasks, but not too complex for the simple tasks. software *is* too complex today because too many people are scattering and multiplying entities, and not enough people are synthesising and gathering entities together. despite "information hiding" being part of the mantra of the object-oriented programmer, not nearly enough information hiding goes on in software today, and the languages in use don't help that. *any* software module that "extends" functionality of another module is guilty of "information exposure" - the most successful, long-lived and reusable software modules of the past have been self-contained modules, reliant on other modules only for their implementation details, not for their interface definitions. class-based languages are some of the worst offenders - because a user of an object must understand not only the interface to that object, but the interface to all its superclasses as well, and how the object relates to its superclasses. responsibilities and obligations of an object become indistinct, and the implementations of objects become set in stone - unchangeable because objects that use them rely on behaviour that should really be implementation-specific but is visible by default; and unreusable because the definition of the object becomes much too tied up in the environment it's defined in. yes, software does do more these days, but the complexity is *not* unavoidable - it's just easier in the short term not to avoid it. creeping featurism riddles today's software - simplicity and minimalist software are the only escapes. for an example of a language which combines simplicity with complete modularity along with great power, while avoiding the pitfalls of the object-oriented crowd, check out Limbo from Bell Labs (and the original creators of unix). it's with languages such as this (and environments such as Inferno) that software useability (and re-useability) could take a real leap forward. (see http://www.lucent-inferno.com/Pages/Developers/index.html) IMHO! rog.
From: dolby@infi.net (Jack Dolby) Newsgroups: comp.sys.mac.programmer.misc,comp.sys.next.programmer Subject: Re: Thoughts on Realbasic Date: Mon, 06 Jul 1998 17:30:16 GMT Organization: InfiNet Message-ID: <35a2092c.11569696@allnews.infi.net> References: <6nqtoo$n4s@drn.newsguy.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Chris, I don't have the address here (at work on a non-mac system... sorry) to subscribe, but there is a very active Realbasic mailing list. there should be a link to it on the realbasic home page. I think the experts there can give you some realy good heads-up info on realbasic. I jsut lurk on that mailinglist now and plan to start experimenting with realbasic this weekend. jack chris wrote: >Real Software has just released their visual basic >look-a-like for the mac. The ide is 92% carbon >compatible. Has anyone had any experiences with >this ide? > >http://www.realsoftware.com/ > >Chris Van Buskirk >Software Engineer >NBC Jack Dolby Gately Communication Company 757.826.8210 x 3012
From: sdm7g@elvis.med.Virginia.EDU (Steven D. Majewski) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 7 Jul 1998 03:39:20 GMT Organization: University of Virginia Message-ID: <6ns598$il4$1@murdoch.acc.Virginia.EDU> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> In article <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net>, Judson McClendon <judmc123@bellsouth.net> wrote: > >Since there were a number of replies to this post, I'll answer them all >here. > >First, I wrote my post hastily and did not clearly articulate what I >meant to say. I used the word 'university', when I should have said >'academia', meaning the broader base of research and theoretical folks, >such as Xerox PARC, not just universities. Sorry. > >Here is an example of how C became popularized. [ ... ] I can agree with some of your comments on complexity, however, your just as wrong with your history of C (as others have pointed out) as you were with your history of OO. I don't think it's a problem of not clearly articulating -- it's just plain wrong! We can argue about opinions about good software design, and other subjective matters, but this stuff is well documented history. Thinking about it harder and articulating it more carefully doesn't help if the facts are all wrong! ---| Steven D. Majewski (804-982-0831) <sdm7g@Virginia.EDU> |--- ---| Department of Molecular Physiology and Biological Physics |--- ---| University of Virginia Health Sciences Center |--- ---| P.O. Box 10011 Charlottesville, VA 22906-0011 |--- "I'm not as big a fool as I used to be, I'm a smaller fool." - Jack Kerouac Some of the Dharma <http://members.aol.com/kerouacsis/SomeDharma.html>
From: David Stes <stes@mundivia.es> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming Subject: Re: Programming and the colossal failure to advance Date: Tue, 07 Jul 1998 10:04:36 +0000 Organization: Mundivia, Santander Message-ID: <35A1F2B4.446B9B3D@mundivia.es> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6nh384$4mj$1@nnrp1.dejanews.com> <359C16EA.5AF08B84@home.com> <slrn6q1udv.bko.rog@talisker.ohm.york.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Roger Peppe wrote: > > for an example of a language which combines simplicity with complete > modularity along with great power, while avoiding the pitfalls of the > object-oriented crowd, check out Limbo from Bell Labs (and the original > creators of unix). it's with languages such as this (and environments > such as Inferno) that software useability (and re-useability) could > take a real leap forward. > (see http://www.lucent-inferno.com/Pages/Developers/index.html) I think what's attractive in Limbo is also due to the fact that Limbo is relatively new, and has just been developed (probably by a small group of people). Sort of like Objective-C or C++ in their first years, when those languages also had only small (and therefore well-informed) user communities. Once Limbo would 'go commercial' ("buy our implementation of Limbo and your problems are over"), and once Limbo would be adopted by a big company for mass-distribution, say, Apple's YellowBox would be in Limbo (instead of Java), then surely people would ask for all sort of garbage and crap to be added to Limbo and completely spoil its original simplicity and elegance. Of course, this pessimist point of view is not at all an argument against Limbo, which is indeed sympathetic and Limbo's a very fine language.
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.pr Subject: Re: Programming and the colossal failure to advance Date: 7 Jul 1998 10:21:29 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6q3tlc.cc3.rog@talisker.ohm.york.ac.uk> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6nh384$4mj$1@nnrp1.dejanews.com> <359C16EA.5AF0 <35A10AA7.F0C361F6@pobox.com> On Mon, 06 Jul 1998 13:34:31 -0400, Ralph Cook <rcook@pobox.com> wrote: > Roger Peppe wrote: > > > that's why unix became so popular - its fundamental concepts were as > > simple as possible, but no simpler. simple enough to encompass an > > enormous range of tasks, but not too complex for the simple tasks. > > Let's not get too starry-eyed about why Unix became popular. The > universities liked it because it was free. The Computer Science people > liked it because (a) it was what their school used, (b) once they liked > it, many of them felt compelled to proselytize it as "the only good > thing ever written", (c) it was a "secret code" with which they could > impress their friends and betters. there were many other systems around as well as unix. i know that i for one liked it, not because it was the only system around (there were also VMS, DOS, and other systems available) but because its designers had obviously had simplicity as one of its main guiding principals. the things that unix did were done well - they often captured the essential simplicity of a problem while keeping its applicability to the problem domain. think about the power of the basic unix system calls: fd = open(file, mode) read(fd, buffer, buffersize) write(fd, buffer, buffersize) close(fd) dup(fd) fork() exec(path, args) in the original unix these extremely simple calls nonetheless conferred great power on the developer, because they presented a uniform and consistent interface to different types of external access. but there's not an ounce of spare fat in those calls, no special-case structures, no optional arguments - their final design is both necessary for their functionality, and sufficient for an enormous range of problems. that, to me, is a mark of good design, and one that is now all too rarely seen - bloatware is the norm. when looking at the documentation for a piece of software, one's reaction should be "wow, just think of all the things i can do with this" (and know immediately how one would go about doing those things), not "hmm, so what's this talking about, and will it really be sufficient for what i want to do?" (and not have the faintest idea of the best way to process). today, many software developers think that because the final application is complex, no part of it is simple, so the simplest tasks gather horribly complex interfaces, and the more complex tasks remain completely untouched. (think about the power of the "make" utility - it's very rare to see a modern application that solves such a general problem in such a simple but powerful way. with the added complexity of windowing, nobody thinks about the really hard bits of the problem) rog.
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.pr Subject: Re: Programming and the colossal failure to advance Date: 7 Jul 1998 10:37:43 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6q3ujp.cc3.rog@talisker.ohm.york.ac.uk> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6nh384$4mj$1@nnrp1.dejanews.com> <359C16EA.5AF0 <35A1F2B4.446B9B3D@mundivia.es> On Tue, 07 Jul 1998 10:04:36 +0000, David Stes <stes@mundivia.es> wrote: > Roger Peppe wrote: > > > > for an example of a language which combines simplicity with complete > > modularity along with great power, while avoiding the pitfalls of the > > object-oriented crowd, check out Limbo from Bell Labs (and the original > > creators of unix). it's with languages such as this (and environments > > such as Inferno) that software useability (and re-useability) could > > take a real leap forward. > > (see http://www.lucent-inferno.com/Pages/Developers/index.html) [...] > Once Limbo would 'go commercial' ("buy our implementation of Limbo and > your problems are over"), and once Limbo would be adopted by a big > company for mass-distribution, say, Apple's YellowBox would be in Limbo > (instead of Java), then surely people would ask for all sort of garbage > and crap to be added to Limbo and completely spoil its original > simplicity and elegance. perhaps... and of course limbo (and inferno) is no panacea. it just seems to be taking a good section of "state of the art" language technology, and packaging it in a way that is completely accessible and deliciously simple, while sacrificing very little generality. its simplicity means that implementations will get better and better; its design (including the superbly designed supporting libraries) means that i'd trust that programs written now in limbo will still work unchanged in 20 years. that's something i couldn't believe of any other language/environment. rog.
From: jonas.v.andersson@seinf.mail.abb.com (Jonas Andersson) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Tue, 07 Jul 1998 04:45:40 GMT Organization: ABB Business Systems AB Message-ID: <35a1a79f.86154874@138.221.200.100> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <35A0A008.41C6@pc-plus.de.remove.this.bit> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit On Mon, 06 Jul 1998 11:59:36 +0200, Stephen Riehm <sr@pc-plus.de.remove.this.bit> wrote: >even the most advanced systems I've seen still rely on the >programmer typing freestyle text, and hoping that the compiler will >be able to make sense of it - as you gain experience you also gain >the skill to be able to do this, but you still have to jump through >hoops if you want to "find that class of object which does xyz" or >to find out exactly what were all those parameters to his_func()? Have a look at VB5 in the: " "find that class of object which does xyz" or to find out exactly what were all those parameters to his_func()?" deparment.
From: <wthock@pop.jaring.my> Newsgroups: comp.sys.next.programmer Subject: LOW PRICE CPU !!! Date: Tue, 07 Jul 1998 18:56:52 PST Organization: The one and only Message-ID: <07071998185652wthock@pop.jaring.my> content-length: 2287 Dear Sir, Now a day, most of the PC builders and PC retailers are complaining on aseveral things: Price Competitive, No Profit Margin, How Can We Beat The Price of Direct Sales Computer Companies such as DELL, GW2000? Here is the solution, most of the branded machine will not bring you profit,and yet not competitive enough. And you start thinking making your own brand, assemble your own brand of PC, but still, how competitive you are on price? The answer is: TO USE SOME PARTS WHICH IS LOWER COST ON SAME OR HIGHER QUALITY!!! Here we come in to introduce you with our new RE CPU, this CPU is absolutelyIntel Original CPU, but with some modiication to run on higher speed. Due to ourhigh technology and good quality control, we have NEVER RECIEVED A SINGLE RETURNon our RE CPU. If you want your PC line to be competitive, try this (shh... asksomeone in Taiwan, and you will know, Acer using RE CPU for the home/business PCline as well....shh..): 1. Intel Pentium II 266MHz - US$175 tray (without Fan) 2. Intel Pentium II 266MHz - US$185 Retail Box - you can sell this on yourrack 3. Intel Pentium II 300MHz - US$188 tray (without Fan) 4. Intel Pentium II 300MHz - US$199 Retail Box - you can sell this on yourrack **FOB Hong Kong **Minimum order for Tray version is 25pcs, Box version is 10pcs **If you need a formal quotation on Ringgit Malaysia, please request by email to aithk@hkid.com , with the quantity you required, and we will fax you theformal quotation as soon as possible. IN CASE YOU STILL ASSEMBLE MMX-200 MACHINE and CUSTOMER is INSIST on INTEL CPU. We are having the following: ** Intel Pentium MMX-200 tcp converted to ppga to accept socket 7. ** note that this is Intel Original Chip build for Notebook, and the factory using an adapter to change the grid to PPGA (socket 7), same performance as MMX-200 (SL27J). ** The Price? USD75@ for 25 ~ 99pcs USD73@ for 100 ~ 499pcs USD72@ for 500+ pcs FOB Taiwan We also carry wide range of Memory Module from SIMM EDO to PC100 DIMM SDRAM,email us at aithk@hkid.com for more info. Thank you for your time. AIT COMPANY OF HONG KONG Well Fung Ind. Centre 58-76 Ta Chuen Ping Street Kwai Chung, N.T., Hong Kong
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <07071998185652wthock@pop.jaring.my> Control: cancel <07071998185652wthock@pop.jaring.my> Date: 07 Jul 1998 12:01:17 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.07071998185652wthock@pop.jaring.my> Sender: <wthock@pop.jaring.my> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: "Judson McClendon" <judmc123@bellsouth.net> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> <6ns598$il4$1@murdoch.acc.Virginia.EDU> Subject: Re: Programming and the colossal failure to advance Organization: Sun Valley Systems Message-ID: <Snpo1.200$Od.706436@news4.mia.bellsouth.net> Date: Tue, 07 Jul 1998 13:28:18 GMT NNTP-Posting-Date: Tue, 07 Jul 1998 09:28:18 EST Steven D. Majewski wrote: >Judson McClendon wrote: >> >>Since there were a number of replies to this post, I'll answer them all >>here. >> >>First, I wrote my post hastily and did not clearly articulate what I >>meant to say. I used the word 'university', when I should have said >>'academia', meaning the broader base of research and theoretical folks, >>such as Xerox PARC, not just universities. Sorry. >> >>Here is an example of how C became popularized. [ ... ] > >I can agree with some of your comments on complexity, however, your >just as wrong with your history of C (as others have pointed out) >as you were with your history of OO. >I don't think it's a problem of not clearly articulating -- >it's just plain wrong! We can argue about opinions about good >software design, and other subjective matters, but this stuff is >well documented history. Thinking about it harder and articulating >it more carefully doesn't help if the facts are all wrong! It's been interesting to read all these comments to my post, particularly those about people who were around before C, because I started programming in 1968. :-) Anyway, it is clear that many of the respondents here have yet to learn the lesson that just because it is in books, and is being taught at school, doesn't mean it is correct! Most of what I say in that post are things I personally was witness to, or was fairly close to. -- Judson McClendon This is a faithful saying and worthy of all Sun Valley Systems acceptance, that Christ Jesus came into the judmc123@bellsouth.net world to save sinners (1 Timothy 1:15) (please remove numbers from email id to respond)
From: "Judson McClendon" <judmc123@bellsouth.net> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> <james-ya023180000407981226410001@news> Subject: Re: Programming and the colossal failure to advance Organization: Sun Valley Systems Message-ID: <sApo1.203$Od.711903@news4.mia.bellsouth.net> Date: Tue, 07 Jul 1998 13:41:44 GMT NNTP-Posting-Date: Tue, 07 Jul 1998 09:41:44 EST James McCartney wrote: >Judson McClendon wrote: > >>the pudding is in the eating." Hard, cold, objective results show that >>the result of these new development environments, including OO, are >>greatly increased costs and development time, and less reliability. > >What results? Trials I've seen posted using good OO environments such >as Smalltalk show reduced costs and development times. C++ is likely not >the best test of the benefits of OO. (In fact a friend told me of his >team of a few Smalltalk programmers who were so fast at getting >their own work done that the larger team of C++ programmers weren't >ready with their component to test on time so the Smalltalk team mocked up >that part as well for testing and it worked so well that the C++ >programmers were all let go.) I mentioned a couple of front page articles from the Wall Street Journal and Info World. Why don't you go read them? And Smalltalk probably represents less than .001% of the real world production programming being done. Wow! >>simpler, not harder. If OO was easier, or more economical, we would have >>seen clear evidence by now in the numbers. What we do see is clear >>evidence to the contrary. If we want economical, reliable software, we > >So you propose that OO makes things too complex and we should go back >to procedural programming? If so, you're a nut case. >OO is a way to deal with complexity. Why didn't you read what I wrote? OO as it is being done today is failing in many areas. Your blind faith doesn't change that one whit, nor does it impress the companies who have wasted billions on projects they could not finish. However, I am not blaiming just OO; it is a broad issue involving many factors. Part of the problem may be poor OO implementations, but the biggest factor is that we are making many systems much more complex than necessary, and we are making programming tools much more complex than we should. Going toward *all* GUI applications is one thing which I believe to be very mis directed. Programming GUIs is more complicated, and many applications simply do not need it. I'm simply saying that we need to look at the situation based on needs and results, not hype and what we were told by others who haven't had to make this stuff work in real world situations. -- Judson McClendon This is a faithful saying and worthy of all Sun Valley Systems acceptance, that Christ Jesus came into the judmc123@bellsouth.net world to save sinners (1 Timothy 1:15) (please remove numbers from email id to respond)
From: "Judson McClendon" <judmc123@bellsouth.net> Newsgroups: comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> <vhIsdqY67dTD-pn2-dGSKqdjMrSfd@dynamic47.pm03.san-rafael.best.com> Subject: Re: Programming and the colossal failure to advance Organization: Sun Valley Systems Message-ID: <zVpo1.206$Od.720988@news4.mia.bellsouth.net> Date: Tue, 07 Jul 1998 14:04:15 GMT NNTP-Posting-Date: Tue, 07 Jul 1998 10:04:15 EST Dan Drake wrote in message ... >Judson McClendon wrote: >> >> Here is an example of how C became popularized. In the 70's, DEC was >> seriously dominant in the university environment. Because Bell Labs had >> written C for the PDP, most of the system software for those computers >> was in C. Students who wanted to seriously get into the system software >> simply had to master C. As those students graduated and moved into >> industry, they knew C, and stumped for C. This is where the main impetus >> for C in the industry came from. > >Popularized schmopularized. I wrote (co-wrote) one of the first PC >applications ever written in C, at the time that Lotus 1-2-3 was being >done in assembler. Was that because I learned C in school? Not quite; >they didn't teach C when I was in school: in 1967! > >By 1982 I had 15, count them, years of working experience in assembler, >Fortran, and other things. (You know -- Cobol, Pascal, Lisp, Simula, and >various things you've never heard of.) No C, except reading K&R and >thinking at first that it looked awfully kludgy. The reason my colleagues >and I decided on C was that no other available choice could possibly be >adequate for the task (though PL/I, believe it or not, came close). So we >learned C and used it. There were no freshly minted CS graduates in the >company -- you'd laugh if you heard how many there were of any kind -- >just programmers who needed to make a CAD program work and knew what they >had to do. Yes, I've heard of those languages. I have done production programming programming in over 40 languages and dialects. Between five and six million lines of COBOL, and something over a million lines of various other languages, since 1968, many of them you've never heard of, because I wrote some of the compilers myself. Your personal experience has no bearing on my description of how C was popularized by universities. I never said they was the *only* reason anybody used C. ;-) I even use VC++ myself. <gasp!> >Somebody who was in school in the very early 80s (before which C had no >real-life non-Unix presence whatever) could maybe tell us whether C was de >rigueur then. My recollection is that Pascal owned the whole academic >show except for the Unixoids (to whom praise and thanks). Pascal *did* own the show, officially. But my point is that, because the PDP was so dominant in many universities, the students who got heavily involved in the system software, of necessity, *had* to do it in C. You should understand that the "Official Position" and what is happening at ground level are sometimes very different, and the results may be unexpected. >>... >> To say that a language developed by academia would be better, because they >> would prefer something elegant, is amusing. Nothing wrong with elegance, >> but unless it works in the real world, on real world applications, the >> point is irrelevant. Most academic folks have little or no experience in >> writing large, real world applications, and in maintaining them over time. >>... > >On this, I'll leave the dissent to someone else, since I mostly agree. My >problem is just that C is a terrible example for your point; its strength >was that it was _not_ developed for academic elegance or big-company >marketing, but for the needs of some programmers who wanted to get a job >done. To be sure, you could call Bell Labs an academic environment within >a big company -- but as to the forces in the invention of C, its further >development, and its early commercial success, that's just misleading. My only reason for using C was as an example of how people can be influenced in ways that may be unexpected. >> The problem, people, is that we need to make programming easier and >> simpler, not harder. If OO was easier, or more economical, we would have >> seen clear evidence by now in the numbers. > >Some of us, including some so old as to remember a time before C, have >seen the clear evidence in their own work and that of colleagues. That >the technology is insanely over-hyped doesn't invalidate the ideas. Or >surprise anyone. My point is not to hack OO. I am trying to say that we are making our programming tasks a lot more difficult by our choices of development languages and environments, and by throwing advanced technology at problems which simply do not warrant it. Because of this, we are seeing the beginnings of what I believe will be a real backlash against overcomplexity in systems and in development environments. Time will show me correct or not. In a related thought, I usually can't stand Steve Jobs, but he said one thing that really made a lot of sense to me. He said "It is not in Microsoft's interest to make programming easier." After all, what difference does it matter to Microsoft if it takes $1 million or $5 million to rewrite Office, since they sell so many copies it amortizes to a few cents per copy? But it makes a *big* difference to all those smaller software developers. Considering how agressive Gates is, do we really think he has overlooked this issue? Remember "It isn't done until Lotus won't run?" Think about that as you're modifying code to work with the next release of VB, or the next changes in the Windows API. -- Judson McClendon This is a faithful saying and worthy of all Sun Valley Systems acceptance, that Christ Jesus came into the judmc123@bellsouth.net world to save sinners (1 Timothy 1:15) (please remove numbers from email id to respond)
From: sdm7g@elvis.med.Virginia.EDU (Steven D. Majewski) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 7 Jul 1998 16:34:16 GMT Organization: University of Virginia Message-ID: <6ntim8$8k9$1@murdoch.acc.Virginia.EDU> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> <james-ya023180000407981226410001@news> <sApo1.203$Od.711903@news4.mia.bellsouth.net> In article <sApo1.203$Od.711903@news4.mia.bellsouth.net>, Judson McClendon <judmc123@bellsouth.net> wrote: >I mentioned a couple of front page articles from the Wall Street Journal and >Info World. Why don't you go read them? And Smalltalk probably represents >less than .001% of the real world production programming being done. Wow! Do you have any basis for that .001% estimate, or is it just a number you made up out of thin air ? ---| Steven D. Majewski (804-982-0831) <sdm7g@Virginia.EDU> |--- ---| Department of Molecular Physiology and Biological Physics |--- ---| University of Virginia Health Sciences Center |--- ---| P.O. Box 10011 Charlottesville, VA 22906-0011 |--- "I'm not as big a fool as I used to be, I'm a smaller fool." - Jack Kerouac Some of the Dharma <http://members.aol.com/kerouacsis/SomeDharma.html>
From: james@unspam.edu (James McCartney) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Tue, 07 Jul 1998 14:20:17 -0700 Organization: The University of Texas at Austin, Austin, Texas Message-ID: <james-ya023180000707981420170001@news> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> <james-ya023180000407981226410001@news> <sApo1.203$Od.711903@news4.mia.bellsouth.net> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <sApo1.203$Od.711903@news4.mia.bellsouth.net>, "Judson McClendon" <judmc123@bellsouth.net> wrote: >I mentioned a couple of front page articles from the Wall Street Journal and >Info World. Why don't you go read them? I don't have them around. If they blame failed projects on increased complexity due to OO then they are based on a fallacy. OO *reduces* complexity by allowing the programmer to express similarities via polymorphism. Without OO you wind up writing lots of extra code. So OO is not the problem. Perhaps it is using a complex implementation of it in something like C++. >And Smalltalk probably represents >less than .001% of the real world production programming being done. Wow! Got a quote on that? comp.lang.smalltalk seems to always have a large number of jobs postings. >Why didn't you read what I wrote? The above remark is obviously only there for incendiary effect. I'll ignore it. >OO as it is being done today is failing >in many areas. define " OO as it is being done today". I'd bet a lot of things may be being blamed on OO here that have nothing to do with OO. >Your blind faith doesn't change that one whit, nor does it More flames.. wa wa wa... >the >biggest factor is that we are making many systems much more complex than >necessary, OK perhaps I'll grant that. However you still beg the question: What are you proposing? to stop using OO? The complexity problem I think has ABSOLUTELY NOTHING to do with OO or other comp sci stuff, but with the following things: Botched APIs. Compatibility with legacy apps, APIs, OSes. (even bug-for-bug) Multiple platform support. Re-porting away from abandoned/obsolete APIs, OSes, hardware, etc.. Users that demand that software do everything and with a GUI. Having to simultaneously support multiple standards for: formats, protocols, drivers, you-name-it.. Having no standards at all. example: GUI APIs. Having to circumlocute standards/APIs due to efficiency concerns. Having to maintain or port code that circumlocutes standards/APIs due to efficiency concerns. Having to maintain/rewrite code that was written with excessive emphasis on optimization(speed/space) rather than longevity/maintainablility/portability. an example: Y2K saved 2 bytes. --- James McCartney
From: Byron Goodman <bgoodman@dev.tivoli.com> Newsgroups: comp.sys.next.programmer Subject: Docs Date: Tue, 07 Jul 1998 12:49:41 -0500 Organization: Tivoli Systems Inc., Austin, TX Message-ID: <35A25FB5.513AC61C@dev.tivoli.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Are there any NeXT or OpenStep Mach developer documention available? I'm needing to find out about system calls that query hardware and devices. I found the Rhapsody Operating System Software Guide, but has very limited information. The documentation I acquired from CMU I have found is not completely consistent with NeXT/OpenStep. -- Byron Goodman bgoodman@dev.tivoli.com Tivoli Systems, an IBM Company
From: NOSPAMmbkennelNOSPAM@yahoo.com (Matt Kennel) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming Subject: Re: Programming and the colossal failure to advance Date: 7 Jul 1998 20:39:20 GMT Organization: University of California at San Diego Message-ID: <slrn6q51ro.cm8.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6nh384$4mj$1@nnrp1.dejanews.com> <359C16EA.5AF08B84@home.com> <slrn6q1udv.bko.rog@talisker.ohm.york.ac.uk> <35A10AA7.F0C361F6@pobox.com> On Mon, 06 Jul 1998 13:34:31 -0400, Ralph Cook <rcook@pobox.com> wrote: :Roger Peppe wrote: : :> that's why unix became so popular - its fundamental concepts were as :> simple as possible, but no simpler. simple enough to encompass an :> enormous range of tasks, but not too complex for the simple tasks. : :Let's not get too starry-eyed about why Unix became popular. The :universities liked it because it was free. The Computer Science people :liked it because (a) it was what their school used, (b) once they liked :it, many of them felt compelled to proselytize it as "the only good :thing ever written", (c) it was a "secret code" with which they could :impress their friends and betters. I go for (d) It came with Sun workstations which were obscenely superior to VT240's hooked up to VAX 11/780's running VMS. -- * Matthew B. Kennel/Institute for Nonlinear Science, UCSD - * "People who send spam to Emperor Cartagia... vanish! _They say_ that * there's a room where he has their heads, lined up in a row on a desk... * _They say_ that late at night, he goes there, and talks to them... _they *- say_ he asks them, 'Now tell me again, how _do_ you make money fast?'"
From: Bill Ghrist <ghristwd@pgh.net> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Tue, 07 Jul 1998 09:41:38 -0400 Organization: Westinghouse NPD, but these are my own opinions. Message-ID: <35A22592.CA7DF837@pgh.net> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <35A0A008.41C6@pc-plus.de.remove.this.bit> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit But what you describe exists now for some applications. Try MathCAD as one example. Precisely because what you see on the screen doesn't look like a traditional computer program, you have overlooked a perfect example of what you are asking for. Regards, Bill Ghrist Stephen Riehm wrote: > > All paradigm and programming language discussions aside, I find it > ironic that programmers really have been left in the dark as far as > software tools go. > > For example, with the exception of a very small minority, no-one > writes documents with a text editor any more. Everyone uses word > processors with automatic indexing, formatting, cross referencing, > spelling and gramitic checks. All the user has to do is type in the > words, and click on a button or two to tell the computer what to > do with those words. The user has no idea what actually gets written > to disk - and doesn't need to either. > > The same should be true for programming. The representation on the > screen doesn't have to be the same as the representation on disk. > If we had code processors, programs could be treated like hypertext > documents, interactive - living - systems. What you see on the screen > is just a representation of the program - formatting becomes a > non-issue, reading someone elses code would be much easier, because > at least the visual representation remains familiar, leaving you > just the task of understanding the algorythim. Finding out what > lies behind a variable, class, method or function is as simple as > clicking on the symbol - the relavant information would be taken > directly from the symbol table, and you would KNOW that it's the > right information. Problems related to multiple definitions of > symbols in different contexts also disapear, because the code > processor knows exactly which symbol you wanted when you wrote your > code, no text based system can do this, instead, we rely on the > black-box concept of a compiler to work out which symbol was meant. > Today's compilers still don't give any live feedback, nor do they > remember what they did before (OK, libraries and pre-compiled headers > are a step in that direction, but it's only the first step). > > even the most advanced systems I've seen still rely on the > programmer typing freestyle text, and hoping that the compiler will > be able to make sense of it - as you gain experience you also gain > the skill to be able to do this, but you still have to jump through > hoops if you want to "find that class of object which does xyz" or > to find out exactly what were all those parameters to his_func()? > > I think this is more the essence of the original posting - we've > already discussed paradigms and languages to death, but the underlying > system of a user writing text and hoping that the computer will cope > is still there. > > Steve > > // __________________________ ._ o ________________________________ > \\\\ Stephen Riehm / //\. Stephen.Riehm@pc-plus.de > //// pc-plus ' \>> | Phone: +49 89 45566 148 > \\ 81675 Munich, Germany \\ ` Fax: +49 89 45566 113 > > -----BEGIN PGP PUBLIC KEY BLOCK----- > Version: 2.6.2 > > mQCNAjWaUeoAAAEEAMSI/hFLWlcAFVSwMHdHFiSbXP+IfbDzHg1IUSjlGseTsi5o > kxDRFjvpxJUpH8S4XHbVAkyAMyysNaBI/LxuCvS66WiMINzJTxeDMJCUnzGpLcQH > h/Co2ymaw0KnjHJtBvFAptntu8wgEM712cU0LSdoD25x1faWDDlT6pxFM0BlAAUR > tChTdGVwaGVuIFJpZWhtIDxTdGVwaGVuLlJpZWhtQHBjLXBsdXMuZGU+ > =+MnD > -----END PGP PUBLIC KEY BLOCK-----
From: Bill Ghrist <ghristwd@pgh.net> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Tue, 07 Jul 1998 09:26:39 -0400 Organization: Westinghouse NPD, but these are my own opinions. Message-ID: <35A2220E.67A2E5F@pgh.net> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359BB163.7A97DEC6@gssec.bt.co.uk> <6nh07l$hl$1@nnrp1.dejanews.com> <359C25D4.7B2A@computer.org> <6nmgrl$tfp$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit hmmueller@my-dejanews.com wrote: > > In article <359C25D4.7B2A@computer.org>, > b.thomas@computer.org wrote: > > > > hmmueller@my-dejanews.com wrote: > > > > > > > > Thanks for saying that! I dont know why, but there is still this idea around > > > that computers should be programmed by laymen - electrical engineers, > > > astronomers, mechanical engineers etc. Nobody would allow me to design > [...] > > > > Most of the founding fathers of computer science came from other fields > > - mathematics, electrical engineering etc. Do you imply that they were > > mediocre too, since they were also self-taught (though most of that > > self-teaching process was part of research). > > No, I dont. My remark is a general one; this includes that there may be many > intelligent non-formally-CS-trained people who are great with CS and > computers and programming. If I say that physicians' knowledge of music > shouldn't be the guideline for professional music, I still accept that Albert > Schweitzer (a physician) was one of the greatest organists and > Bach-specialists around. > > What I say is exactly what I say: That programming is wrongly assumed to be > something everybody (or at least every engineer) should be able to do, whereas > this does not hold for other engineering disciplines. > > Re the argument that it is easier to train a biologist in programming than a > programmer in biology: > > In the one (university) biology lab where I had friends, they had a lab > technician. If they needed a new cage; or a special machine to to do an > experiment with chimpanzees (3 switches, a few food outlets and two lamps, in > one case), it was clear that the design and building of the machine was NOT > the biologist's job; he would put a sketch on paper and talk with the > technician who would then design and build the machine; in the case of the > cage, he would usually only do the design, and the cage would be built by > some welding company or so. All that is obvious. > > So why was it assumed that the same biologist would write the program for > controlling the lamps and food outlets and measuring the timing of the > chimpanzees' responses himself (where he failed miserably)? > Or that me (the programmer coming in afterwards) would have to understand > biology to write a small real-time program? > > Neither has the biologist to learn programming, nor the programmer biology. > They just have to talk to each other. It works; and is almost always much > cheaper, and more more certainly gets the job done. > > And again: If someone has all the required expertise, from biology to > programming to turning some part on a lathe, let him/her do it. But, please, > dont assume that it is the fault of CS if he/she cant. It is simply > unreasonable to require programming to be simple. > > Harald M. Mueller > > -----== Posted via Deja News, The Leader in Internet Discussion ==----- > http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum But the example you have given is not a biology problem, it is a control system problem. I am not surprised that the technician is better than the biologist at programming a control system. But if the biologist needs a program that would take the results of the experiments and perform some analysis on them according to the principles of biology (don't ask me for an example, I'm not a biologist) then I would expect the biologist to be better at that (after some education in programming, of course). Regards, Bill Ghrist
From: tj_sam@my-dejanews.com Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re:Programming and the colossal failure to advance - VR Date: Tue, 07 Jul 1998 21:21:22 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6nu3gi$97i$1@nnrp1.dejanews.com> Interesting thread... I basically agree with the original posting (there is an overall increase in totally unnecessary complexity). Some replies have been in agree and a few disagree with historical evidence to back up there view. These are all comments on the present state of computer programming, the follow is a speculative view into the possible cause of this failure and even a partial solution thrown in. ........ On the assumption that a major advance to programming would represent a overall simplification, I would agree that there has been a colossal failure to advance programming. I would speculate that a possible reason for this failure is due to the fact that we have reached the practical limits for pure semantic programming. There is an intrinsic level of noise to semantics definitions which is required in human languages, however this noise does creates problems and limits when applied to computer programming. The most basic effect of reaching this "semantic noise limit" would be that any attempts to advance programming would simply lead to an increased complexity rather than a simplification. A reoccurring theme in the replies to this thread has been the suggestion to move programming to more visual or graphic presentations. Generalized, this represents a trend (or at least a wishlist) towards a type of virtual reality programming. The possibility of actual programming in a VR environment would constitute an advancement, however it remains fiction. The question I've often asked myself is: Why not? IMO, the primary factor preventing a move towards VR programming is the totally semantic structure of source code. Program editors and the code they generate have no actual graphic elements. You can attach graphic elements such as graphically defined class structure in C++, but these are secondary elements and the core of programming remains totally and completely a semantically based process (its all text files). Question:Is this semantic based limitation to programming a natural limitation or an imposed limitation? Answer: It's a very unnatural limitation and goes back to the very concept as programming as a "Language". Within its core, programming has always contained both graphic and text definable elements. The graphically based elements are represented by the control flow of the program. Conditional and unconditional branches and loops constitute a core programming element that can be directly represented by graphic vectors. For example, "goto" in C can be directly represented visually as a graphic vector with source and destination directly in the actual source code (you don't need semantic labels). A basic VR program editor would essential be a flowcharting program with the actual code embedded within the flowchart (not the simplex flowchart but actual source level flowcharting). A multiple column text editor in which major branches and loop would be directly represented by visual vectors. Individual subroutines or functions would be represented as 2D (X,Y) wire frame objects. If subroutine or function calls are represented as Z plane bi-directional vector (recursive function are outlawed) then modules and complex programs can be represented as Virtual 3D wireframe objects. With our present semantically based programming, source code amounts to a meaningless and endless one dimensional string of computer code held together by semantic directives and the mindset of the programmers. We can change this to a very simple and direct visualization of complex source code as a 3D objects, which would be step towards Virtual Reality Programming. Creating a large programs has the same level of complexity as building a skyscraper, can you image the architect and engineers using primarily semantic definitions (a long narrative book (source code) with a few pictures)? Essential this is what programmers have been doing all along, working without software blueprints. Flames anyone :-) T.J. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: Martin@nezumi.demon.co.uk (Martin Tom Brown) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Tue, 07 Jul 1998 20:05:19 GMT Organization: Nezumi Message-ID: <899841919snz@nezumi.demon.co.uk> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> On Saturday, in article <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> judmc123@bellsouth.net "Judson McClendon" wrote: > Here is an example of how C became popularized. In the 70's, DEC was > seriously dominant in the university environment. Because Bell Labs had > written C for the PDP, most of the system software for those computers > was in C. Students who wanted to seriously get into the system software I find it a bit hard to believe that the UK was a decade behind, and if my memory serves correctly the 70's was a period when all sorts of C predecessors like BCPL were prevalent in universities. C really started to get going in the early 80's when critical mass was achieved and compilers were ported to newly arrived PC's. It may have been a year or two earlier in the US, but a decade ???? Regards, -- Martin Brown <martin@nezumi.demon.co.uk> __ CIS: 71651,470 Scientific Software Consultancy /^,,)__/
From: "Charles W. Swiger" <chuck@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: Docs Date: 7 Jul 1998 21:15:01 GMT Organization: Spacelab.net Internet Access Message-ID: <6nu34l$hmp$1@news.spacelab.net> References: <35A25FB5.513AC61C@dev.tivoli.com> Byron Goodman <bgoodman@dev.tivoli.com> wrote: >Are there any NeXT or OpenStep Mach developer documention available? I'm >needing to find out about system calls that query hardware and devices. >I found the Rhapsody Operating System Software Guide, but has very >limited information. The documentation I acquired from CMU I have found >is not completely consistent with NeXT/OpenStep. Could you provide more information about what you're trying to do? Your question is so broad that you could be directed towards anything from the details of Mach device drivers, to generic Unix /dev/ stuff, to docs on SCSI, etc.... -Chuck Charles Swiger | chuck@codefab.com | standard disclaimer ---------------+-------------------+-------------------- "Microsoft: we make the easy almost impossible."
From: Tim McDermott <mcdermott@draper.com> Newsgroups: comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming Subject: Re: Programming and the colossal failure to advance Date: Tue, 07 Jul 1998 18:32:26 -0400 Organization: CSDL-DC Message-ID: <35A2A1F9.F0FFF7B5@draper.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> <vhIsdqY67dTD-pn2-dGSKqdjMrSfd@dynamic47.pm03.san-rafael.best.com> <zVpo1.206$Od.720988@news4.mia.bellsouth.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Judson McClendon wrote: > Yes, I've heard of those languages. I have done production programming > programming in over 40 languages and dialects. Between five and six > million lines of COBOL, and something over a million lines of various other > languages, since 1968, many of them you've never heard of, because I wrote > some of the compilers myself. So here is a perfect example of what on Bentley called "back-of-the-envelope" calculations in _Programming Pearls_. 30 years X 2000 hour/year = 60,000 hours 6,000,000 lines of COBOL, et al / 60,000 hours = 100 lines per hour It's a wonder Judson has time to cruise the news groups. Tim
From: cbbrowne@news.hex.net (Christopher Browne) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 7 Jul 1998 23:36:33 GMT Organization: Hex.Net Superhighway, DFW Metroplex 817-329-3182 Message-ID: <6nube1$c8n$6@blue.hex.net> References: <6nu3gi$97i$1@nnrp1.dejanews.com> On Tue, 07 Jul 1998 21:21:22 GMT, tj_sam@my-dejanews.com <tj_sam@my-dejanews.com> wrote: >On the assumption that a major advance to programming would >represent a overall simplification, I would agree that there has >been a colossal failure to advance programming. I would speculate >that a possible reason for this failure is due to the fact that we >have reached the practical limits for pure semantic programming. >There is an intrinsic level of noise to semantics definitions which >is required in human languages, however this noise does creates >problems and limits when applied to computer programming. The most >basic effect of reaching this "semantic noise limit" would be that >any attempts to advance programming would simply lead to an >increased complexity rather than a simplification. > >A reoccurring theme in the replies to this thread has been the >suggestion to move programming to more visual or graphic >presentations. Generalized, this represents a trend (or at least a >wishlist) towards a type of virtual reality programming. The >possibility of actual programming in a VR environment would >constitute an advancement, however it remains fiction. The question >I've often asked myself is: Why not? The answers are likely the same as those to the question: "Why do engineers still write out equations rather than ``simply'' drawing the relationships?" The written language of mathematics has been, and still is, symbolic equations as opposed to graphical pictures. There may be instances in which visual representations provide insight; the common means of representing mathematical concepts nonetheless still remains that of symbolic equations. I would suggest that in the same way, the notion of some sort of written language is critical to computing, and that while visual representations can provide some insight, they are not so fundamentally important as the symbolic representations. I'm not merely waving my hands furiously trying to suggest some vague notion of a connection between computing and mathematics; the base theories surrounding computation involve such things as: - Lambda calculus - Turing machines - The Church/Turing thesis and these three things represent concepts that are both mathematical as well as being fundamentally symbolic in nature. I would argue that the common conception that computers work with "numbers" is incorrect, and that they are fundamentally *symbol* processors. Symbol systems commonly represent numerical systems; it is nonetheless a better understanding to treat the numerical systems used by computers as systematic sets of symbols as the symbol sets do not generally correspond to any of the sets associated with conventional mathematical constructs. For instance, an "int" or "float" in C does not fully correspond to an integer or a real number. This digression hopefully underlines the fact that the way computers behave can only be predicted via symbolic systems. It takes a *tremendous* amount of work to make them do anything "visual;" the limited "virtual reality" capabilities at present may increase, but still will not soon be the "essence" of computing, but merely an *application* of computing. The failure of the Japanese "Fifth Generation" project to produce results that people would recognize as success further displays that there is no "silver bullet" in terms of having automated symbol processing recursively improve computing in a fundamental way. The end result of that project was a "better Prolog" and a generation of educated Japanese computer scientists; they *didn't* come up with a fundamentally different understanding of computing. "Virtual Reality" is similarly unlikely to provide further insight, as computing is fundamentally not unlike mathematics in the way its development has proceeded. VR still leaves you with the need to determine an appropriate set of symbols and ways to manipulate those symbols. -- "Problem solving under linux has never been the circus that it is under AIX." (By Pete Ehlke in comp.unix.aix) cbbrowne@ntlug.org- <http//www.hex.net/~cbbrowne/lsf.html>
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <07071998230151wthock@pop.jaring.my> Control: cancel <07071998230151wthock@pop.jaring.my> Date: 08 Jul 1998 00:46:37 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.07071998230151wthock@pop.jaring.my> Sender: <wthock@pop.jaring.my> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: dd@dandrake.com (Dan Drake) Newsgroups: comp.lang.misc,comp.software-eng,comp.programming,comp.sys.next.programmer Subject: Re: Programming and the colossal failure to advance Date: 8 Jul 1998 00:53:45 GMT Message-ID: <vhIsdqY67dTD-pn2-crqqitqHey3d@dynamic59.pm02.san-rafael.best.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> <vhIsdqY67dTD-pn2-dGSKqdjMrSfd@dynamic47.pm03.san-rafael.best.com> <bksEvp5Jy.945@netcom.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 8bit On Mon, 6 Jul 1998 23:35:58, bks@netcom.com (Bradley K. Sherman) wrote: >... > > Even at Berkeley in that period Pascal was the choice for undergraduate > work. Thanks for confirming this. >... > There was a BDS C > compiler that sacrificed some parts of the language but was really, > really fast under CP/M. BDS C! My life would have been very different without that tool, and not better. > C took over because you could get work done > and you had real reuse of the code when porting. Exactly my experience. In all honesty > I would say that C is still, in some sense, more portable than Java. > Wow. Not sure I'd go that far, but what the heck...... -- Dan Drake dd@dandrake.com http://www.dandrake.com/index.html Pedant: A man who likes what he says to be true. -- Bertrand Russell
From: dd@dandrake.com (Dan Drake) Newsgroups: comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming Subject: Re: Programming and the colossal failure to advance Date: 8 Jul 1998 00:56:04 GMT Message-ID: <vhIsdqY67dTD-pn2-LVFwXiy7k1JR@dynamic59.pm02.san-rafael.best.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> <6ns598$il4$1@murdoch.acc.Virginia.EDU> <Snpo1.200$Od.706436@news4.mia.bellsouth.net> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 8bit On Tue, 7 Jul 1998 13:28:18, "Judson McClendon" <judmc123@bellsouth.net> wrote: > Steven D. Majewski wrote: >... > > It's been interesting to read all these comments to my post, particularly > those about people who were around before C, because I started programming > in 1968. :-) Congratulations. Only 4 or 5 years after me. > Anyway, it is clear that many of the respondents here have > yet to learn the lesson that just because it is in books, and is being > taught at school, doesn't mean it is correct! An amazing statement, considering the number of people who have been describing their own experience! Most of what I say in that > post are things I personally was witness to, or was fairly close to. >... Different planet, I guess. -- Dan Drake dd@dandrake.com http://www.dandrake.com/index.html
From: "Judson McClendon" <judmc123@bellsouth.net> Newsgroups: comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> <vhIsdqY67dTD-pn2-dGSKqdjMrSfd@dynamic47.pm03.san-rafael.best.com> <zVpo1.206$Od.720988@news4.mia.bellsouth.net> <35A2A1F9.F0FFF7B5@draper.com> Subject: Re: Programming and the colossal failure to advance Organization: Sun Valley Systems Message-ID: <Qyzo1.2514$JB5.1181286@news1.atl.bellsouth.net> Date: Wed, 08 Jul 1998 01:02:40 GMT NNTP-Posting-Date: Tue, 07 Jul 1998 21:02:40 EST Tim McDermott wrote in message <35A2A1F9.F0FFF7B5@draper.com>... >Judson McClendon wrote: >> Yes, I've heard of those languages. I have done production programming >> programming in over 40 languages and dialects. Between five and six >> million lines of COBOL, and something over a million lines of various other >> languages, since 1968, many of them you've never heard of, because I wrote >> some of the compilers myself. > >So here is a perfect example of what on Bentley called "back-of-the-envelope" >calculations in _Programming Pearls_. > >30 years X 2000 hour/year = 60,000 hours > >6,000,000 lines of COBOL, et al / 60,000 hours = 100 lines per hour > >It's a wonder Judson has time to cruise the news groups. That just goes to show that you don't necessarily need OO to get a high percentage of code reuse. I write a large number of online programs for mainframes, many of which are over 50,000 lines each. I have developed a number of techniques for code reuse and automated code generation for certain types of routines. When I am programming heavily, I do well over 100 lines/hour; often several thousand lines/hour. -- Judson McClendon This is a faithful saying and worthy of all Sun Valley Systems acceptance, that Christ Jesus came into the judmc123@bellsouth.net world to save sinners (1 Timothy 1:15) (please remove numbers from email id to respond)
From: "Brian lLuethke" <Luethke@worldnet.att.net> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Tue, 7 Jul 1998 21:58:52 -0400 Organization: AT&T WorldNet Services Message-ID: <6nujo7$3vm@bgtnsc02.worldnet.att.net> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> Judson McClendon wrote in message ... >First, I wrote my post hastily and did not clearly articulate what I >meant to say. I used the word 'university', when I should have said >'academia', meaning the broader base of research and theoretical folks, >such as Xerox PARC, not just universities. Sorry. >-- >Judson McClendon This is a faithful saying and worthy of all >Sun Valley Systems acceptance, that Christ Jesus came into the >judmc123@bellsouth.net world to save sinners (1 Timothy 1:15) >(please remove numbers from email id to respond) > > > I would like to point out that most of the time. anyone coming up with a new language is doing RESEARCH and could be classified as researchers, therefore academia. Most people don't invent languages in their spare time,there may be a few, I can't say that for sure ( correct me if am wrong ). Many teachers and researchers spent a lot of time "in the trenches". My two cents worth Luethke@worldnet.att.net
From: wyoinst@trib.com Newsgroups: comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer Subject: $$$DOUBLE YOUR GAS MILEAGE AND PAY 1/2 PRICE FOR GASOLINE!$$$ Date: 8 Jul 1998 05:49:41 GMT Organization: Wyoming Instruments Message-ID: <13094272820068864@trib.com> Hi! I posted this using an unregistered copy of Newsgroup AutoPoster PRO! You can download your own copy for FREE from: http://www.autoposter.cc or just click this line: http://www.autoposter.cc/files/newspro1.exe --- ============================================================== Pay 1/2 Price For Your Gasoline?.. Sounds Impossible?.. Well, Read On.. =============================================================== Introducing the invention that has astounded the "Experts". The Fuel Atomizer 2000, introduced to the market in 1991 by Wyoming Instruments (Glenrock, Wyoming), has actually PRODUCED GAS MILEAGE INCREASES AS HIGH AS 100% on some vehicles. It has been read about in such magazines as Popular Science and Popular Mechanics. Kit Carson, of the Rush Limbaugh program says, "The Fuel Atomizer 2000 is a very fascinating invention." Dr. Graber, who's car doubled in gas mileage says, "It is just like paying half price at the pump." The Fuel Atomizer 2000 offers your vehicle some if not all the following benefits: 1) INCREASED GAS MILEAGE 2) INCREASED POWER AND PERFORMANCE 3) INCREASED LIFE SPAN OF YOUR EXAUST SYSTEM 4) DECREASED TAILPIPE EMMISSIONS (Independent testing shows hydrocarbons and carbon monoxide levels reduced to near zero on new cars!) 5) DECREASED ENGINE WEAR AND TEAR BEST OF ALL...The Fuel Atomizer 2000 costs less than $150.00 and comes with an incredible 60 DAY MONEY BACK GUARANTEE IF YOU ARE NOT COMPLETELY SATISFIED! =================================================================== DISTRIBUTORSHIPS AVAILABLE =================================================================== Now you can make money showing people how to save on their gasoline expenses! Start on a part time or full time basis and watch your income soar! For more information, or to place an order, please call us at (307) 436-2601, or visit us on the web by using your mouse to double click on the web site address listed here: www.trib.com/WYOINST/ For God and Country, R.C. Davis, CEO
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer Subject: cmsg cancel <13094272820068864@trib.com> Control: cancel <13094272820068864@trib.com> Date: 08 Jul 1998 05:51:25 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.13094272820068864@trib.com> Sender: wyoinst@trib.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: nigel@access4.digex.net (Nigel Tzeng) Newsgroups: comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming Subject: Re: Programming and the colossal failure to advance Organization: Express Access Online Communications, Greenbelt, MD USA Message-ID: <6nuqn3$6a6@access4.digex.net> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <vhIsdqY67dTD-pn2-YqJWEd2Zw85n@dynamic47.pm03.san-rafael.best.com> Date: Wed, 08 Jul 1998 03:56:00 GMT NNTP-Posting-Date: Tue, 07 Jul 1998 23:56:00 EDT In article <vhIsdqY67dTD-pn2-YqJWEd2Zw85n@dynamic47.pm03.san-rafael.best.com>, Dan Drake <dd@dandrake.com> wrote: >On Thu, 2 Jul 1998 13:48:58, boracay@hotmail.com wrote: [snip] >GUI code -- which would be harder to write and harder to debug than the >algorithms ever were. That's why it looks as if there has been no, or >negative, progress in making programming more efficient. Er...no one really writes GUI code anymore...at least not like we used to. Coding a cool looking GUI is relativly trivial affair (in a relative sense) given VB, VC++, UIM/X, Visual Cafe, etc, etc... As a GUI developer it's been a looooooong long time since I've touched raw X code and I've never touched any real windows GUI code (lay it out and write the callbacks...done!). As close as I've come is some Tcl code or the code to generate HTML. That's not quite as horrible as what we used to do a decade ago. >Dan Drake Nigel
From: rex@smallandmighty.com (Eric King) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: Wed, 08 Jul 1998 02:30:17 -0400 Organization: The Small & Mighty Group Message-ID: <rex-0807980230180001@cc497300-a.wlgrv1.pa.home.com> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> In article <6nube1$c8n$6@blue.hex.net>, cbbrowne@hex.net wrote: :The answers are likely the same as those to the question: : :"Why do engineers still write out equations rather than ``simply'' :drawing the relationships?" I'd argue that when they write out an equation that they *are* drawing the relationships. A better question might be: why do programmers still draw flowcharts rather than simply writing out a description of their programs? :The written language of mathematics has been, and still is, symbolic :equations as opposed to graphical pictures. But aren't those symbols just groups of small graphical pictures? I think you're making a very artificial distinction. An 'O' could be the representation of a letter in the English alphabet or it could just be a small picture of a circular object. :There may be instances in which visual representations provide insight; I'll go one step further and say there are instances where a graphical representation makes a lot more sense than a textual representation. Not all computing tasks can be conveniently represented by a linear stream of text. There is a huge need for tools that help programmers manage the complexity of their projects. Class browsers, GUI builders, flowcharting tools, etc. are all attempts to augment textual representations of programs. The sheer number of those types of tools leads me to question whether a purely textual representation is really appropriate. IMO, the (distant) future of programming will be based upon languages like Prograph, which mix graphical and textual representations of programming constructs. Purely textual programming languages are trying to encode a vast quantity of information using a very limited set of symbols and layout rules. This results in a steady stream of new programming languages which really aren't any more expressive than preexisting languages. Furthermore, a lot of effort is being spent developing large libraries, frameworks, toolkits, etc. which try to anticipate what one might need to do in a program. IMO, more effort needs to be spent on trying to help one manage the complexity of what one is trying to do. ::Eric
From: Andrey V Khavryutchenko <akhavr@compchem.kiev.ua> Newsgroups: comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming Subject: Re: Programming and the colossal failure to advance Date: 08 Jul 1998 11:40:18 +0300 Organization: Computational Chemistry Group Sender: akhavr@netmaster.compchem.kiev.ua Message-ID: <x71zrw3fd9.fsf@netmaster.compchem.kiev.ua> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> <vhIsdqY67dTD-pn2-dGSKqdjMrSfd@dynamic47.pm03.san-rafael.best.com> <zVpo1.206$Od.720988@news4.mia.bellsouth.net> <35A2A1F9.F0FFF7B5@draper.com> <Qyzo1.2514$JB5.1181286@news1.atl.bellsouth.net> Mime-Version: 1.0 (generated by tm-edit 7.108) Content-Type: text/plain; charset=US-ASCII >>>>> "JM" == Judson McClendon writes: JM> That just goes to show that you don't necessarily need OO to get a high JM> percentage of code reuse. I write a large number of online programs JM> for mainframes, many of which are over 50,000 lines each. I have JM> developed a number of techniques for code reuse and automated code JM> generation for certain types of routines. When I am programming JM> heavily, I do well over 100 lines/hour; often several thousand JM> lines/hour. You count lines in a very strange manner. Size metrics may have different goals, but if you want measure the maintability (which is seriously impacted by size), then you have to count artifacts, that you have to maintain. If you use LOC, then count lines of code, that _you_ wrote. Function call is 1 line, not those 100 lines in a function. For automated code generation, you would count lines of specification code, not the lines of generated code. Well, may be I'm wrong and you use LOC for other purposes, I can't think of. I hope that it's not to impress your customer (wow, we've got 10'000'000 bits program) or usenet community (which is rather immune). -- SY, Andrey V Khavryutchenko http://www.kbi.kiev.ua/~akhavr Good judgment comes from experience, and a lot of that comes from bad judgment.
Message-ID: <35A22694.3512D129@planetall.com> Date: Tue, 07 Jul 1998 13:45:56 +0000 From: Mike Percy <mpercy@planetall.com> Organization: Scientific Research Corp. MIME-Version: 1.0 Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6ng8c3$vb52@onews.collins.rockwell.com> <6nh8pk$82k$1@murdoch.acc.Virginia.EDU> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Steven D. Majewski wrote: > In article <6ng8c3$vb52@onews.collins.rockwell.com>, > Erik M. Buck <embuck@palmer.cca.rockwell.com> wrote: > >This is just silly. > > > >A computer scientist does not expect to understand geology, genetics, > >surgery, particle physics, systems engineering, or even ditch digging to any > >greater degree than the average population and certainly not to any degree of > >professional skill. Why would a geologist, surgeon, physicist, systems > >engineer, or ditch digger expect to know enough about computer science to > >exhibit professional skill in computer science ? > > > >It is condescending in the extreme to claim the computer science should be > >easy enough for anyone. This is as large and hard a proffesion as any other > >and people do not expect geology to be instantly understood by every novice > >that comes along. > > > > But there's a difference between computer science and programming. > One of the goals of computer science is to make programming easier. > The difference is that someone who took a "Intro to Programming in C" class once can get a job as a programmer and be woefully unprepared for the reality of most such positions. I took multiple courses in physics in high-school and college, but probably couldn't get a job as a physicists assistant! The premise that a few programming classes form the basis of being a "software engineer" or "computer scientist is fundamentally flawed. > If it is exactly as difficult to teach programming to geologists > or teach geology to programmers, then we will never get many good > geology application programs. One of them has got to become easier. What I was taught to do was effectively communicate with domain experts, like the radar engineers I currently work with, like the people at the Department of Corrections whose "offender management" systems I helped design and code, like the HR people I helped design & implement personnel databases, etc. My point is that I don't need to become an expert in geology or biology, that's what the expert is for. My skill is in extracting, understanding, and codifying information from the experts such that I can construct (correctly and skillfully) a software system that meets their needs. > I expect the number of facts and techniques needed to master both > geology and computer science to grow. However, if computer scientists > do their job, programming should get easier and simpler. And in fact, > it you hold constant the complexity of the job you're trying to > accomplish, and you avoid poorly designed systems ( like C++ and MFC ), > programming has in fact gotten easier. > > ---| Steven D. Majewski (804-982-0831) <sdm7g@Virginia.EDU> |--- > ---| Department of Molecular Physiology and Biological Physics |--- > ---| University of Virginia Health Sciences Center |--- > ---| P.O. Box 10011 Charlottesville, VA 22906-0011 |--- Programming itself is not particularly hard -- if it were we wouldn't have the problems we have now. Skillfully/professionally designing and implementing software systems that correctly implement the requirements is the hard part. I can build a deck or a dog-house, but would you want me to build your house without some training *and* experience? Why, then, does anyone think it reasonable to let unskilled, inexperienced people construct your software systems?
From: Jim Webber <james.webber@ncl.ac.uk> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Re:Programming and the colossal failure to advance - VR Date: Wed, 8 Jul 1998 15:08:48 +0100 Organization: University of Newcastle upon Tyne Message-ID: <6nvuhh$jl8$1@ucsnew1.ncl.ac.uk> References: <6nu3gi$97i$1@nnrp1.dejanews.com> I think that you could well be correct in your assumption that a graphical programming paradigm could be the next stage in the development of software engineering. However, most visual programming techniques (my own included) are relatively immature, and the software industry as a whole takes little notice of them. For proof you just need to look at the volume of traffic through say comp.lang.c++ compared to comp.lang.visual. I think the jump from textual to hybrid visual-textual programming (not meaning VB or similar in this sense) is going to be significantly slower than the jump from say assembly to high level languages (which also seemed to be resisted (though it was a little before my time)). However that's not to say that we can't take paradigms with us when we move from one programming language type to another, and I still think that it's the paradigms that count, after all programming is largely arbitrary :) The reason for going to visual languages would be that we have some particularly complicated hardware system to program (parallel and distributed come to mind) which we wish to abstract. I think that VLs do this rather better than TLs. Jim - http://painshaw.ncl.ac.uk/
Message-ID: <35A35D80.1A7ADD1C@cadvision.com> Date: Wed, 08 Jul 1998 11:52:32 +0000 From: Christopher Clark <compudata@cadvision.com> Organization: Compudata Systems Consultin, Inc. MIME-Version: 1.0 Newsgroups: comp.lang.misc,comp.software-eng,comp.programming,comp.sys.next.programmer Subject: Re: Programming and the colossal failure to advance References: <6ng34a$j2g$1@nnrp1.dejanews.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> <vhIsdqY67dTD-pn2-dGSKqdjMrSfd@dynamic47.pm03.san-rafael.best.com> <bksEvp5Jy.945@netcom.com> <vhIsdqY67dTD-pn2-crqqitqHey3d@dynamic59.pm02.san-rafael.best.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Dan Drake wrote: > On Mon, 6 Jul 1998 23:35:58, bks@netcom.com (Bradley K. Sherman) > wrote: > > In all honesty I would say that C is still, in some sense, more > > portable than Java. > Wow. Not sure I'd go that far, but what the heck...... I like the way it was put the the Perl language reference (I don't have my copy handy, so I'm paraphrasing)... "We programmed perl in C not because it's portable, but because it's ubiquitous. Standard C is about as portable as Chuck Yaeger on foot." (For the context-impaired: Chuck Yaeger was the pilot who first broke the sound barrier.) IMO, Java (big J), the language, is more portable than C. java (small j), the veritable cornucopia of interpreters, are still horrendously incompatible. C is becoming more portable (look at the <inttypes.h> (I think) header in the new draft standard for an example), but backwards combatibility, and the cost of stripping out all the temporary solutions that have been knocked together over the years, is going to impede it. The java interpreters should become compatible soon, just because of the sheer amount of code being written. If you're not compatible, you won't get used. Period. Even if you do integrate it into the operating system. Not that I'm naming names. Of course, if programmers continue to hard-code things like directory separators, strings, etc., this is all moot anyways. -- @>-`--,-- Chris Clark kitanin@hotmail.com Compudata Systems Consulting, Inc. compudata@cadvision.com ICQ#9342023 (eventually...) "So those lizard things...they were demons?" "Unmistakably. Although I didn't know they came in pink."
From: dd@dandrake.com (Dan Drake) Newsgroups: comp.lang.misc,comp.software-eng,comp.programming,comp.sys.next.programmer Subject: Re: Programming and the colossal failure to advance Date: 8 Jul 1998 16:23:30 GMT Message-ID: <vhIsdqY67dTD-pn2-Plvn6jluOldP@dynamic33.pm02.san-rafael.best.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <vhIsdqY67dTD-pn2-YqJWEd2Zw85n@dynamic47.pm03.san-rafael.best.com> <6nuqn3$6a6@access4.digex.net> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 8bit On Wed, 8 Jul 1998 03:56:00, nigel@access4.digex.net (Nigel Tzeng) wrote: > In article <vhIsdqY67dTD-pn2-YqJWEd2Zw85n@dynamic47.pm03.san-rafael.best.com>, > Dan Drake <dd@dandrake.com> wrote: > >On Thu, 2 Jul 1998 13:48:58, boracay@hotmail.com wrote: > > [snip] > > >GUI code -- which would be harder to write and harder to debug than the > >algorithms ever were. That's why it looks as if there has been no, or > >negative, progress in making programming more efficient. > > Er...no one really writes GUI code anymore...at least not like we used > to. Coding a cool looking GUI is relativly trivial affair (in a > relative sense) given VB, VC++, UIM/X, Visual Cafe, etc, etc... > >... True enough; my comment on non-progress is a bit out of date. But I'm not sure it's really out of date to put a little question mark next to the "no one". -- Dan Drake dd@dandrake.com http://www.dandrake.com/index.html
From: dd@dandrake.com (Dan Drake) Newsgroups: comp.software-eng,comp.programming,comp.sys.next.programmer,comp.lang.misc Subject: Re: Programming and the colossal failure to advance Date: 8 Jul 1998 16:32:16 GMT Message-ID: <vhIsdqY67dTD-pn2-dlRGwXS1mmkf@dynamic33.pm02.san-rafael.best.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> <vhIsdqY67dTD-pn2-dGSKqdjMrSfd@dynamic47.pm03.san-rafael.best.com> <bksEvp5Jy.945@netcom.com> <vhIsdqY67dTD-pn2-crqqitqHey3d@dynamic59.pm02.san-rafael.best.com> <35A35D80.1A7ADD1C@cadvision.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 8bit Did a previous non-reply of mine make out to the net? If so, sorry. On Wed, 8 Jul 1998 11:52:32, Christopher Clark <compudata@cadvision.com> wrote: >... > > IMO, Java (big J), the language, is more portable than C. java (small > j), the veritable cornucopia of interpreters, are still horrendously > incompatible. Well said. > ... > > The java interpreters should become compatible soon, just because of > the sheer amount of code being written. If you're not compatible, you > won't get used. Period. Even if you do integrate it into the operating > system. Not that I'm naming names. > > Of course, if programmers continue to hard-code things like directory > separators, strings, etc., this is all moot anyways. You hit a nerve there. What you say about hard-coded stuff is absolutely true; but the File class, which is supposed to do away with that, is one of most appallingly bad jobs in capital-J Java. I mean, it officially lies about whether a name is absolute when run on any DOS-descended system, which means most of the computers in the world. Phaugh. I still love Java though; the mark of greatness in both Java and C may be the intensity of the love-hate relation that so many of us have with them. -- Dan Drake dd@dandrake.com http://www.dandrake.com/index.html A hard heart is no infallible protection against a soft head. --C. S. Lewis
From: JGARBUZ@postoffice.worldnet.att.net Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Sat, 04 Jul 1998 17:44:04 -0700 Organization: AT&T WorldNet Services Message-ID: <6o08ei$6sr@bgtnsc03.worldnet.att.net> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6ng8c3$vb52@onews.collins.rockwell.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Erik M. Buck wrote: > > This is just silly. > > A computer scientist does not expect to understand geology, genetics, > surgery, particle physics, systems engineering, or even ditch digging to any > greater degree than the average population and certainly not to any degree of > professional skill. Why would a geologist, surgeon, physicist, systems > engineer, or ditch digger expect to know enough about computer science to > exhibit professional skill in computer science ? > > People are tantalized by how easy and powerful programming seems at first. > It is not a problem that computers are too hard; it is that they are > approachable. Almost anyone can learn intro programming. Why would that > person expect to be any more capable as a computer scientist than a computer > scientist would be at geology after an intro geology course.< Programming is problem solving. Good programmers enjoy solving problems. Adapting the solution to a particular computer system is the art of efficient coding, and knowing how to apply the best available tools is a major part of the game. A good mechanic is only as good as his tools, but tools are no substitute for a brilliant, problem-solving mind.
From: Matthias Fischmann <fis@mpi-sb.mpg.de> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Followup-To: comp.lang.misc Date: 08 Jul 1998 20:09:17 +0200 Organization: CIP-Pool der Philosophischen Fakultaet der Universitaet des Saarlandes Message-ID: <y9iemvw5i5u.fsf@warren.mpi-sb.mpg.de> References: <6nu3gi$97i$1@nnrp1.dejanews.com> tj_sam@my-dejanews.com writes: > A basic VR program editor would essential be a flowcharting program > with the actual code embedded within the flowchart (not the simplex > flowchart but actual source level flowcharting). A multiple column > text editor in which major branches and loop would be directly > represented by visual vectors. Individual subroutines or functions > would be represented as 2D (X,Y) wire frame objects. If subroutine > or function calls are represented as Z plane bi-directional vector > (recursive function are outlawed) then modules and complex programs > can be represented as Virtual 3D wireframe objects. This seems a little vague to me. What exactly do those objects look like? At some level of abstraction, you just don't have any other tools than language to describe what you need from your virtual environment. And turing complete programming languages, or the tasks they are designed to handle *are* abstract, and have to be. It may be possible to *use* specific programs in cyberspace. I can think of designing molecules, taxi scheduling, or many other tasks. But if you try to *describe* a completely new task, you fall into the limitations of a predefined 3D environment very quickly. Given a language without watchdogs already provided by some library, define an object that dials a phone number if my pc crashes. curious - Matthias -- Max-Planck-Institut für Informatik | Deutsches Forschungszentrum für KI fis@mpi-sb.mpg.de | fischman@dfki.de http://www.mpi-sb.mpg.de/~fis |
From: Dan Wellman <wellman@students.uiuc.edu> Newsgroups: comp.lang.objective-c,comp.sys.next.programmer Subject: Generating Procedure Execution Log Date: 8 Jul 1998 20:57:36 GMT Organization: University of Illinois at Urbana-Champaign Message-ID: <6o0mg0$beu$1@vixen.cso.uiuc.edu> I'm running OpenStep 4.2 for Windows NT and am looking for a way to generate a text file/log that shows all the functions that are called as they are called.. sort of like a procedure history log for an application's lifetime. Does such a program exist, and if so, where? Thanks in advance for your help. dan -- Dan Wellman <> wellman@uiuc.edu <> http://www.cen.uiuc.edu/~wellman/ "A million thoughts in one night can't be wrong" - Cause & Effect
From: nreed@indiana.edu Newsgroups: comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming Subject: Re: Programming and the colossal failure to advance Date: Wed, 08 Jul 1998 20:48:33 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6o0lv1$47o$1@nnrp1.dejanews.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> <vhIsdqY67dTD-pn2-dGSKqdjMrSfd@dynamic47.pm03.san-rafael.best.com> <bksEvp5Jy.945@netcom.com> In article <bksEvp5Jy.945@netcom.com>, bks@netcom.com (Bradley K. Sherman) wrote: > In all honesty > I would say that C is still, in some sense, more portable than Java. C more portable than Java? That's a good one. (You're joking, right?) -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: hmmueller@my-dejanews.com Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Wed, 08 Jul 1998 20:53:37 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6o0m8h$4id$1@nnrp1.dejanews.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359BB163.7A97DEC6@gssec.bt.co.uk> <6nh07l$hl$1@nnrp1.dejanews.com> <359C25D4.7B2A@computer.org> <6nmgrl$tfp$1@nnrp1.dejanews.com> <35A2220E.67A2E5F@pgh.net> In article <35A2220E.67A2E5F@pgh.net>, ghristwd@pgh.net wrote: > hmmueller@my-dejanews.com wrote: > > [...] > > > > In the one (university) biology lab where I had friends, they had a lab > > technician. If they needed a new cage; or a special machine to to do an > > experiment with chimpanzees (3 switches, a few food outlets and two lamps, in > > one case), it was clear that the design and building of the machine was NOT > > the biologist's job; he would put a sketch on paper and talk with the > > technician who would then design and build the machine; in the case of the > > cage, he would usually only do the design, and the cage would be built by > > some welding company or so. All that is obvious. > > > > So why was it assumed that the same biologist would write the program for > > controlling the lamps and food outlets and measuring the timing of the > > chimpanzees' responses himself (where he failed miserably)? > > Or that me (the programmer coming in afterwards) would have to understand > > biology to write a small real-time program? > > > > Neither has the biologist to learn programming, nor the programmer biology. > > They just have to talk to each other. It works; and is almost always much > > cheaper, and more more certainly gets the job done. > > > > And again: If someone has all the required expertise, from biology to [...] > > But the example you have given is not a biology problem, it is a control > system problem. I am not surprised that the technician is better than > the biologist at programming a control system. But if the biologist > needs a program that would take the results of the experiments and > perform some analysis on them according to the principles of biology > (don't ask me for an example, I'm not a biologist) then I would expect > the biologist to be better at that (after some education in programming, > of course). > > Regards, > Bill Ghrist > D'accord. Better example (but never finished): A friend, who was (is?) in ???cat breeding??? - where the "best" cats get awards if they adhere to some "beauty standard", based on race [sorry - I do not know the English expressions ...; hopefully you get the idea]; anyway, we wanted to write a program that would propose certain matings, based on the genotype of the available breeding cats, which would have to be "re-engineered" (computed) from the phenotypes of a cat and its relatives. The necessary biology/cat-breeding knowledge was standard "cat gene knowledge" and knowledge about basic genetic inheritance. It turned out that the program (which would have to do some dynamic programming and/or recursive search with probabilities) could only be written by both of us - almost every biologist (here I generalize from "my" cat breeder) would not even know about e.g. dynamic programming. Making "programming" easier would not have helped (in the sense that the programmer/CStist wouls become superfluous) - if you understand dynamic programming [including when/when not to use it etc.], you ARE a [proably not too bad] SW engineer. So we HAD to do it together (because, of course, I would not know that the gene for "white" is stronger than almost any other color gene [except tabby??][?? if I remember correctly). Harald P.S. Which should not say that programming say of a simple statistic evaluation of a list of given cats should not be possible by my cat-friend (but preferably not thru "programming", but by "clicking on a few buttons"). P.P.S. As I said, we did not finish :( But I'd still like to know whether anyone has ever written a program like that ... I've not yet found out. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: NOSPAMmbkennelNOSPAM@yahoo.com (Matt Kennel) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 8 Jul 1998 23:54:28 GMT Organization: University of California at San Diego Message-ID: <slrn6q81lj.j00.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> <rex-0807980230180001@cc497300-a.wlgrv1.pa.home.com> On Wed, 08 Jul 1998 02:30:17 -0400, Eric King <rex@smallandmighty.com> wrote: ::The written language of mathematics has been, and still is, symbolic ::equations as opposed to graphical pictures. : : But aren't those symbols just groups of small graphical pictures? Yes. But they are a language in the strong Chomskian sense in the way that 'just pictures' are not. Humans have specific neurobiology for understanding and producing language. -- * Matthew B. Kennel/Institute for Nonlinear Science, UCSD - * "People who send spam to Emperor Cartagia... vanish! _They say_ that * there's a room where he has their heads, lined up in a row on a desk... * _They say_ that late at night, he goes there, and talks to them... _they *- say_ he asks them, 'Now tell me again, how _do_ you make money fast?'"
Newsgroups: comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming From: bks@netcom.com (Bradley K. Sherman) Subject: Re: Programming and the colossal failure to advance Message-ID: <bksEvsppn.Jq0@netcom.com> Organization: DNA + Sunlight References: <6ng34a$j2g$1@nnrp1.dejanews.com> <vhIsdqY67dTD-pn2-dGSKqdjMrSfd@dynamic47.pm03.san-rafael.best.com> <bksEvp5Jy.945@netcom.com> <6o0lv1$47o$1@nnrp1.dejanews.com> Date: Wed, 8 Jul 1998 21:44:11 GMT Sender: bks@netcom18.netcom.com In article <6o0lv1$47o$1@nnrp1.dejanews.com>, <nreed@indiana.edu> wrote: >> I would say that C is still, in some sense, more portable than Java. > >C more portable than Java? That's a good one. (You're joking, right?) > In the sense that C is more stable. Java, in principle, is more portable, but in practice there are problems. I am not bashing Java here, so don't take this wrong. If there are not problems with Java portability, what is Sun's gripe with MicroSoft? --bks
From: zegelin@actonline.com.au (Peter) Newsgroups: comp.sys.next.programmer Subject: URL for Draw.App Date: Thu, 09 Jul 1998 13:00:36 +1000 Organization: Totally Disorganized Message-ID: <zegelin-0907981300370001@dialup232.canberra.net.au> Cache-Post-Path: genie.cyberone.com.au!unknown@dialup232.canberra.net.au Hi there! I cant seem to locate the source code for Draw.App at Apples' web site. Even their search engine cant help. Does someone know the URL? thanks! Peter
From: scott_sa@my-dejanews.com Newsgroups: comp.sys.mac.programmer.misc,comp.sys.next.programmer Subject: Re: Thoughts on Realbasic Date: Thu, 09 Jul 1998 04:52:33 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6o1iah$fst$1@nnrp1.dejanews.com> References: <6nqtoo$n4s@drn.newsguy.com> In article <6nqtoo$n4s@drn.newsguy.com>, chris wrote: > > Real Software has just released their visual basic > look-a-like for the mac. The ide is 92% carbon > compatible. Has anyone had any experiences with > this ide? Yup, its sweet! Needs to mature a little but is quite functional for many things esp. learning right now. There are still some issues with regards to the 1.0 release (has not happened yet, MacWorld and all). I don't recommend commercial development for a while but I expect sometime in early August this will be resolved. In about 3 hours I re-built the interface (3 dilaogs, an about box, a tabbed window with four panes, 30-ish controls and two menus - all functioning ) of a simple app. I built a year ago. Compare that to about 12 hours! The OOP implementation is a little different with regards to overshadowing & polymorphism, but more than usable. Scott -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: kcd@babylon5.jumpgate.com (Kenneth C. Dyke) Newsgroups: comp.sys.next.programmer Subject: Re: Strange assembler-code? Date: 9 Jul 1998 05:27:51 GMT Message-ID: <6o1kcn$pg3$1@nntp2.ba.best.com> References: <35a0901c.0@uni-wuerzburg.de> In-Reply-To: <35a0901c.0@uni-wuerzburg.de> On 07/05/98, Sven Droll wrote: >Hello. > >I have one last problem for getting kaffe 0.10.1 to work correctly on black >NeXT (at least I hope so). > >There is a part written in machine code and kaffe relies heavily on this! But >it seems to me like a mix between motorola (surely not the NeXT/MIT) syntax >and intel-code (the % and %% are very nice, but were never found in any NeXT >or motorola-assembler-code). > >As mot2mit fails to convert the code, my question is if somebody could >translate the following 24 directives (especially the "%number" and the >"4(...)") or at least give me a hind what m68k-assembler this was written That code snippet is set up to be used with GNU C inline assembly. From what I can tell, it should 'just work' with the included NeXT compiler. The %0, %1, etc. refer to 'parameters' passed into the assembly code. These will be replaced by registers or memory locations by the C compiler depending on the restrictions placed on them. The %% is required so that the % can be escaped, as real CPU registers are referred to as %sp, %pc, etc. by the assembler. Here's a quick explanation of what the meanings of the paremeters are at the end of the asm section: " : /* Nothing here, so no outputs */ : "r" ((CALL)->nrargs), /* This is for %0, and means it must be a data register. */ "a" ((CALL)->args), /* This is for %1, and means it must be an address register. */ "a" ((CALL)->callsize), /* For %2 */ "a" ((CALL)->function), /* For %3 */ "m" ((CALL)->rettype), /* For %4, must be a memory location */ "a" ((CALL)->ret) /* For %5 */ Check out the GNU C compiler documentation for more detailed information. GNU inline assembly is very cool, but it can be a bit gnarly to use until you get used to it. -Ken -- Kenneth Dyke, kcd@jumpgate.com (personal), kdyke@ea.com (work) Nuclear Strike and OPENSTEP Tools Engineer, Electronic Arts C++: The power, elegance and simplicity of a hand grenade.
From: jmenna@malpractice.com Newsgroups: comp.sys.mac.programmer.misc,comp.sys.next.programmer Subject: Re: Thoughts on Realbasic Date: Thu, 09 Jul 1998 05:32:41 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6o1klp$jbg$1@nnrp1.dejanews.com> References: <6nqtoo$n4s@drn.newsguy.com> In article <6nqtoo$n4s@drn.newsguy.com>, chris wrote: > > Real Software has just released their visual basic > look-a-like for the mac. The ide is 92% carbon > compatible. Has anyone had any experiences with > this ideH Very slick interface on a solid product. Im not an MIS guy but I mannage to piss off the corpoate MIS guys a@lot when their VB apps are ported and look better.@iusing realbasic) The company is very responsive and the user base is@as dedicated as heck. You can download it and play with it fr free for 30(?) days. www.realsoftware.com -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: jmenna@malpractice.com Newsgroups: comp.sys.mac.programmer.misc,comp.sys.next.programmer Subject: Re: Thoughts on Realbasic Date: Thu, 09 Jul 1998 05:34:37 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6o1kpd$jbp$1@nnrp1.dejanews.com> References: <6nqtoo$n4s@drn.newsguy.com> <35a2092c.11569696@allnews.infi.net> In article <35a2092c.11569696@allnews.infi.net>, dolby@infi.net (Jack Dolby) wrote: > > Chris, I don't have the address here (at work on a non-mac system... > sorry) to subscribe, but there is a very active Realbasic mailing > list. there should be a link to it on the realbasic home page. > > I think the experts there can give you some realy good heads-up info > on realbasic. > Oh yea I forgot to mention. The people who do the macaddict Cd are tossing macromedia in favor of realbasic. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: rex@smallandmighty.com (Eric King) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: Thu, 09 Jul 1998 02:06:41 -0400 Organization: The Small & Mighty Group Message-ID: <rex-0907980206430001@cc497300-a.wlgrv1.pa.home.com> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> <rex-0807980230180001@cc497300-a.wlgrv1.pa.home.com> <slrn6q81lj.j00.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu> In article <slrn6q81lj.j00.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu>, mbkennel <replace this with '@'> yahoo.com wrote: :On Wed, 08 Jul 1998 02:30:17 -0400, Eric King <rex@smallandmighty.com> wrote: :::The written language of mathematics has been, and still is, symbolic :::equations as opposed to graphical pictures. :: :: But aren't those symbols just groups of small graphical pictures? : :Yes. But they are a language in the strong Chomskian sense in the way that :'just pictures' are not. I agree that a grouping of arbitrary pictures do not constitute a language, but why must programming languages be constrained to such a narrow set of symbols and layout rules? :Humans have specific neurobiology for understanding and producing language. Yes, and that neurobiology is capable of understanding and producing much more varied forms of language than the simple streams of text that make up most programs. ::Eric
From: rex@smallandmighty.com (Eric King) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: Thu, 09 Jul 1998 02:39:01 -0400 Organization: The Small & Mighty Group Message-ID: <rex-0907980239020001@cc497300-a.wlgrv1.pa.home.com> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> <rex-0807980230180001@cc497300-a.wlgrv1.pa.home.com> <slrn6q81lj.j00.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu> <james-ya023180000807982343080001@news> In article <james-ya023180000807982343080001@news>, james@unspam.edu (James McCartney) wrote: :I've yet to see a graphical language that could express something in less :space than a text equivalent. Why does it need to express something in less space than a text equivalent? Space is cheap, especially if you have robust browsing facilities. I will mention that in Prograph and probably other graphical languages, one can select an arbitrary group of icons and 'compress' it down to one. This just can't be done with most textual languages, and is probably the single feature I miss the most about the environment. The freedom to attach comments to any icon and have them follow that icon around is another wonderful feature. :Graphical languages I've seen are one of graphs with nodes and edges, (easy to translate to text) containment diagrams, (even easier -> trees) But is the text easier to understand and maintain than the graphical representation? Does it give you a better idea of the functionality? How many lines of a textual code can you actually keep in your head at once? I'm not advocating the abandonment of textual languages, but rather the augmentation of them. I think the baseline editing environment needs to be much more sophisticated than a text editor; it needs to know about manipulating graphics as well as text. Once such environments are ubiquitous, I think language implementors will start experimenting more with alternative syntax representations. That time's a long way off, though. :All of them can be translated pretty easily to a lisp like syntax. That's somewhat of a moot point. Programs can often be translated from one language to another. Will the Lisp-like syntax be easier to follow and edit than the graph? :Some advantages to graphical environments is that they can :provide default arguments, and popup help for their arguments and usage. :However this is not impossible for text environments, just not usually :implemented as well. Everthing that can be done in a textual environment can be done in a graphical environment. But not the reverse. ::Eric
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: URL for Draw.App Date: 9 Jul 1998 04:06:53 GMT Organization: Digital Fix Development Message-ID: <6o1fkt$8aq$1@news.digifix.com> References: <zegelin-0907981300370001@dialup232.canberra.net.au> In-Reply-To: <zegelin-0907981300370001@dialup232.canberra.net.au> On 07/08/98, Peter wrote: >Hi there! > > I cant seem to locate the source code for Draw.App at Apples' web site. >Even their search engine cant help. Does someone know the URL? > Its not on the ftp or web site as far as I know.. It does come on the OpenStep/Rhapsody Developer CDs though.. -- Scott Anguish <sanguish@digifix.com> Stepwise - OpenStep/Rhapsody Information <URL:http://www.stepwise.com>
From: james@unspam.edu (James McCartney) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: Wed, 08 Jul 1998 23:43:08 -0700 Organization: The University of Texas at Austin, Austin, Texas Message-ID: <james-ya023180000807982343080001@news> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> <rex-0807980230180001@cc497300-a.wlgrv1.pa.home.com> <slrn6q81lj.j00.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <slrn6q81lj.j00.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu>, mbkennel <replace this with '@'> yahoo.com wrote: >Yes. But they are a language in the strong Chomskian sense in the way that >'just pictures' are not. > >Humans have specific neurobiology for understanding and producing language. I try to convince people sometimes that text based programming actually is a graphical programming environment, and a very dense and sophisticated one. But they usually don't buy it. I've yet to see a graphical language that could express something in less space than a text equivalent. Graphical languages I've seen are one of graphs with nodes and edges, (easy to translate to text) containment diagrams, (even easier -> trees) tiles. ( just a variation of nodes and edges) But then I'm not really up on what else might be out there. (I'm not counting graphical languages whose end product is the graphics themselves.) All of them can be translated pretty easily to a lisp like syntax. I think graphical browsers that can display relationships in code are probably more beneficial and might be a good VR direction. Some advantages to graphical environments is that they can provide default arguments, and popup help for their arguments and usage. However this is not impossible for text environments, just not usually implemented as well. --- James McCartney
From: james@unspam.edu (James McCartney) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: Thu, 09 Jul 1998 02:52:11 -0700 Organization: The University of Texas at Austin, Austin, Texas Message-ID: <james-ya023180000907980252110001@news> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> <rex-0807980230180001@cc497300-a.wlgrv1.pa.home.com> <slrn6q81lj.j00.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu> <rex-0907980206430001@cc497300-a.wlgrv1.pa.home.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <rex-0907980206430001@cc497300-a.wlgrv1.pa.home.com>, rex@smallandmighty.com (Eric King) wrote: > I agree that a grouping of arbitrary pictures do not constitute a >language, but why must programming languages be constrained to such a >narrow set of symbols and layout rules? > Yes, and that neurobiology is capable of understanding and producing >much more varied forms of language than the simple streams of text that >make up most programs. Just as graphical languages use a very limited palette of symbols and graphical relations. The nature of the communication is a very specific one, not as general as natural language or art, and therefore uses a certain vocabulary and structure. --- James McCartney
From: james@unspam.edu (James McCartney) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: Thu, 09 Jul 1998 03:00:15 -0700 Organization: The University of Texas at Austin, Austin, Texas Message-ID: <james-ya023180000907980300150001@news> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> <rex-0807980230180001@cc497300-a.wlgrv1.pa.home.com> <slrn6q81lj.j00.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu> <james-ya023180000807982343080001@news> <rex-0907980239020001@cc497300-a.wlgrv1.pa.home.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <rex-0907980239020001@cc497300-a.wlgrv1.pa.home.com>, rex@smallandmighty.com (Eric King) wrote: >In article <james-ya023180000807982343080001@news>, james@unspam.edu >(James McCartney) wrote: > >:I've yet to see a graphical language that could express something in less >:space than a text equivalent. > > Why does it need to express something in less space than a text >equivalent? Space is cheap, especially if you have robust browsing >facilities. No space is not cheap. screens are small and VRs remain cumbersome to navigate. Yes I can keep quite a lot of source in my head and more fits on the screen as text that as icons. > I will mention that in Prograph and probably other graphical languages, >one can select an arbitrary group of icons and 'compress' it down to one. >This just can't be done with most textual languages, Sure it can. Just hasn't been. I think in some languages with simple syntax rules like Forth, Scheme, Smalltalk this would be easy. There ARE such things as folding editors and refactoring browsers too. > But is the text easier to understand and maintain than the graphical >representation? I've found it easier to manipulate and format text than the endless prettying up that I've had to do in MAX to prevent the snarl of wires look. Does it give you a better idea of the functionality? How >many lines of a textual code can you actually keep in your head at once? I've not seen a graphical language that was significantly easier to follow than a text language. In fact I've often found the opposite. I can more easily read a large structure of text than icons. That has been my experience with MAX. > I'm not advocating the abandonment of textual languages, but rather the >augmentation of them. As was I. >:All of them can be translated pretty easily to a lisp like syntax. > > That's somewhat of a moot point. Programs can often be translated from >one language to another. Will the Lisp-like syntax be easier to follow and >edit than the graph? I was just saying that I've seen no graphical language whose paradigm was so different that it could not be expressed concisely with text. I was not advocating a lisp syntax as the way to do it, just as an example. Now if there came to be some graphical representation that were quite difficult to express in text then that would be a good argument for a graphical paradigm, I've just not seen it yet. But lisp editors are quite smart about editing thanks to the syntax. > Everthing that can be done in a textual environment can be done in a >graphical environment. But not the reverse. A rather meaningless statement, as true as it is false. Yes, text cannot be its own GUI browser, so what? Icons need text as well. Your statement hinges on definitions of terms that amount to splitting hairs. Where do you differentiate graphical/text? I think no one wants to go back to ascii only CRTs.. I am quite interested in graphical languages. I'm not arguing against them, in fact I'm working on one, but text still does many things well that iconic languages don't. And to make simple statements like "I can do everything you can do but better" is just short sighted. --- James McCartney
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 8 Jul 98 09:11:00 Organization: Is a sign of weakness Message-ID: <SCOTT.98Jul8091100@slave.doubleu.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6ng8c3$vb52@onews.collins.rockwell.com> <SCOTT.98Jul2212935@slave.doubleu.com> <359CDF2B.D3BA8A8@mindspring.com> In-reply-to: "Scott P. Duncan"'s message of Fri, 03 Jul 1998 09:40:00 -0400 In article <359CDF2B.D3BA8A8@mindspring.com>, "Scott P. Duncan" <softqual@mindspring.com> writes: Scott Hess wrote: > You don't have to tell them to close the paint can before putting > it in the trunk. You don't have to tell them to close the > freezer door after putting the ice cream away. I guess you don't have any children! :-) Except for the smaller units, children generally know very well what you mean. Because they don't do what you mean is an entirely different issue from whether they are capable of understanding what you mean. This can range from actively opposing what you meant to simply ignoring you. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: Ian Wild <ian@cfmu.eurocontrol.be> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: Thu, 09 Jul 1998 09:52:49 +0200 Organization: Hierarchical Message-ID: <35A476D1.6B38ABD0@cfmu.eurocontrol.be> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> <rex-0807980230180001@cc497300-a.wlgrv1.pa.home.com> <slrn6q81lj.j00.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu> <james-ya023180000807982343080001@news> <rex-0907980239020001@cc497300-a.wlgrv1.pa.home.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Eric King wrote: > I will mention that in Prograph and probably other graphical languages, > one can select an arbitrary group of icons and 'compress' it down to one. > This just can't be done with most textual languages, Isn't this why the gods gave us (DEFMACRO)?
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.mac.programmer.misc,comp.sys.next.programmer Subject: Re: Thoughts on Realbasic Date: 9 Jul 1998 09:55:57 GMT Organization: The World's Usenet -- http://www.Supernews.com Message-ID: <6o243d$pjd$1@supernews.com> References: <6nqtoo$n4s@drn.newsguy.com> <6o1iah$fst$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: scott_sa@my-dejanews.com scott_sa@my-dejanews.com may or may not have said: [snip] -> In about 3 hours I re-built the interface (3 -> dilaogs, an about box, a tabbed window with four -> panes, 30-ish controls and two menus - all -> functioning ) of a simple app. I built a year ago. -> Compare that to about 12 hours! This would take about 1 hour in OpenStep, assuming that the Tabbed window wasn't even implemented in an interface builder palette. With the tabs integrated with IB, as apple intends to do in the first customer release of the Yellow Box development environment, this would be about a 15-20 minute task. -jcr
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: A truly eye-opening experience. Date: 9 Jul 1998 10:06:15 GMT Organization: The World's Usenet -- http://www.Supernews.com Message-ID: <6o24mn$pjd$2@supernews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Guys, As you know, I've always maintaned that computers suck. They're unreliable, they're more trouble to use than they should be, and I could go on and on. I've just been introduced to a project going on at the University of Pennsylvania, that simply blows me away. It's called EROS (Extremely Reliable Operating System) and it's a derivative of KeyKOS, which is a very mature capabilities-based operating system. Picture a system where when someone trips over the power cord, and when you turn it back on, everything's the way you left it. I've known all along that UNIX sucked, and I've just had a revelation as to just how much it sucks. If you want to know how computing *could* be, do a web search for "extremely reliable operating system" and "KeyKOS". Read the papers, and then take an hour or so to *ponder* the implications. Now, *that's* what the world needs in terms of basic reliability, security, and performance. -jcr
From: "Mike Anderson" <mikea@knowmed.com> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Thu, 9 Jul 1998 03:58:43 -0700 Organization: The World's Usenet -- http://www.Supernews.com Message-ID: <6o27rs$2s$1@supernews.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> Judson McClendon wrote in message ... >Judson McClendon wrote in message ... >>Scott Hess wrote: >>> It wasn't like some egghead invented object oriented programming and >>> design and then everyone in the world decided that we should try to >>> cram every project into that mold. >> >>Actually, that's a pretty good description of what DID happen. Had OO >>been invented by the folks in the trenches, not only would it have been >>far easier to learn and use than it is, but it would be time and cost >>effective, which it has certainly not proven to be, in real world >>practice. The movement for OO came from the universities, not the field. >>Any time you see all the universities trumpeting for something which has >>never been proven practical in practice, you can *know* it comes from >>academia. :-) > >Since there were a number of replies to this post, I'll answer them all >here. > >First, I wrote my post hastily and did not clearly articulate what I >meant to say. I used the word 'university', when I should have said >'academia', meaning the broader base of research and theoretical folks, >such as Xerox PARC, not just universities. Sorry. > >Here is an example of how C became popularized. In the 70's, DEC was >seriously dominant in the university environment. Because Bell Labs had >written C for the PDP, most of the system software for those computers >was in C. Students who wanted to seriously get into the system software >simply had to master C. As those students graduated and moved into >industry, they knew C, and stumped for C. This is where the main impetus >for C in the industry came from. Why else do you think companies like >DEC spend so much money assisting universities? The fact that C was >developed at a company became almost irrelevant, because it had saturated >the universities. > >To say that a language developed by academia would be better, because they >would prefer something elegant, is amusing. Nothing wrong with elegance, >but unless it works in the real world, on real world applications, the >point is irrelevant. Most academic folks have little or no experience in >writing large, real world applications, and in maintaining them over time. >You cannot learn to play a musical instrument, or ride a motorcycle, or >program a computer by reading about it in a book. You must actually do it >for a long time to learn to do it well. Not to say that study is of no or >little benefit, only that it is insufficient of itself. It is one thing >to read in a book, or be told by an instructor, that such and such is or >is not, important. But it is like being burned by fire. You can be told >a million times that fire will burn you, but when you stick your hand into >fire for the first time, your knowledge suddenly takes on a whole new >dimension! And the lasting burn gives you a respect, and a 'feel' for what >a burn is, that you could *never* learn any other way. This experiential >knowledge is essential in making 'judgment' calls and making decisions. > >In my opinion, the most serious problem we have today with development >environments is too much complexity. I say that because the cost of >new systems development and maintenance has skyrocketed. About 45% of >new systems started are not completed because of cost or time overruns, >and inability to meet design goals. Back in the early days of computing, >hardware was the expensive item, and software was relatively cheap. But >over the years, this has flip-flopped dramatically, and development >costs continue to rise at an alarming rate. But more importantly, as >the pace of change in business, and society in general, quickens, the >life span of new systems dwindles. At the same time, the time it takes >to develop new applications is increasing. When the time to develop a >new application becomes as great or greater than the life span of the >application, then you are in a "can't get there from here" scenario. >Many of the failed systems development efforts today die from just this >problem. > >It was clear by 1980 that this problem was becoming critical. One reason >development time is increasing is that applications are becoming more and >more complex. Some of this increased complexity is unavoidable. However, >much of this added complexity is created by the move to 'gee-whiz' >features like GUI, which may offer absolutely no benefit whatever, in >certain applications. Computers may offer absolutely no benefit whatever, in certain applications. (In other applications they seem kind of useful). > At the same time, software development tools and >environments are becoming more and more complex, for similar reasons. On >the shelves over my desk are my software reference manuals. There are >about 2 feet (.6 m) of manuals for VC++, and about 1 foot (.3 m) of >Windows reference manuals, plus a host of other manuals for VB, Office 97, >Java, etc. That represents a truly *enormous* amount of information to >assimilate and deal with on a daily basis. Part of the reason for the >complexity is the Windows API, MFC, and C++. It is an undeniable fact >that when people must deal with more complex environments, they will make >more mistakes. A programming language environment should be *only* as >complex as necessary, and definitely as simple as possible. I agree completely, but I don't understand how this relates to OO. >It is fairly easy to tell when a new method is effective. Simply put it >into practice and observe the results. As the saying goes "The proof of >the pudding is in the eating." Hard, cold, objective results show that >the result of these new development environments, including OO, are >greatly increased costs and development time, and less reliability. >After all, we must not forget that complexity increases geometrically. >Unfortunately, some people seem attracted to these methodologies with >little or no regard to the actual results from the field. > >Recently I encountered two prominent articles, one on the front page of >the Wall Street Journal (April 30, 1998, "Pulling the Plug"), and one on >the front page of Info World (June 29, 1998, "Return of the Mainframe"). >The people on the board of large companies are interested in the bottom >line, people. If these new technologies were delivering as they have >been touted, companies *would not* be backing off their use. What kind >of fool would consider abandoning technologies that were delivering the >goods? Not the people who make seven figures sitting on the boards of >large corporations, of that you can be sure. They just want what works. > >The problem, people, is that we need to make programming easier and >simpler, not harder. If OO was easier, or more economical, we would have >seen clear evidence by now in the numbers. Whose 'we'? I have seen it. > What we do see is clear >evidence to the contrary. 'We'? 'We' who? Is it the royal 'we'? >If we want economical, reliable software, we >will do this. We can't make people smarter. If we want more programmer >productivity, and better reliability, we had better make our tools easier >to use. We will never do it by ignoring the evidence that is clearly at >hand. And we had better learn to apply complex technology where it is >needed or useful, not spray it across everything, whether it is warranted >or not. We are free to ignore the evidence. But we will have to deal >with the consequences, and we are. It's seems like you have little or no experience in producing object orient programs and in changing them over time. You cannot learn to play a musical instrument, or ride a motorcycle, or learn object oriented programming by reading about it in a book. You must actually do it for a long time to learn to do it well. Not to say that study is of no or little benefit, only that it is insufficient of itself. It is one thing to read in a book, or be told by a WSJ article, that such and such is or is not, important. But it is like being burned by fire. You can be told a million times that fire will burn you, but when you stick your hand into fire for the first time, your knowledge suddenly takes on a whole new dimension! And the lasting burn gives you a respect, and a 'feel' for what a burn is, that you could *never* learn any other way. This experiential knowledge is essential in making 'judgment' calls and making decisions. ...Mike
From: *johnnyc*@or.psychology.dal.ca (John Christie) Newsgroups: comp.sys.mac.programmer.misc,comp.sys.next.programmer Subject: Re: Thoughts on Realbasic Date: Thu, 09 Jul 1998 09:42:32 -0300 Organization: ISINet, Nova Scotia Message-ID: <*johnnyc*-0907980942330001@jchristie.halifaxcable.dal.ca> References: <6nqtoo$n4s@drn.newsguy.com> <6o1iah$fst$1@nnrp1.dejanews.com> <6o243d$pjd$1@supernews.com> In article <6o243d$pjd$1@supernews.com>, jcr.remove@this.phrase.idiom.com wrote: > scott_sa@my-dejanews.com may or may not have said: > [snip] > > -> In about 3 hours I re-built the interface (3 > -> dilaogs, an about box, a tabbed window with four > -> panes, 30-ish controls and two menus - all > -> functioning ) of a simple app. I built a year ago. > -> Compare that to about 12 hours! > > This would take about 1 hour in OpenStep, assuming that the Tabbed window > wasn't even implemented in an interface builder palette. With the tabs > integrated with IB, as apple intends to do in the first customer release of > the Yellow Box development environment, this would be about a 15-20 minute > task. Hmmm, I've seen OpenStep development and it is relatively simple. However, I think this person was relating times to development in a brand new environment. Your times for OpenStep are for individuals experienced in that environment. Experienced RealBasic programmers would come back with similar times. In addition, the speed you talk about in OpenStep (OS X server?) is not as "obviously" obtained as in RealBasic. Now, the power of what you can do is another matter. -- You aren't free if you CAN choose - only if you DO choose. All you are is the decisions you make. Remove "*" and ohnny (i.e. jc@) to reply via email
From: "Scott P. Duncan" <softqual@mindspring.com> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Thu, 09 Jul 1998 09:25:06 -0400 Organization: SoftQual (Quality That Lasts) Message-ID: <35A4C4AF.44E3047D@mindspring.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6ng8c3$vb52@onews.collins.rockwell.com> <SCOTT.98Jul2212935@slave.doubleu.com> <359CDF2B.D3BA8A8@mindspring.com> <SCOTT.98Jul8091100@slave.doubleu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; x-mac-type="54455854"; x-mac-creator="4D4F5353" Content-Transfer-Encoding: 7bit Scott Hess wrote: > In article <359CDF2B.D3BA8A8@mindspring.com>, > "Scott P. Duncan" <softqual@mindspring.com> writes: > I guess you don't have any children! :-) > > Except for the smaller units, children generally know very well what > you mean. Because they don't do what you mean is an entirely > different issue from whether they are capable of understanding what > you mean. This can range from actively opposing what you meant to > simply ignoring you. Well, while the original was intended more for its (attempted) humorthan literal content (hence the smiley), when it comes to certain work- related directions/requirements that get a bit more complex (and personal), your latter sentence somewhat represents a real problem (in that the behavior exhibited is often more sophisticated in how it manifests the basic behavior/attitude you describe). And what one person would expect aother to be "capable of understanding" compared to what the other person seems to understand is often an issue as well.
From: Byron Goodman <bgoodman@dev.tivoli.com> Newsgroups: comp.sys.next.programmer Subject: Re: Docs Date: Thu, 09 Jul 1998 10:40:52 -0500 Organization: Tivoli Systems Inc., Austin, TX Message-ID: <35A4E484.70876595@dev.tivoli.com> References: <35A25FB5.513AC61C@dev.tivoli.com> <6nu34l$hmp$1@news.spacelab.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: "Charles W. Swiger" <chuck@codefab.com> Yea, I'm trying to gather information about hardware specific issues. Like query the bios about Y2K info, etc. Opening /dev/kmem and searching isn't working very well. Also, I'm really ticked there isn't a POSIX library, when the MAN pages are there for POSIX functionality. What I really need is some documentation on how to write a kernel service module. But the docs that are provided are completely out of date, and are for Mach 2.5. Charles W. Swiger wrote: > Byron Goodman <bgoodman@dev.tivoli.com> wrote: > >Are there any NeXT or OpenStep Mach developer documention available? I'm > >needing to find out about system calls that query hardware and devices. > >I found the Rhapsody Operating System Software Guide, but has very > >limited information. The documentation I acquired from CMU I have found > >is not completely consistent with NeXT/OpenStep. > > Could you provide more information about what you're trying to do? > > Your question is so broad that you could be directed towards anything from > the details of Mach device drivers, to generic Unix /dev/ stuff, to docs on > SCSI, etc.... > > -Chuck > > Charles Swiger | chuck@codefab.com | standard disclaimer > ---------------+-------------------+-------------------- > "Microsoft: we make the easy almost impossible." -- Byron Goodman bgoodman@dev.tivoli.com
From: phil@pfsystems.com (Phil Stenson) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: Thu, 09 Jul 1998 14:56:39 -0400 Organization: Present & Future Systems, Inc. Message-ID: <nJRp10sM/4DZ090yn@pfsystems.com> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> NNTP-Posting-Time: 9 Jul 1998 19:45:25 GMT While reading comp.software-eng on 7 Jul 1998 23:36:33 GMT, I saw cbbrowne@news.hex.net (Christopher Browne) write: >"Why do engineers still write out equations rather than ``simply'' >drawing the relationships?" > >The written language of mathematics has been, and still is, symbolic >equations as opposed to graphical pictures. Well, I submit that if civil engineers could put together a virtual building to see if it would stand up, they would. They would specify the various standard parts that are available (in order to keep costs down), erect the building in their 'holodeck' and subject it to the various tests (also all standardized). If something failed, they'd replace it with the next more costly design (larger part, more complicated contruction, etc.) until it withstood the tests. That would be the design. This is Design by Trial-and-Error. Sure, the better companies would have had experience and perhaps even developed systems by which they could eliminate some of the 'trial' possiblities as not capable. The reason that designs are still done by equations is that the above is not yet possible and the only other alternative (ie: real building as prototype) would be too costly for each failure. So the paper design is done (and everything is multiplied by 2 to make sure <g>). The advantage of building design is that we've had hundreds of years of practice. During that time, many buildings and other structures *have* fallen down due to lack of knowledge. However engineers are very good at integrating the lessons learned into present practice. Essentially, today's software design is being done using the 'holodeck' method, simply because it is possible. Unfortunately, there are few standard components available with which to reduce costs. And there are certainly no standardized tests available except for the poor end-user. This results in costly products which have a high degree of probability of 'falling down', if they ever get finished at all. Perhaps after several hundred years of software development has passed, we'll get to the same point as civil engineering is today. Developers of that future era will look back on the 'early days' (ie today) and wonder how we ever produced anything that *didn't* fail given our primitive tools and even more primitive methods. PS. I are an engineer! -- Phil Stenson Present & Future Systems, Inc. Toronto phil@pfsystems.com ------ ]We never respond to SPAM. Period.
Newsgroups: comp.sys.next.programmer From: cdouty@netcom.com (Chris Douty) Subject: Re: Docs Message-ID: <cdoutyEvuF7I.M61@netcom.com> Organization: Netcom References: <35A25FB5.513AC61C@dev.tivoli.com> <6nu34l$hmp$1@news.spacelab.net> <35A4E484.70876595@dev.tivoli.com> Date: Thu, 9 Jul 1998 19:52:30 GMT Sender: cdouty@netcom15.netcom.com In article <35A4E484.70876595@dev.tivoli.com>, Byron Goodman <bgoodman@dev.tivoli.com> wrote: >Yea, I'm trying to gather information about hardware specific issues. Like >query the bios about Y2K info, etc. Opening /dev/kmem and searching isn't >working very well. > >Also, I'm really ticked there isn't a POSIX library, when the MAN pages are >there for POSIX functionality. What I really need is some documentation on how >to write a kernel service module. But the docs that are provided are completely >out of date, and are for Mach 2.5. There is a MachOS guide in the Apple developer docs site. I don't have the exact URL handy. This is a PDF of NeXT's old book on the Mach kernel and covers such things as Loadable Kernel Servers and the like. Still, the information is a little sketchy. NeXT stopped providing even the headers for some hardware access in the NS 2.x timeframe. There isn't a POSIX library in OS/mach because it was just horribly broken. NeXT removed POSIX compilation support between NS3.3 and OS/mach 4.0, IIRC. Yes, it sucks, but Rhapsody (er, MacOS X Server) will be better. I'm guessing that you are running OS/mach on Intel hardware. Is that correct? Sorry if this doesn't help. -Chris -- Christopher Douty - Rogue Engineer trapped in a land of software cdouty@netcom.com "Frequently the messages have meaning; that is they refer to or are correlated according to some system with physical or conceptual entities. These semantic aspects of communication are irrelevant to the engineering problem." -Shannon
From: cbbrowne@news.hex.net (Christopher Browne) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 9 Jul 1998 23:45:36 GMT Organization: Hex.Net Superhighway, DFW Metroplex 817-329-3182 Message-ID: <6o3kn0$7bh$10@blue.hex.net> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> <rex-0807980230180001@cc497300-a.wlgrv1.pa.home.com> On Wed, 08 Jul 1998 02:30:17 -0400, Eric King <rex@smallandmighty.com> wrote: >In article <6nube1$c8n$6@blue.hex.net>, cbbrowne@hex.net wrote: > >:The answers are likely the same as those to the question: >: >:"Why do engineers still write out equations rather than ``simply'' >:drawing the relationships?" > > I'd argue that when they write out an equation that they *are* drawing >the relationships. Yes and no. Equations are a *symbolic* language. They contain some visual components, but are certainly not "visually oriented" in the way an engineering diagram is. > A better question might be: why do programmers still draw flowcharts >rather than simply writing out a description of their programs? Do they? I probably have a flowchart template somewhere; it hasn't been used professionally in on the order of 30 years. (Note that I'm not much more than 30; the template belonged to Mom, and she ceased working as a programmer not long after I came along...) If I want a flowchart, I'm more liable to take the source code and run it through some preprocessor rather than to take a flowchart and try to turn it into a program. >:The written language of mathematics has been, and still is, symbolic >:equations as opposed to graphical pictures. > > But aren't those symbols just groups of small graphical pictures? I >think you're making a very artificial distinction. An 'O' could be the >representation of a letter in the English alphabet or it could just be a >small picture of a circular object. The notion of "zero" doesn't indicate roundness in contrast with the notion of "one" indicating some sort of straightness/verticality. No, those symbols are *not* just groups of graphical pictures. Chinese and Egyptian ideograms may be pictures of things; the symbols used in Latin-based languages and mathematics are *not* pictures of things. >:There may be instances in which visual representations provide insight; > > I'll go one step further and say there are instances where a graphical >representation makes a lot more sense than a textual representation. Not >all computing tasks can be conveniently represented by a linear stream of >text. One of my professors once said that all proofs are one line proofs; it's just a question of how long that line is... To be sure, that suggests (by its evident silliness) that there is merit to having structures other than simply that of a linear stream of text. Donald E Knuth's efforts with respect to the TeX system display that nicely; a *HUGE* amount of effort has gone, over the years, into providing better representations for information. But to the extent to which it is symbolic, that is something that adding pictures doesn't help. >there is a huge need for tools that help programmers manage the >complexity of their projects. Class browsers, GUI builders, flowcharting >tools, etc. are all attempts to augment textual representations of >programs. The sheer number of those types of tools leads me to question >whether a purely textual representation is really appropriate. And \[ x = \frac{ y^2 + z^2 } { y^2 - z^2} \] is in similar fashion best displayed in a fashion that is not "purely textual." It is nonetheless still quite purely symbolic. It is *not* a picture of an equation by any stretch of the imagination. > IMO, the (distant) future of programming will be based upon languages >like Prograph, which mix graphical and textual representations of >programming constructs. Purely textual programming languages are trying to >encode a vast quantity of information using a very limited set of symbols >and layout rules. This results in a steady stream of new programming >languages which really aren't any more expressive than preexisting >languages. Furthermore, a lot of effort is being spent developing large >libraries, frameworks, toolkits, etc. which try to anticipate what one >might need to do in a program. IMO, more effort needs to be spent on >trying to help one manage the complexity of what one is trying to do. If you are indicating "graph" as in "a set of nodes and arcs," then I'll agree. And I have no problem with the notion that some of those graphs might have some form of visual representation. But that's a rather different thing from suggesting a "Virtual Reality" approach to programming. -- Microsoft has brought the microcomputer OS to the point where it is more bloated than even OSes from what was previously larger classes of machines altogether. This is perhaps Bill's single greatest accomplishment. cbbrowne@hex.net- <http//www.ntlug.org/~cbbrowne/lsf.html>
From: enders@bolshoi.cc.misu.nodak.edu (Todd Enders) Newsgroups: comp.sys.next.programmer Subject: Re: A truly eye-opening experience. Date: 10 Jul 1998 01:25:54 GMT Organization: North Dakota Higher Ed. Network Message-ID: <6o3qj2$pmi$1@node2.nodak.edu> References: <6o24mn$pjd$2@supernews.com> In article <6o24mn$pjd$2@supernews.com> jcr.remove@this.phrase.idiom.com (John C. Randolph) writes: > Picture a system where when someone trips over the power cord, and when you > turn it back on, everything's the way you left it... > I've an old (vintage 1976) PDP 11 with core memory that pretty well does this. Pull the plug, come back days/weeks/months later, power the beast up, and everything in core is just as it was. If CPU regs and memory were still non-volatile, it wouldn't make a whole lot of difference what OS you were running -- the machine would "remember" what state it was in, where it was in the code, etc., and would pick up right where it left off when the power came back. Unfortunately, silicon forgets when the juice goes away. Amazing how far we've come in 30-odd years... :-)
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: A truly eye-opening experience. Date: 10 Jul 1998 03:40:47 GMT Organization: The World's Usenet -- http://www.Supernews.com Message-ID: <6o42fv$cqn$1@supernews.com> References: <6o24mn$pjd$2@supernews.com> <6o3qj2$pmi$1@node2.nodak.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: enders@bolshoi.cc.misu.nodak.edu Todd Enders may or may not have said: -> In article <6o24mn$pjd$2@supernews.com> jcr.remove@this.phrase.idiom.com (John -> C. Randolph) writes: -> -> > Picture a system where when someone trips over the power cord, and when you -> > turn it back on, everything's the way you left it... -> > -> -> I've an old (vintage 1976) PDP 11 with core memory that pretty well does -> this. Pull the plug, come back days/weeks/months later, power the beast up, -> and everything in core is just as it was. If CPU regs and memory were still -> non-volatile, it wouldn't make a whole lot of difference what OS you were -> running -- the machine would "remember" what state it was in, where it was in -> the code, etc., and would pick up right where it left off when the power came -> back. Unfortunately, silicon forgets when the juice goes away. Amazing how -> far we've come in 30-odd years... :-) Of course, Silicon also offers several million times the bit density, and it takes a trivial amount of power relative to magnetic cores. What EROS and KeyKOS give you is the ability to recover everything up to the last checkpoint (checkpoints are made every five minutes or so.) Apps that need smaller granualrity than that can append memory pages to the checkpoint at any time, so that transactions can be preserved. -jcr
From: eiffelpgmr@nni.com (Tom Morrisette) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Thu, 09 Jul 1998 02:04:47 GMT Message-ID: <35a42498.4135708@news.nni.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> <Kzsn1.114$pr5.299198@news4.atl.bellsouth.net> <6ns598$il4$1@murdoch.acc.Virginia.EDU> <Snpo1.200$Od.706436@news4.mia.bellsouth.net> On Tue, 07 Jul 1998 13:28:18 GMT, "Judson McClendon" <judmc123@bellsouth.net> wrote: >It's been interesting to read all these comments to my post, particularly >those about people who were around before C, because I started programming >in 1968. :-) Anyway, it is clear that many of the respondents here have >yet to learn the lesson that just because it is in books, and is being >taught at school, doesn't mean it is correct! Most of what I say in that >post are things I personally was witness to, or was fairly close to. >-- Well, if seniority is the criterion for correctness, I feel obligated to note that I've been programming since 1963, observerved the state of programming pretty carefully since that time, and am in strong agreement with those who feel you distorted history.
From: rex@smallandmighty.com (Eric King) Newsgroups: comp.programming,comp.lang.prograph,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: Fri, 10 Jul 1998 01:25:50 -0400 Organization: The Small & Mighty Group Message-ID: <rex-1007980125510001@cc497300-a.wlgrv1.pa.home.com> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> <rex-0807980230180001@cc497300-a.wlgrv1.pa.home.com> <slrn6q81lj.j00.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu> <james-ya023180000807982343080001@news> <rex-0907980239020001@cc497300-a.wlgrv1.pa.home.com> <james-ya023180000907980300150001@news> In article <james-ya023180000907980300150001@news>, james@unspam.edu (James McCartney) wrote: :No space is not cheap. screens are small and VRs remain cumbersome to :navigate. I find scrolling and jumping about through linear blocks of code cumbersome. Paned editors and browsers help, but they still feel somewhat clumsy to me. Incidentally, a side benefit of the extra space that Prograph code takes up is that there's much more room for descriptive multiline comments. :> I will mention that in Prograph and probably other graphical languages, :>one can select an arbitrary group of icons and 'compress' it down to one. :>This just can't be done with most textual languages, : :Sure it can. Note, I said 'most' textual languages, not all. :I think in some languages with simple syntax rules Exactly, some languages. :like Forth, Scheme, Smalltalk this would be easy. Scheme possibly, but I'm not sure about Smalltalk.. I think you would have to limit the ways in which you group code. I'm also unsure of what other operations could be applied to the groupings once you made them. Part of the reason this is so easy to do in Prograph is that there are no variables, and hence no name-clashes can occur when doing a grouping. :I've found it easier to manipulate and format text than the :endless prettying up that I've had to do in MAX to prevent the snarl :of wires look. Then that's a problem with Max's interface. Prograph provides lots of aids for automatically aligning icons, nodes, and connections. :I've not seen a graphical language that was significantly easier to :follow than a text language. I don't think there ever will be one generally. A lot is dependent on the coding style of the programmer, but I do find Prograph programs very easy to follow. They've got an inherent and very apparent tree-like structure. :In fact I've often found the opposite. I can more easily read a large structure of text than icons. That has been my experience with MAX. I've had a different experience with Prograph. Since one can easily collapse arbitrary groups of icons into local or global methods, developers often take advantage of this fact so it's rare to see more than 10-15 icons in a single code window. As a result, code is usually presented in very manageable chunks. I've heard Max mentioned before, but I've never seen it in action. Do you have any pointers to it? Is it a general purpose visual language or an application-specific one? :But lisp editors are quite smart about editing thanks to the syntax. As is Prograph. IMO, the real difficulty in graphical languages is in anticipating what sorts of shortcuts and aids a programmer is going to need. There have been decades of research and experience with text editors to know what features are important. Visual languages are still in their infancy. The Prograph folks figured out a lot of what is needed for a visual language to work (to my knowledge Prograph is still the only 'pure' visual language that has been used to build commercial off-the-shelf apps.), but after 10 years there are still rough edges. :> Everthing that can be done in a textual environment can be done in a :>graphical environment. But not the reverse. : :A rather meaningless statement, as true as it is false. Yes, text cannot :be its own GUI browser, so what? Icons need text as well. Icons do *not* always need text. For instance, consider the Prograph 'injection' node annotation which is applied to a blank method icon. The name of that method is determined by a string that flows into the injection node. There are other types of input and output node annotations, which alter the functionality of the method to which they are applied. However, this altering of functionality has no textual representation onscreen. :hinges on definitions of terms that amount to splitting hairs. Where do :you differentiate graphical/text? I think no one wants to go back to ascii :only CRTs.. No, but most programming languages these days still seem to be targeting such a barebones editing environment. I'm hoping that portable graphical development environments like Squeak will help nurture the development of visual languages. Squeak really raises the bar of what kinds of graphics capabilities a language implementor can rely on. Smalltalk's also a wonderful language to experiment in. ::Eric
From: "Mike Anderson" <mikea@knowmed.com> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Re:Programming and the colossal failure to advance - VR Date: Fri, 10 Jul 1998 01:21:42 -0700 Organization: The World's Usenet -- http://www.Supernews.com Message-ID: <6o4j6l$ltu$1@supernews.com> References: <6nu3gi$97i$1@nnrp1.dejanews.com> tj_sam@my-dejanews.com wrote in message <6nu3gi$97i$1@nnrp1.dejanews.com>... [snip] >For example, "goto" in C >can be directly represented visually as a graphic vector with >source and destination directly in the actual source code (you >don't need semantic labels). > >A basic VR program editor would essential be a flowcharting program >with the actual code embedded within the flowchart (not the simplex >flowchart but actual source level flowcharting). A multiple column >text editor in which major branches and loop would be directly >represented by visual vectors. Individual subroutines or functions >would be represented as 2D (X,Y) wire frame objects. If subroutine >or function calls are represented as Z plane bi-directional vector >(recursive function are outlawed) It seems like your trying to limit programming constructs to things that are analogous to things in the physical world (the phrase "VR programming" implies that also). This is similar to what has been done to the user interface with GUIs. They try to mimic the real world in many ways: drag, drop, push button, etc. This generally makes things easier to learn, but places severe limits on the kinds of things they can do. The most notably missing thing is abstraction. GUIs are impaired because of their inability to handle abstraction, but it's not much of a problem because people don't expect much of them. Programming without abstraction would be a very tedious excersise (but it would in some sense be simpler). Abstraction is a powerful tool, and it should be refined, not ignored or outlawed. But "visual" programming doesn't have to be overly concrete. I think the potential of visual programming is in expressing abstraction in ways that are easier for people to deal with. ...Mike
From: mmh@dcs.qmw.ac.uk (Matthew M. Huntbach) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Followup-To: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Date: 10 Jul 1998 09:07:36 GMT Organization: Queen Mary & Westfield College, London, UK Message-ID: <6o4lko$rs1$14@beta.qmw.ac.uk> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <SCOTT.98Jul2210220@slave.doubleu.com> <EJ4n1.1946$C%.2847671@news3.atl.bellsouth.net> Judson McClendon (judmc123@bellsouth.net) wrote: > Scott Hess wrote: > > It wasn't like some egghead invented object oriented programming and > > design and then everyone in the world decided that we should try to > > cram every project into that mold. > Actually, that's a pretty good description of what DID happen. Had OO > been invented by the folks in the trenches, not only would it have been > far easier to learn and use than it is, but it would be time and cost > effective, which it has certainly not proven to be, in real world > practice. The movement for OO came from the universities, not the field. Eh? Which universities were using C++ before industry? Academia was still pushiong functional and/or logic programming as the paradigm of the future when the C++ revolution happened. And, yes, I know, C++ is not true OO programming, but it's what most people think of OO programming. Matthew HUntbach
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: A truly eye-opening experience. Date: 9 Jul 98 08:28:40 Organization: Is a sign of weakness Message-ID: <SCOTT.98Jul9082840@slave.doubleu.com> References: <6o24mn$pjd$2@supernews.com> In-reply-to: jcr.remove@this.phrase.idiom.com's message of 9 Jul 1998 10:06:15 GMT In article <6o24mn$pjd$2@supernews.com>, jcr.remove@this.phrase.idiom.com (John C. Randolph) writes: As you know, I've always maintaned that computers suck. They're unreliable, they're more trouble to use than they should be, and I could go on and on. <...> Picture a system where when someone trips over the power cord, and when you turn it back on, everything's the way you left it. I've known all along that UNIX sucked, and I've just had a revelation as to just how much it sucks. <...> Now, *that's* what the world needs in terms of basic reliability, security, and performance. Sometimes, my mood WRT the computer industry gets black indeed. People love Windows and MacOS, they defend them nigh unto death, yet... 95% of the specific items people bitch about in their computer experience have been solved. And I don't mean "in theory", I mean that they could be there _today_. There are, of course, that last 5%, and once you solved the surface complaints, you'd most likely find a whole bunch of others behind them that people didn't even realize were complaints, but it still sucks. An _easy_ for-instance is the use of drive letters under Windows NT and 95. Drive letters cause no end of trouble. In a relative sense, people get a lot of work done with no reference to them - but in an absolute sense, hundreds of man-years have surely been spent working around drive letters, making things work right. This didn't need to happen, single-rooted filesystems have proven their value over the past 30 years. Windows could _trivially_ take a rooted filesystem and associate drive letters with it - but it's nigh impossible to do the reverse trick. [I've even heard of a product for Windows 95 which lets you combine all partitions into a single logical C: drive.] I'm not saying "Unix is better" across the board. Instead, what I'm saying is that all of the good ideas in Unix have been there for 10-15 years, and Microsoft _willfully_ ignored most of them in favor of inventing their own good ideas. The difference being that the good ideas in Unix have had 15 years of field testing, and the ones that looked good but were really bad have fallen by the wayside. I suspect human nature is at the root. Microsoft (and others, Microsoft is just the biggest) would rather control the market than deliver a decent user experience. Using good ideas from elsewhere in the industry would imply that Microsoft isn't the sole font of good ideas. I just hope that Apple breaks their own version of the cycle, and _embraces_ the technology the past has to offer rather than rejecting it. CPUs are fast, memory is large - all the past arguments against abstraction are quickly falling by the wayside, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: tj_sam@my-dejanews.com Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: Fri, 10 Jul 1998 12:46:43 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6o52fj$tvg$1@nnrp1.dejanews.com> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> <rex-0807980230180001@cc497300-a.wlgrv1.pa.home.com> <slrn6q81lj.j00.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu> <rex-0907980206430001@cc497300-a.wlgrv1.pa.home.com> In article <rex-0907980206430001@cc497300-a.wlgrv1.pa.home.com>, rex@smallandmighty.com (Eric King) wrote: > I agree that a grouping of arbitrary pictures do not constitute a > language, but why must programming languages be constrained to such a > narrow set of symbols and layout rules? Again that problem of typecasting programming as a language, perhaps its better to resolve a new definition for programming and what programs are: Based on a previous posting, control flow elements were defined as an essential core element in all computer coding (languages) with a graphic or visual representation. In the presentation of source code as 3D wireframe (control flow) model, the actual code is embedded within the control flow. The specific code (language) is important but at a higher level of abstract the coding becomes generic and a secondary element. What is left of the model are the control flow elements of conditional branches, loops and subroutines/functions (bi-directional vectors) and defines the elements common to all types of programming. Mechanical engineers design basic machines which can be defined as wheels and levers held in structural frameworks in the "real world". Programmers design virtual machines (generic term - not Java) with loops (wheels) and conditional branches (levers) held in subroutines (structural frameworks) in the "virtual world". This amounts to defining programming as a branch of mechanical engineering. A conceptual requirement for turning machines? T.J. ---------- -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: "Stu" <stu@grayechnologies.com> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 10 Jul 1998 13:07:30 GMT Organization: Gray Technologies, Inc. Message-ID: <01bdac03$b070be20$0fd29cce@kiki.graytechnologies.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359BB163.7A97DEC6@gssec.bt.co.uk> <359E92B5.5E4A5A33@earthlink.net> <gio+van+no+ni+8-0407982316390001@dialup12.tlh.talstar.com> Giovanni 8 <gio+van+no+ni+8@tal+star+spam.com> wrote in article <gio+van+no+ni+8-0407982316390001@dialup12.tlh.talstar.com>... > > Charles Hixson wrote: > >> Alan Gauld wrote: > >>> boracay@hotmail.com wrote: > >> I would have agreed with this between 1990 and 1995 but > >> the tools available in recent years have changed my mind > >> - its now only as hard as it was 20 years ago... > > > Actually, I think that programming is, in fact, getting harder. > > This is because the simple things that are worth doing tend to > > already be done. I don't currently see and reason to design a > > spread-sheet, it's been done. The current models even have > > tail-fins. > [snip] ... > But, the main point I want to make is that the fancy GUIs > and IDEs are actually making development more difficult in > some ways. I mean, how complex is binary? It's just a few > 1s and 0s, after all. You just line them up in the right > clumps in the right order, and you've got a program. Obviously, you are either being completley facitious or you have been so isolated from any real concepts of programming that you don't know what your talking about. This statement would be like saying "how complex is writing a best-selling novel? It's just a few letters, after all. You just line them up in the right clumps in the right order and you've got the next War and Peace." Sheesh! > Now, some of the IDEs & code generators get in the way. A simple > SQL query with a couple inner joins, gets turned into a > nested mess by some of the "tools". And they hide things > from you that make it more difficult to debug. Ideally, > they'd reveal all, though in a conveniently organized > manner, with e.g. forms, form elements, DB tables, code > and all the attributes related to each all nicely inter- > related graphically. In order for a processor(s), which is really what is doing the work your talking about here. No, there aren't little people inside your machine reading your source code and running around real fast to paint your screen or type out your files to the printer. Just a dumb hunk of silicon that does exactly, and only, what your program told it to do. BUT, in order for your simple SQL query to be implemented a whole helluva lot of layers of software HAD to be implemented. Software that most people seem to have no clue is even running. Try doing an SQL query on a PDP-11. You can't ... because there is NO SOFTWARE SUPPORT for SQL on a PDP-11. And, if some enterprising person decided they wanted to support a "simple SQL query" on a PDP-11, it would certainly be a of "limited use" by the user's expectations. These "superflous" IDE's and code generators are certainly not necessary, but without them, someone who has NO CLUE what a processor actually does to support High level constructs like "Simple SQL queries" would be unable to accomplish the task that they want to do - Database management - and would instead have to do system program to make the support necessary to do "simple SQL queries" on their platform. [snip] ... -- Stu Potter Principal Engineer Gray Technologies, Inc -- to reply delete the REMOVETHIS from the e-mail address
From: "Stu" <stu@grayechnologies.com> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Program99, was (Re: Programming and the colossal failure to advance) Date: 10 Jul 1998 14:35:16 GMT Organization: Gray Technologies, Inc. Message-ID: <01bdac0f$f6d84f20$0fd29cce@kiki.graytechnologies.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <35A0A008.41C6@pc-plus.de.remove.this.bit> Stephen Riehm <sr@pc-plus.de.remove.this.bit> wrote in article <35A0A008.41C6@pc-plus.de.remove.this.bit>... > All paradigm and programming language discussions aside, I find it > ironic that programmers really have been left in the dark as far as > software tools go. > > For example, with the exception of a very small minority, no-one > writes documents with a text editor any more. Everyone uses word > processors with automatic indexing, formatting, cross referencing, > spelling and gramitic checks. [snip] > > The same should be true for programming. The representation on the > screen doesn't have to be the same as the representation on disk. > If we had code processors, programs could be treated like hypertext > documents, interactive - living - systems. I concur with this, but remember (assuming your old enough to remember) the issues that brought us to todays word processors, principally the incompatibilities. Today using a word processor on a document is pretty mindless. Just open the document with your favorite word processor and, in general, viola - you can edit your document. The reason we can do this is because the word processor makers have gone to the effort of supporting the majority of document formats in use today. AND, makers of new word processors are formatting to one or more of these document formats. This was not always the way it was. Market pressures moved us to this (maybe not efficient, but tremendously easier) state. > What you see on the screen > is just a representation of the program - formatting becomes a > non-issue, reading someone elses code would be much easier, because > at least the visual representation remains familiar, leaving you > just the task of understanding the algorythim. Formatting, is infact the issue. Not the format of the source code, but the format of the document that now represents the "visual representation" of the code. Most integrated design environments and NextG language systems do this kind of high level representation for you. The problem may is like in the "early days" of word processors ... lack of compatibility between makers formats. Having the system go back to the pure text "source" code is not a big help. Since the source code would now be a translation of your "visual" representation of your software by the system you used, a different system may "visually" represent that source code differently. There are no standards, after all. Just like when you made a doc in WP with two columns of text, now Word takes the doc and displays it as one column of text. > Finding out what > lies behind a variable, class, method or function is as simple as > clicking on the symbol - the relavant information would be taken > directly from the symbol table, and you would KNOW that it's the > right information. Problems related to multiple definitions of > symbols in different contexts also disapear, because the code > processor knows exactly which symbol you wanted when you wrote your > code, no text based system can do this, instead, we rely on the > black-box concept of a compiler to work out which symbol was meant. The statement is true, if you use the same system all the time or you have a standard. Right now, the only standard is the language itself, which is DEFINED as text - not symbols. UML might be considered one attempt to standardize a "visual" representation for a software language. But, it is more of a design tool/methodology, not a programming language and any two "translation" tools developed by different manufacturers will translate the symbols into different source code. > Today's compilers still don't give any live feedback, nor do they > remember what they did before (OK, libraries and pre-compiled headers > are a step in that direction, but it's only the first step). > > even the most advanced systems I've seen still rely on the > programmer typing freestyle text, and hoping that the compiler will > be able to make sense of it - as you gain experience you also gain > the skill to be able to do this, but you still have to jump through > hoops if you want to "find that class of object which does xyz" or > to find out exactly what were all those parameters to his_func()? > Maybe we are at the early stages of "code processors" and the same market pressures that moved us to the current word processor state-of-affairs would begin. But there was a lot of incompatibilites to begin with when word processors first "hit the street". I agree with this idea, being able to RELIABLY generate code at a higher level of abstraction (above text) like the EDA world would be a large step in the right direction of software "manufacturing". > I think this is more the essence of the original posting - we've > already discussed paradigms and languages to death, but the underlying > system of a user writing text and hoping that the computer will cope > is still there. -- Stu Potter Principal Engineer Gray Technologies, Inc -- to reply delete the REMOVETHIS from the e-mail address
From: mmh@dcs.qmw.ac.uk (Matthew M. Huntbach) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Followup-To: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Date: 10 Jul 1998 15:09:15 GMT Organization: Queen Mary & Westfield College, London, UK Message-ID: <6o5aqr$7e9$5@beta.qmw.ac.uk> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> <rex-0807980230180001@cc497300-a.wlgrv1.pa.home.com> <slrn6q81lj.j00.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu> <rex-0907980206430001@cc497300-a.wlgrv1.pa.home.com> <6o52fj$tvg$1@nnrp1.dejanews.com> tj_sam@my-dejanews.com wrote: > Based on a previous posting, control flow elements were defined as > an essential core element in all computer coding (languages) with > a graphic or visual representation. In the presentation of source > code as 3D wireframe (control flow) model, the actual code is > embedded within the control flow. The specific code (language) is > important but at a higher level of abstract the coding becomes > generic and a secondary element. What is left of the model are the > control flow elements of conditional branches, loops and > subroutines/functions (bi-directional vectors) and defines the > elements common to all types of programming. Control flow is very much tied to the old fashioned von Neumann model of the computer. It ought to be the first thing that is thrown away as peripheral when one is truing to move to a more abstract model. Matthew Huntbach
Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng From: bks@netcom.com (Bradley K. Sherman) Subject: Re: Program99, was (Re: Programming and the colossal failure to advance) Message-ID: <bksEvw2oI.4xA@netcom.com> Organization: DNA + Sunlight References: <6ng34a$j2g$1@nnrp1.dejanews.com> <35A0A008.41C6@pc-plus.de.remove.this.bit> <01bdac0f$f6d84f20$0fd29cce@kiki.graytechnologies.com> Date: Fri, 10 Jul 1998 17:17:06 GMT Sender: bks@netcom19.netcom.com In article <01bdac0f$f6d84f20$0fd29cce@kiki.graytechnologies.com>, Stu <stu@grayREMOVETHIStechnologies.com> wrote: >incompatibilities. Today using a word processor on a document is pretty >mindless. Just open the document with your favorite word processor and, in >general, viola - you can edit your document. The reason we can do this is Unless of course you are using Word Version N-1 and the document was created by your colleague using Word Version N. --bks
From: ajmas@bigfoot.com Newsgroups: comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming Subject: Re: Programming and the colossal failure to advance Date: Fri, 10 Jul 1998 19:01:16 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6o5ods$k6$1@nnrp1.dejanews.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <vhIsdqY67dTD-pn2-dGSKqdjMrSfd@dynamic47.pm03.san-rafael.best.com> <bksEvp5Jy.945@netcom.com> <6o0lv1$47o$1@nnrp1.dejanews.com> <bksEvsppn.Jq0@netcom.com> In article <bksEvsppn.Jq0@netcom.com>, bks@netcom.com (Bradley K. Sherman) wrote: > In article <6o0lv1$47o$1@nnrp1.dejanews.com>, <nreed@indiana.edu> wrote: > >> I would say that C is still, in some sense, more portable than Java. > > > >C more portable than Java? That's a good one. (You're joking, right?) > > > > In the sense that C is more stable. Java, in principle, is more > portable, but in practice there are problems. I am not bashing > Java here, so don't take this wrong. If there are not problems > with Java portability, what is Sun's gripe with MicroSoft? > Java can be seen as two elements. The first is the programming language and the second is the platform. The platform consists of all the portable classes and the Java code being in byte form. Microsoft is using the programming language part of it for implementing Windows only applications. In certain sense the is fine, if you see Java as a C++ or Visual Basic replacement, though although you solve the problems of C++ or VB you are still left with something that is not portable. The thing is most people, and Sun for that matter, when they think of Java, and its whole reason for existance, is as a way of being able to write once run anywhere. For this to work the standard classes, on which the vast majority of Java applications rely, need to be available in all the JVM. Take that away and there is no way the Java application can run on any platform, in esence this what Microsoft has apparently done on their platform (I am not up to date on the specifics). I hope this explanation helps you understand why Microsoft's approach has got on people's nerves (Microsoft makes things non-cross-platorm in order to encourage people to use Windows (damn, I wish we could educate the world, sniff ;-) ). AJ -- http://www.bigfoot.com/~ajmas/ -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: Ralph Cook <rcook@pobox.com> Newsgroups: comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming Subject: Re: Programming and the colossal failure to advance Date: Fri, 10 Jul 1998 15:22:18 -0400 Organization: Ericsson, Inc. Message-ID: <35A669EA.B06434BF@pobox.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <vhIsdqY67dTD-pn2-dGSKqdjMrSfd@dynamic47.pm03.san-rafael.best.com> <bksEvp5Jy.945@netcom.com> <6o0lv1$47o$1@nnrp1.dejanews.com> <bksEvsppn.Jq0@netcom.com> <6o5ods$k6$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit ajmas@bigfoot.com wrote: > <somebody else wrote...> > > In the sense that C is more stable. Java, in principle, is more > > portable, but in practice there are problems. I am not bashing > > Java here, so don't take this wrong. If there are not problems > > with Java portability, what is Sun's gripe with MicroSoft? > > > ... > something that is not portable. The thing is most people, and Sun for > that > matter, when they think of Java, and its whole reason for existance, > is as a > way of being able to write once run anywhere. For this to work the > standard > classes, on which the vast majority of Java applications rely, need to > be > available in all the JVM. Take that away and there is no way the Java > application can run on any platform, in esence this what Microsoft has > > apparently done on their platform (I am not up to date on the > specifics). > > I hope this explanation helps you understand why Microsoft's approach > has > got on people's nerves (Microsoft makes things non-cross-platorm in > order > to encourage people to use Windows (damn, I wish we could educate the > world, > sniff ;-) ). > > AJ I think this is correct as far as philosophy goes. On the more legal side, Sun licenses Java with certain restrictions concerning the language, many of them having to do with this philosophy. They are taking Microsoft to court on the argument that Microsoft has violated that agreement. So it isn't just Sun whining that Microsoft is doing something that Sun doesn't like; Sun's argument is that, in order to call it Java, it must conform to the licensing agreement, and that Microsoft has violated it. I think this is correct as far as philosophy goes. On the more legal side, Sun licenses Java with certain restrictions concerning the language, many of them having to do with this philosophy. They are taking Microsoft to court on the argument that Microsoft has violated that agreement. So it isn't just Sun whining that Microsoft is doing something that Sun doesn't like; Sun's argument is that, in order to call it Java, it must conform to the licensing agreement, and that Microsoft has violated it. rc -- When I do speak for my company, I list my title and use "we".
From: "Mike Anderson" <mikea@knowmed.com> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: Fri, 10 Jul 1998 12:48:26 -0700 Organization: The World's Usenet -- http://www.Supernews.com Message-ID: <6o5r90$1a6$1@supernews.com> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> <rex-0807980230180001@cc497300-a.wlgrv1.pa.home.com> <slrn6q81lj.j00.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu> <rex-0907980206430001@cc497300-a.wlgrv1.pa.home.com> <6o52fj$tvg$1@nnrp1.dejanews.com> tj_sam@my-dejanews.com wrote in message <6o52fj$tvg$1@nnrp1.dejanews.com>... >Again that problem of typecasting programming as a language, >perhaps its better to resolve a new definition for programming and >what programs are: [snip] >Programmers design virtual machines (generic term - not Java) with >loops (wheels) and conditional branches (levers) held in >subroutines (structural frameworks) in the "virtual world". This mechanical analogy works for me but I don't see how it in any way escapes the "programming language" model. I consider diagrams, blueprints, object-object interaction, human-computer interaction, etc. to be expressions of a language. Some languages are more well defined than others. I don't have a well thought out definition of "language", but off the top of my head I'd define it as a collection of symbols and a system rules that determine how the symbols can be arranged and manipulated. Can you give your definition of what is (or what isn't) language? I'm not asking for anything formal or comprehensive, a few examples of what is and isn't language could satisfy me. However you feel best to clarify this. Another thing: it seems like you are calling for a greater distinction between control structure and noncontrol structure parts of the program. Smalltalk goes the other way: there's no distinction between messages that do affect execution flow and messages that do not (though there is a return operator: '^' which is built in to the language). The concept of execution flow control is not within the language. It is part of the class libraries and new flow controlling methods can be created (which is not often done but most Smalltalkers don't hesitate to do so if it provides clarity or conciseness). That things are extensible at this level is part of the reason there is a comprehensive set of methods to iterate over a collection, which in turn helped enable Smalltalk's collection family of classes, which are incredibly useful (Java's attempt at duplicating Collections is not nearly as powerful). So 1: am I reading you right: you are suggesting a greater distinction between flow-control and non-flow-control parts and 2) is so, might this be the wrong direction? ...Mike
From: Greg_Anderson@afs.com (Gregory H. Anderson) Newsgroups: comp.sys.next.programmer Subject: Rhapsody beta testers sought Date: 10 Jul 1998 21:16:28 GMT Organization: Anderson Financial Systems Inc. Message-ID: <6o60bc$cq1@shelob.afs.com> This seemed like the best group to post this, since we programmers are the only ones who have the ability to run Rhapsody right now, eh? With the impending release of Mac OSX Server (formerly Rhapsody Customer Release 1), AFS is building a variety of architectures of our WriteUp word processing and PasteUp page layout products. Expected architectures include OS 4.2 (Motorola and Intel), Rhapsody (PowerPC and Intel), and Yellow Box for Windows. We do not plan to build OPENSTEP Enterprise for Windows (the one that was parallel to OS 4.2) unless there is specific demand for that older version. Let us know if that's important to you. If you would like to participate in this very brief test cycle, please respond to <greg@afs.com> with the following information: Full name Email address for notices (apps will be downloaded, not mailed) List of architectures you can test We do have the API working through Distributed Objects, so if your programming needs include access to word processing features, this might prove a great way to test that. Thanks! -- Gregory H. Anderson | "We're in the land of the blind, Visionary Ophthalmologist | selling working eyeballs, and they Anderson Financial Systems | balk at the choice of color." -- Tony greg@afs.com (NeXTmail OK) | Lovell, on Mac user reactions to NeXT
From: Eric Hermanson <eric@alum.mit.edu> Newsgroups: comp.sys.next.programmer Subject: Decompiling Java Classes. Wow. Date: Fri, 10 Jul 1998 14:34:48 -0700 Organization: Digital Universe Corporation Message-ID: <35A688F8.AC8C70D6@alum.mit.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NNTP-Posting-Date: 10 Jul 1998 21:37:22 GMT Maybe this is old news (in fact, I am sure it is old news), but not being an expert in compilers, I was very surprised to find that there are several public domain decompilers available that will literally generate java source files given *.class files. See: http://meurrens.ml.org/ip-Links/Java/codeEngineering/decomp.html Two of these decompilers, Mocha/Jasmine, and Jad, seem to work with amazing accuracy. I was able to decompile all class files I tried it on, including class files that ship with some major commercial products. I was able to gererate what looks like the original source files, in effect allowing anyone to reverse engineer any and all java bytecode. This seems like a major, major problem with Java. What vendor in their right mind would ship class files (or .zip or .jar files for that matter) when their competitors can easily reverse engineer their code (even though it may be "illegal" to do so)?? Is there a way to compile Java code so as to prevent decompilation? I read about something named "Crema" which is supposed to prevent decompilation by "Mocha", but then somebody came along and invented "Jasmine", which counteracts the effects of "Crema"! Also, what prevents someone from importing an existing class file (i.e. import com.microsoft.xxx.xxx) and calling methods on that commercial library from their own separate application? Eric
From: "heidi" <heidi@gestalt.org> Newsgroups: comp.sys.next.programmer Subject: connecting to sybase Date: Fri, 10 Jul 1998 16:52:48 -0600 Organization: Canada Connect Corp. Message-ID: <6o65ud$i3l$1@cleavage.canuck.com> We get the following error when trying to connect to Sybase running on NT host using EOModeler on Rhapsody for PC compatibles, DR2: The context allocation routine failed when it tried to load localization files!! One or more following problems may caused the failure Cannot access the sybase home directory, please check environment variable SYBASE or ~sybase Your sybase home directory is . Check the environment variable SYBASE if it is not the one you want! Cannot access file *‚^C Jul 10 16:19:22 Workspace: EOModeler was killed by signal 10 We have also set the environment variable with the following command: setenv SYBASE /usr/sybase The same environment variable is passed to the app from the Launcher window settings. Our Rhapsody root directory has folder /usr/sybase, with the interfaces file sitting there and pointing to the Sybase server. We have no problem connecting to the NT-based Sybase server from NEXTSTEP 3.3 environment, using DB-Library and the same interfaces file, and from another NT host using CT-Library. thanks, heidi
From: "Mike Anderson" <mikea@knowmed.com> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming Subject: Re: Programming and the colossal failure to advance Date: Fri, 10 Jul 1998 16:54:41 -0700 Organization: The World's Usenet -- http://www.Supernews.com Message-ID: <6o69rd$og4$1@supernews.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6nh384$4mj$1@nnrp1.dejanews.com> <359C16EA.5AF08B84@home.com> <slrn6q1udv.bko.rog@talisker.ohm.york.ac.uk> Roger Peppe wrote in message ... I agree with your theme of simplicity, but regarding this section: >for an example of a language which combines simplicity with complete >modularity along with great power, while avoiding the pitfalls of the >object-oriented crowd, check out Limbo from Bell Labs (and the original >creators of unix). it's with languages such as this (and environments >such as Inferno) that software useability (and re-useability) could >take a real leap forward. about Limbo - I just took a brief glance at the language docs at the site you pointed to. From what I could see it didn't look any simpler than Modula family languages. Of course I should dig into it more to really get a feel for it, but could you (or anyone) point out what qualities might bring that "real leap forward"? And what are the pitfalls of the object oriented crowd? (I'm an OO fan, but certainly not one of the crowd.) ...Mike
From: paipai@tin.it (Paolo Di Francesco) Newsgroups: comp.sys.next.programmer Subject: Re: Getting Started with Rhapsody Date: Sun, 05 Jul 1998 13:09:58 GMT Organization: Telecom Italia Net Message-ID: <359e13ba.1807258@news.tin.it> References: <andrewp-2706981427020001@ts5port13.port.net> <1998063022544100.SAA13890@ladder01.news.aol.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit On 30 Jun 1998 22:54:41 GMT, rafkah@aol.com (Rafkah) wrote: >I'm also enthousiastic about beginning programming with the Yellow Box. It >really sounds fantastic (at least when one sees Andrew Stone's sWord : >http://www.stone.com/dev/sWord/). Well, yellow box and ObjectiveC, are two great things. But, if you are NOW in the Apple market, you'll be entusiastic, if you are in the Intel market, these are a not-so-important things... >My own problem is that I have no experience in programming (in that way getting >started with the NeXT's tools seems to me as a great beginning to developping). try this: http://developer.apple.com/techpubs/rhapsody/System/Documentation/Developer/YellowBox/HomePages/HomePage.html I think you can find more docs in the same site, and navigating in the directories.... good luck!
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Program99, was (Re: Programming and the colossal failure to advance) Date: 10 Jul 98 12:50:32 Organization: Is a sign of weakness Message-ID: <SCOTT.98Jul10125032@slave.doubleu.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <35A0A008.41C6@pc-plus.de.remove.this.bit> <01bdac0f$f6d84f20$0fd29cce@kiki.graytechnologies.com> In-reply-to: "Stu"'s message of 10 Jul 1998 14:35:16 GMT In article <01bdac0f$f6d84f20$0fd29cce@kiki.graytechnologies.com>, "Stu" <stu@grayechnologies.com> writes: Stephen Riehm <sr@pc-plus.de.remove.this.bit>, wrote in article <35A0A008.41C6@pc-plus.de.remove.this.bit>... > The same should be true for programming. The representation on > the screen doesn't have to be the same as the representation on > disk. If we had code processors, programs could be treated like > hypertext documents, interactive - living - systems. I concur with this, but remember (assuming your old enough to remember) the issues that brought us to todays word processors, principally the incompatibilities. Today using a word processor on a document is pretty mindless. Just open the document with your favorite word processor and, in general, viola - you can edit your document. The reason we can do this is because the word processor makers have gone to the effort of supporting the majority of document formats in use today. The above paragraph pretty much destroys your credibility, in my eyes. You can only swap word processors with ease for the simplest documents. I've been involved for about five years with a project who's goal is to make it easy to build custom documents quickly, and one way you get the document is as a word processor format file. We've (well, one of us) has spent the past couple years working with WordPerfect and Word to accomplish this. The problem appears to be arbitrarily complex, because the 900lb gorilla has documented the _syntax_ of their file formats, but not the _semantics_. So we can emit files that are "correct" to our heart's content - but getting things to format precisely to customer's specifications can be a real nightmare. Sometimes even having a correctly formatted sample file doesn't help, because while we can see the differences, we can't figure out why they make a difference without moving beyond the published specs. The oddest thing is that WordPerfect seems to be tremendously more forgiving in this area. Not nearly as much voodoo required. For Word, though, we sometimes have to emit different output for different versions of Word! It's been such a nightmare that we're strongly pondering a move to XML as our output format. This doesn't so much fix the problem as give us an easy way to export the problem out of the main portions of the program. At least it won't be _our_ problem, then :-). Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: ell@access.digex.net (Ell) Newsgroups: comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.programming Subject: Re: Programming and the colossal failure to advance Date: Sat, 11 Jul 1998 06:21:45 GMT Organization: Universe Message-ID: <35aa0304.68013918@news.erols.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <vhIsdqY67dTD-pn2-dGSKqdjMrSfd@dynamic47.pm03.san-rafael.best.com> <bksEvp5Jy.945@netcom.com> <6o0lv1$47o$1@nnrp1.dejanews.com> <bksEvsppn.Jq0@netcom.com> <6o5ods$k6$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Isn't the assertion that software development is a colossal failure a falsehood? In terms of one aspect the high "failure" rate (which is ambiguous, and may be based on the wrong measurements--like initial project resource estimations) mirrors the failure rate in other fields thought to be "successful". (Someone said this on Usenet 3 or 4 years ago, but I can't recall what fields they noted.) Elliott -- :=***=: Objective * Pre-code Modelling * Holistic :=***=: Hallmarks of the best SW Engineering "The domain object model is the foundation of OOD." Check out SW Modeller vs SW Craftite Central : www.access.digex.net/~ell Copyright 1998 Elliott. exclusive of others' writing. may be copied without permission only in the comp.* usenet and bitnet groups.
Newsgroups: comp.sys.next.programmer From: leffert@cs.uchicago.edu (Jonathan B. Leffert) Subject: Objective C woes Message-ID: <leffert.900164607@cs.uchicago.edu> Sender: news@midway.uchicago.edu (News Administrator) Organization: University of Chicago -- Academic Computing Services Date: Sat, 11 Jul 1998 13:43:27 GMT I've been attempting to do some IB/PB work under OpenStep Mach 4.2. I however had run into difficulties in the area of classes I have defined to work with my IB classes for the interface. So, I attempted to complete a simple class as I have not worked with Objective C in about a year and it's coming back fairly slowly. I defined a simple class (aptly named foo) as follows: -- foo.h -- #import <AppKit/AppKit.h> @interface foo : NSObject { id bar; } - init; - (void) dealloc; - getBar; - (void) setBar: (id) b; @end -- foo.m -- #import "foo.h" @implementation foo - init { [super init]; return self; } - (void) dealloc { [bar release]; [super dealloc]; } - getBar { return bar; } - (void) setBar: (id) b { [bar autorelease]; bar = [b copy]; } @end Everything, I think, looks ok. Then, I created x.m to test the foo class: -- x.m -- #import "foo.h" int main() { id s = @"hello"; id f = [foo init]; [f setBar: s]; printf("%s\n",[[f getBar] cString]); return 0; } I then compiled as follows % cc -c foo.m x.m % cc -o x foo.o x.o /NextLibrary/Frameworks/AppKit.framework/AppKit And when i ./x, things do not exactly work correctly: kowalevsky% cc -c foo.m x.m kowalevsky% cc -o x foo.o x.o /NextLibrary/Frameworks/AppKit.framework/AppKit kowalevsky% ./x Jul 11 08:31:10 x[3130] *** _NSAutoreleaseNoPool(): Object 0xc64c of class NSInlineCString autoreleased with no pool in place - just leaking Jul 11 08:31:11 x[3130] *** +[foo setBar:]: selector not recognized Jul 11 08:31:11 x[3130] *** _NSAutoreleaseNoPool(): Object 0x14af4 of class NSInlineCString autoreleased with no pool in place - just leaking Jul 11 08:31:11 x[3130] *** _NSAutoreleaseNoPool(): Object 0x15070 of class NSCStringWithGap autoreleased with no pool in place - just leaking Jul 11 08:31:11 x[3130] *** _NSAutoreleaseNoPool(): Object 0xc97c of class NSException autoreleased with no pool in place - just leaking Jul 11 08:31:11 x[3130] *** _NSAutoreleaseNoPool(): Object 0xca04 of class NSConcreteMutableString autoreleased with no pool in place - just leaking Jul 11 08:31:11 x[3130] *** Uncaught exception: <NSInvalidArgumentException> *** +[foo setBar:]: selector not recognized stack: 0x18067c81 0x18067b3a 0x5087612 0x508785d 0x18064c23 0x1806381a 0x18063795 0x18063a60 0x1803da1c 0x5002065 0x3775 0x353e exiting! So the question is, what am I doing wrong? I think it's probably something small and stupid, but I've been staring at this code for far too long. Thanks, Jonathan -- Jonathan B. Leffert <leffert@cs.uchicago.edu> "Indeed it is generally the case that men are readier to call rogues clever than simpletons honest, and are as ashamed of being the second as they are proud of being the first." -- Thucydides, The Peloponnesian War
Newsgroups: comp.sys.next.programmer From: haller@nowhere.romulus.uwaterloo.ca (Kirk Haller) Subject: Re: Objective C woes Sender: news@undergrad.math.uwaterloo.ca (news spool owner) Message-ID: <Evxvyw.1Mu@undergrad.math.uwaterloo.ca> Cc: leffert@cs.uchicago.edu Date: Sat, 11 Jul 1998 16:47:20 GMT Content-Transfer-Encoding: 7bit References: <leffert.900164607@cs.uchicago.edu> Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Organization: University of Waterloo In <leffert.900164607@cs.uchicago.edu> Jonathan B. Leffert wrote: > I've been attempting to do some IB/PB work under OpenStep Mach 4.2. I > however had run into difficulties in the area of classes I have defined to > work with my IB classes for the interface. So, I attempted to complete a > simple class as I have not worked with Objective C in about a year and it's > coming back fairly slowly. I defined a simple class (aptly named foo) as > follows: > > -- foo.h -- > #import <AppKit/AppKit.h> > > @interface foo : NSObject > { > id bar; > } > > - init; > - (void) dealloc; > - getBar; > - (void) setBar: (id) b; > > @end > > -- foo.m -- > #import "foo.h" > > @implementation foo > > - init > { > [super init]; > return self; > } > > - (void) dealloc > { > [bar release]; > [super dealloc]; > } > > - getBar > { > return bar; > } > > - (void) setBar: (id) b > { > [bar autorelease]; > bar = [b copy]; > } > > @end > > Everything, I think, looks ok. Then, I created x.m to test the foo class: > > -- x.m -- > #import "foo.h" > > int main() > { > id s = @"hello"; > id f = [foo init]; > [f setBar: s]; > printf("%s\n",[[f getBar] cString]); > return 0; > } > > I then compiled as follows > > % cc -c foo.m x.m > % cc -o x foo.o x.o /NextLibrary/Frameworks/AppKit.framework/AppKit > > And when i ./x, things do not exactly work correctly: > > kowalevsky% cc -c foo.m x.m > kowalevsky% cc -o x foo.o x.o > /NextLibrary/Frameworks/AppKit.framework/AppKit > kowalevsky% ./x > Jul 11 08:31:10 x[3130] *** _NSAutoreleaseNoPool(): Object 0xc64c of class > NSInlineCString autoreleased with no pool in place - just leaking > Jul 11 08:31:11 x[3130] *** +[foo setBar:]: selector not recognized > Jul 11 08:31:11 x[3130] *** _NSAutoreleaseNoPool(): Object 0x14af4 of class > NSInlineCString autoreleased with no pool in place - just leaking > Jul 11 08:31:11 x[3130] *** _NSAutoreleaseNoPool(): Object 0x15070 of class > NSCStringWithGap autoreleased with no pool in place - just leaking > Jul 11 08:31:11 x[3130] *** _NSAutoreleaseNoPool(): Object 0xc97c of class > NSException autoreleased with no pool in place - just leaking > Jul 11 08:31:11 x[3130] *** _NSAutoreleaseNoPool(): Object 0xca04 of class > NSConcreteMutableString autoreleased with no pool in place - just leaking > Jul 11 08:31:11 x[3130] *** Uncaught exception: > <NSInvalidArgumentException> *** > +[foo setBar:]: selector not recognized > stack: 0x18067c81 0x18067b3a 0x5087612 0x508785d 0x18064c23 0x1806381a > 0x18063795 0x18063a60 0x1803da1c 0x5002065 0x3775 0x353e > exiting! > > So the question is, what am I doing wrong? I think it's probably something > small and stupid, but I've been staring at this code for far too long. > > Thanks, > > Jonathan > -- Kirk Haller haller@romulus.uwaterloo.ca ************************************************************ Computer Graphics Lab Department of Computer Science University of Waterloo Waterloo, ON Canada N2L 3G1 ************************************************************
Newsgroups: comp.sys.next.programmer From: haller@nowhere.romulus.uwaterloo.ca (Kirk Haller) Subject: Re: Objective C woes Sender: news@undergrad.math.uwaterloo.ca (news spool owner) Message-ID: <EvxwAt.o1z@undergrad.math.uwaterloo.ca> Cc: leffert@cs.uchicago.edu Date: Sat, 11 Jul 1998 16:54:28 GMT Content-Transfer-Encoding: 7bit References: <leffert.900164607@cs.uchicago.edu> Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Organization: University of Waterloo In <leffert.900164607@cs.uchicago.edu> Jonathan B. Leffert wrote: > I've been attempting to do some IB/PB work under OpenStep Mach 4.2. I > however had run into difficulties in the area of classes I have defined to > work with my IB classes for the interface. So, I attempted to complete a > simple class as I have not worked with Objective C in about a year and it's > coming back fairly slowly. I defined a simple class (aptly named foo) as > follows: Sorry about the last message, I hit the wrong button. I think that your problem is that you have not initialized the autorelease pool. This is normally done with an invocation of NSApplication. Since you are writing your own main, you will have to set up the pool yourself. Here is the relevant piece from the on-line documentation of the NSApplication object: The NSApplication class sets up autorelease pools (instances of the NSAutoreleasePool class) during initialization and inside the event loop specifically, within its init (or sharedApplication) and run methods. Similarly, the methods that the Application Kit adds to NSBundle employ autorelease pools during the loading of nib files. These autorelease pools aren't accessible outside the scope of the respective NSApplication and NSBundle methods. Typically, an application creates objects either while the event loop is running or by loading objects from nib files, so this usually isn't a problem. However, if you do need to use OpenStep classes within the main() function itself (other than to load nib files or to instantiate NSApplication), you should create an autorelease pool before using the classes and then release the pool when you're done. For more information, see the NSAutoreleasePool class specification in the Foundation Framework Reference. Hope that helps -- Kirk Haller haller@romulus.uwaterloo.ca ************************************************************ Computer Graphics Lab Department of Computer Science University of Waterloo Waterloo, ON Canada N2L 3G1 ************************************************************
Newsgroups: comp.sys.next.programmer From: ergo@camelot.in-berlin.de (Olaf Foellinger) Subject: Re: Objective C woes Content-Type: text/plain; charset=iso-8859-1 Message-ID: <EvyDBF.59w@camelot.in-berlin.de> Sender: news@camelot.in-berlin.de Content-Transfer-Encoding: 8bit Cc: leffert@cs.uchicago.edu Organization: none References: <leffert.900164607@cs.uchicago.edu> Mime-Version: 1.0 Date: Sat, 11 Jul 1998 23:02:02 GMT In <leffert.900164607@cs.uchicago.edu> Jonathan B. Leffert wrote: > [Source code deleted] > > kowalevsky% cc -c foo.m x.m > kowalevsky% cc -o x foo.o x.o > /NextLibrary/Frameworks/AppKit.framework/AppKit > kowalevsky% ./x > Jul 11 08:31:10 x[3130] *** _NSAutoreleaseNoPool(): Object 0xc64c of class > NSInlineCString autoreleased with no pool in place - just leaking > Jul 11 08:31:11 x[3130] *** +[foo setBar:]: selector not recognized > Jul 11 08:31:11 x[3130] *** _NSAutoreleaseNoPool(): Object 0x14af4 of class > NSInlineCString autoreleased with no pool in place - just leaking > Jul 11 08:31:11 x[3130] *** _NSAutoreleaseNoPool(): Object 0x15070 of class > NSCStringWithGap autoreleased with no pool in place - just leaking > Jul 11 08:31:11 x[3130] *** _NSAutoreleaseNoPool(): Object 0xc97c of class > NSException autoreleased with no pool in place - just leaking > Jul 11 08:31:11 x[3130] *** _NSAutoreleaseNoPool(): Object 0xca04 of class > NSConcreteMutableString autoreleased with no pool in place - just leaking > Jul 11 08:31:11 x[3130] *** Uncaught exception: > <NSInvalidArgumentException> *** > +[foo setBar:]: selector not recognized > stack: 0x18067c81 0x18067b3a 0x5087612 0x508785d 0x18064c23 0x1806381a > 0x18063795 0x18063a60 0x1803da1c 0x5002065 0x3775 0x353e > exiting! > > So the question is, what am I doing wrong? I think it's probably something > small and stupid, but I've been staring at this code for far too long. Object allocation and releasing depends on an existing autorelease pool (see the docs). Try to add one by inserting #import "foo.h" int main() { NSAutoreleasePool * pool = [[NSAutoreleasePool alloc] init]; id s = @"hello"; id f = [foo init]; [f setBar: s]; printf("%s\n",[[f getBar] cString]); [pool release]; return 0; } Greetings Olaf -- Olaf Foellinger NeXTMail & MIME welcome!
From: uli@tallowcross.uni-frankfurt.de (Uli Zappe) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.programmer Subject: Desperately seeking SPARC users Date: 12 Jul 1998 01:27:53 GMT Organization: NEXTTOYOU Message-ID: <6o93ep$88o$1@tallowcross.uni-frankfurt.de> NNTP-Posting-Date: 12 Jul 1998 01:27:53 GMT Hi, anybody out here that still uses NS/OS on SPARC? If so, *please* contact me! To compile an app 4fat, I need a certain information from a SPARC user. Thank you! Bye Uli -- _____________________________________________________________________ Uli Zappe E-Mail: uli@tallowcross.uni-frankfurt.de (NeXTMail,Mime,ASCII) PGP on request Lorscher Strasse 5 WWW: - D-60489 Frankfurt Fon: +49 (69) 9784 0007 Germany Fax: +49 (69) 9784 0042 staff member of NEXTTOYOU - the German NEXTSTEP/OPENSTEP magazine _____________________________________________________________________
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <668899611234@digifix.com> Date: 12 Jul 1998 03:48:54 GMT Organization: Digital Fix Development Message-ID: <26220900216033@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1993. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: tj_sam@my-dejanews.com Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: Sun, 12 Jul 1998 13:15:30 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6oacti$5o4$1@nnrp1.dejanews.com> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> <rex-0807980230180001@cc497300-a.wlgrv1.pa.home.com> <slrn6q81lj.j00.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu> <rex-0907980206430001@cc497300-a.wlgrv1.pa.home.com> <6o52fj$tvg$1@nnrp1.dejanews.com> <6o5r90$1a6$1@supernews.com> In article <6o5r90$1a6$1@supernews.com>, "Mike Anderson" <mikea@knowmed.com> wrote: >>would be represented as 2D (X,Y) wire frame objects. If subroutine >>or function calls are represented as Z plane bi-directional vector >>(recursive function are outlawed) >mikea@knowmed.com wrote in message >The most notably missing thing is abstraction. GUIs are impaired >because oftheir inability to handle abstraction, but it's not much of >a problem because people don't expect much of them. Programming >without abstraction would be a very tedious excersise (but it would in >some sense be simpler). >Abstraction is a powerful tool, and it should be refined, not ignored >or outlawed. I would agree, in programming abstractions create a simplification. However it must be understood (assuming OOP) that abstractions only mask an internal complex, with or without abstraction (classes) the overall complexity of any representation remains the same. A point here, class structures seem to work when dealing with internalize computer representation (MFC) however in the "real world" use of complex classes (an airline booking system) breaks down totally. IMO this difference is caused by the fact that computers are definable as consistent mechanical system, while "real world" representation have a complexity and interaction that defeats any attempt to define a constant abstraction. ------------------------------- In sorry, I really didn't give a proper perspective. The VR editor with its hybrid graphic-textural source code is a "bridge", a small part of the actual system and does not define the possibilities for VR programming. I emphasized it because it's a concrete factor, I have a simple working model. Assume that the 3D source code could be encapsulated into a sphere. Only then do you entire a VR programming environment. Touch the sphere and it should project out "Hello World" in an attached IO cone. If it doesn't work: open up the sphere and using a GUI styled 3D rendered debug and VR editor -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.pr Subject: Re: Programming and the colossal failure to advance Date: 12 Jul 1998 14:06:43 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6qhgnn.ih4.rog@talisker.ohm.york.ac.uk> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6nh384$4mj$1@nnrp1.dejanews.com> <359C16EA.5AF0 <6o69rd$og4$1@supernews.com> On Fri, 10 Jul 1998 16:54:41 -0700, "Mike Anderson" <mikea@knowmed.com> wrote: > >for an example of a language which combines simplicity with complete > >modularity along with great power, while avoiding the pitfalls of the > >object-oriented crowd, check out Limbo from Bell Labs (and the original > >creators of unix). it's with languages such as this (and environments > >such as Inferno) that software useability (and re-useability) could > >take a real leap forward. > > about Limbo - I just took a brief glance at the language docs at the site > you pointed to. From what I could see it didn't look any simpler than > Modula family languages. Of course I should dig into it more to really get > a feel for it, but could you (or anyone) point out what qualities might > bring that "real leap forward"? i've heard good reports of the modula family of languages, but i only have a small amount of experience (with modula 1) so i can't really comment in that respect. i find limbo a particularly simple language because all the language semantics seem completely natural while remaining powerful. something as simple as the declaration syntax, which allows the type to be ommitted, while retaining type security, makes an enormous amount of difference to the "feel" of the language. i've done a fair amount of thinking about this, and i think that one feature stands out above the rest: lack of unpredicable side-effects. if you see a bit of limbo code (and the same is true of C, to a some extent) you know precisely what is guaranteed by the language and when you're going to have to refer to the library documentation. for an example of what i mean, take Pascal as an example. in pascal, the way you call a procedure with no arguments is just to name it: e.g. result := procedurename; all very well until you come to maintain the code, and you're not quite sure whether _procedurename_ is a function, or a variable. the difference is only a minor syntactic one, but this one change makes a large difference when trying to work out what a piece of code is doing. ideally, when you've located a bug in a large piece of software, you should be able to modify the piece of code in question while being *sure* that you're not affecting the rest of the code. i.e. it's very important that you can manipulate sections of code while maintaining external invariants. some culprits in this regard: C++ overloading operators. you never know quite what the result of "a=b+c;" is going to be. call by reference you can't tell without referring to the declaration whether a variable might be changed. arbitrary pointers to objects once you've taken the address of a variable, you're never quite sure whether the variable might be changed. methods overriden by a subclass it's often not possible to tell if the original semantics of the method have been preserved, and whether *all* the superclass methods are still valid. i find that it's very easy to reason about limbo programs because it's been designed with this as an aim. it hasn't gone all the way to that extreme (not like Occam, for example, which was mathematically semantically proven but had some limitations as a result) but the result is a language that feels nice and in which code largely works independently of its context (a consequence of enforced modularity; no global variables across modules, a strictly hierarchical module structure, and a lovely set of well-defined language primitives) > And what are the pitfalls of the object oriented crowd? (I'm an OO fan, but > certainly not one of the crowd.) the idea that subclassing (and the class hierarchy in general) is a genuine boon, rather than a hack. the idea that functions are somehow primitive (even though they are actually far cleaner than objects) the lack of local, but persistent per-object variables in some widely used OO languages. that'll do for now! cheers, rog.
From: blenko-tom@cs.yale.edu (Tom Blenko) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 12 Jul 1998 10:39:23 -0400 Organization: Yale University Computer Science Dept., New Haven, CT 06520-2158 Distribution: world Message-ID: <6oahqrINNrkm@RA.DEPT.CS.YALE.EDU> Matthew M. Huntbach writes | ... | Control flow is very much tied to the old fashioned von Neumann model of | the computer. It ought to be the first thing that is thrown away as | peripheral when one is truing to move to a more abstract model. Perhaps I'm overreading your comments, but it's far from clear to me that control flow can be "thrown away", and I also don't understand what is "old-fashioned" about the von Neumann model -- do you know of some alternative model that haven't been throughly discredited at this point? Tom
From: james@unspam.edu (James McCartney) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: Sun, 12 Jul 1998 10:11:07 -0700 Organization: The University of Texas at Austin, Austin, Texas Distribution: world Message-ID: <james-ya023180001207981011070001@news> References: <6oahqrINNrkm@RA.DEPT.CS.YALE.EDU> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <6oahqrINNrkm@RA.DEPT.CS.YALE.EDU>, blenko-tom@cs.yale.edu (Tom Blenko) wrote: >Matthew M. Huntbach writes >| ... >| Control flow is very much tied to the old fashioned von Neumann model of >| the computer. It ought to be the first thing that is thrown away as >| peripheral when one is truing to move to a more abstract model. > >Perhaps I'm overreading your comments, but it's far from clear to me >that control flow can be "thrown away", and I also don't understand >what is "old-fashioned" about the von Neumann model -- do you know of >some alternative model that haven't been throughly discredited at this >point? Well you write from a CS dept, so I'd assume you know the following anyway. Surely you don't think lazy FP has been "throughly discredited" at this point. Lazy functional languages order of evaluation is demand driven. The order of operations is not written explicitly by the programmer but is determined by the runtime structure of the problem. This allows one to program by specifying the problem and not how to do it step by step, which is quite liberating. It has been reported that programs written in lazy FPs look, to those unfamiliar with how they work, like only a description of the problem and not a solution. When sequential order of operations are necessary there are ways to do it. Uhh I'll leave it at that.. (tip toeing so as not to wake the sleeping Monad.. :) --- James McCartney james@clyde.as.utexas.edu If you have a PowerMac check out SuperCollider, a real time synth program: http://www.audiosynth.com
From: macghod@concentric.net (Steve Sullivan) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.system,comp.sys.next.programmer,comp.sys.mac.programmer.codewarrior Subject: Apple heading into obscurity? part deux Date: Sun, 12 Jul 1998 18:12:33 -0700 Organization: Great till Apple got rid of the newton Message-ID: <macghod-1207981812330001@sdn-ar-001casbarp113.dialsprint.net> Apple has a page for the wanna be programmer. So you wanna be a mac programmer, ey? These directions mostly apply to a newby programmer, who, if follows these directions wont have any meaningful software out for a year or two. So what is Apple's future a year or two from now? NO YELLOWBOX. These directions basically apply to people who wont write full mac programs for a year or two, and you will notice their is not one word about yellow box. This page contains an introduction to Macintosh programming. Developers new to the Macintosh platform will find many useful links, resources and suggestions for starting points. Last updated 22-June-98 Things you should know before you start on your new life as a Macintosh programmer... It is helpful to have some programming experience, ideally with C or Pascal, and to be able to understand reading Pascal source since much of Inside Macintosh is in Pascal, some is in C. It is possible to learn to program for a Macintosh without prior programming experience. It requires learning basic programming syntax and fundamentals before you learn Macintosh-specific features. Be aware of the concepts of GUI programming, such as: windows, dialogs, menus, buttons, etc. Things you should have before you start your new life as a Macintosh programmer... A Power Macintosh or 68040 based Macintosh (A Power Macintosh is recommended) A development environment such as CodeWarrior or MPW Reference Material, such as Inside Macintosh Things you'll learn along your way as a new Macintosh programmer... Apple Events Font Management Macintosh Toolbox Memory Management QuickDraw Step 1. You need a Macintosh You can buy a Macintosh from the Apple Online Store if you don't already have one. A good entry level Power Macintosh, is the G3 233mhz with 64 megs of RAM, and 4 gigs of disk. Step 2. Get a Development Environment In order to actually write, compile and run your code, you will need a compiler and a debugger. I know some of you might be thinking, "Those are really expensive!" Well MPW is available free from Apple Computer. Basically, you have the choice of MPW or CodeWarrior. Step 3. Get some instruction Read through the information available through the "Documentation" section to start learning programming on the Macintosh. Macintosh C, is a good starting point if you are new to programming. Apple Computer's Inside Macintosh contains all the Macintosh-specific documentation for both hardware and software. Tools Programming, like any other task in which you build something, requires that you have the right tools for the job. Without the right tools, you'll spend a lot of time and effort that would be better spent on your application. There are hundreds of applications available in the marketplace that you can use to develop for the Macintosh. We have attempted to provide links to the most commonly used tools. Many of the tools listed are available to students through educational programs by the companies themselves, or through your local college. Compilers MPW: Apple's Macintosh Programmers Workshop CodeWarrior: Metrowerks's Development Environment| Editors and Debuggers BBEdit: A good text editor Resorcerer: A resource editor ResEdit: A resource editor MacsBug: A debugger Spotlight and QC: Memory debugger, and stress tester MemoryMime: A debugger Smartheap: A debugger MacNosy and The Debugger: A debugger SDKs and more Apple Tools: Development tools from Apple Apple SDK's: Apple Software Development Kits Sample Code Starting with some sample code that is similar in function to what you need, or using sample code to learn programming style and function is never a bad plan. You're time is valuable; use the lessons and experience of others to help you learn the basics. Apple's official sample code and sample snippets of code alt.sources.mac: The entire alt.sources.mac archive MacTech's Code Archive: MacTech's source archive Documentation There are many good books and other documentation sources for programming the Macintosh. The Metrowerks CodeWarrior reference CD has several good books (in electronic form) on it. The Usenet newsgroups and mailing lists are frequently watched by knowledgeable Macintosh programmers who are happy to answer questions. Inside Macintosh: Official reference site for MacOS Apple's Technotes Apple's Q&As Macintosh C: A Hobbyist's guide to programming a Mac MIT's Porting to PowerMac page: Porting to the Macintosh Programming with CodeWarrior: Metrowerks' Discover site for new programmers Mac Games FAQ Foundations of Mac Programming Metrowerks CodeWarrior Programming for the Mac Newsgroups When all else fails, ask your question on a newsgroup or mailing list that seems to fit your topic and you will be surprised at the response you will get. comp.sys.mac.programmer.info comp.sys.mac.programmer.help comp.sys.mac.programmer.tools comp.sys.mac.programmer.codewarrior comp.sys.mac.programmer.games comp.sys.mac.programmer.misc comp.sys.mac.oop.macapp3 comp.sys.mac.oop.powerplant comp.sys.mac.oop.tcl comp.sys.mac.oop.misc comp.programming Mailing Lists Apple Users Group Bulletin Apple Script Implementers Apple Script Users Cocoa Developers ColorSync Developers Mac Games Developers MPW Developers MRJ Developers QuickTime Developers QuickTime-VR Developers All Apple sponsored mailing lists -- So many pedestrians, so little time.
From: blenko-tom@cs.yale.edu (Tom Blenko) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 13 Jul 1998 09:33:15 -0400 Organization: Yale University Computer Science Dept., New Haven, CT 06520-2158 Distribution: world Message-ID: <6od2arINNq98@RA.DEPT.CS.YALE.EDU> |>| Control flow is very much tied to the old fashioned von Neumann model |>| of the computer. It ought to be the first thing that is thrown away |>| as peripheral when one is truing to move to a more abstract model. |> |>Perhaps I'm overreading your comments, but it's far from clear to me |>that control flow can be "thrown away", and I also don't understand |>what is "old-fashioned" about the von Neumann model -- do you know of |>some alternative model that haven't been throughly discredited at |>this point? |Well you write from a CS dept, so... Surely you don't think lazy FP |has been "throughly discredited" at this point. I haved worked with non-strict functional languages for the past 13 years. It never crossed my mind that anyone would cite them in a response. The claim that functional programs can be written thinking about what is to be computed but not how -- a claim which was popular in the FP community in the 1980's -- has been strongly discredited. "Thoroughly" is probably a bit too much because the idea was that the compiler writers would solve the problems, and the compiler writers haven't, but there are probably a few compilers yet to be written. The advocates are still around but these days I think you are more likely to hear them say that such languages "have a declarative semantics" or that one can "reason equationally" about them. In other words, real functional programmers have to worry about the operational semantics, including control flow, and it often is not a pretty sight. |When sequential order of operations are necessary there are ways to do |it. Uhh I'll leave it at that.. (tip toeing so as not to wake the |sleeping Monad.. :) Indeed, and no need for monads, this was done for a long time. That's what continuation-passing conversion, bracket abstraction, supercombinator abstraction, staging to compute with partial inputs, and, arguably, frequency reduction, program slicing, dead code elimination, and a few more are about. Which is so because they're all staging transformations (that's for Torben, who used to do staging transformations :-). Tom
From: mike@erix.ericsson.se (Mike Williams) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 13 Jul 1998 14:34:11 GMT Organization: Ericsson Telecom, Stockholm, Sweden Sender: mike@morgan (Mike Williams) Distribution: world Message-ID: <6od5t3$lv9$2@news.du.etx.ericsson.se> References: <6od2arINNq98@RA.DEPT.CS.YALE.EDU> NNTP-Posting-User: mike In article <6od2arINNq98@RA.DEPT.CS.YALE.EDU>, blenko-tom@cs.yale.edu (Tom Blenko) writes: |> I haved worked with non-strict functional languages for the past 13 |> years. It never crossed my mind that anyone would cite them in a |> response. Why not? We use Erlang (a strict functional language) to program several very large distributed control systems (> 300 KLOCS). |> The claim that functional programs can be written thinking about what |> is to be computed but not how -- a claim which was popular in the FP |> community in the 1980's -- has been strongly discredited. "Thoroughly" |> is probably a bit too much because the idea was that the compiler |> writers would solve the problems, and the compiler writers haven't, but |> there are probably a few compilers yet to be written. The advocates |> are still around but these days I think you are more likely to hear |> them say that such languages "have a declarative semantics" or that one |> can "reason equationally" about them. I can't say that being able to "reason equationally" or that they "have a declarative semantics" thrills me that much. What does interest me is that we have metrics which show that the source code of programs written in Erlang is about one fourth the size of equivalent software written in C++. We also have reliable metrics of (total_lines_of_code) / (total_time_taken_for_whole_project) which shows that an Erlang line of code costs roughly the same as a C++ line of code. Note that the total_time_taken_for_whole_project includes specification, programming, integration, testing, system testing etc etc. Metrics of fault intensities per line of code are also equivalent. The conclusion is obvious, to build a large system, it cost one quarter of the (human) resources if you use a functional language like Erlang instead of C++ and the quality is four times better. The ecconomic consequences of this are astounding! Without going into details (I will if asked :-), the reason a program in Erlang is shorter than the corresponding program in C++ is that a lot of extraneous details can be omitted from the Erlang version. Which brings us a lot closer to "thinking about what is to be computed but not how" even if we still have some way to go. /Mike -- mike@erix.ericsson.se ++46 8 7197855 Ericsson Telecom AB, Dept ÄT2/ETX/DN/S, S 126 25 Stockholm, Sweden
From: eiffelpgmr@nni.com (Tom Morrisette) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng,comp.pr Subject: Re: Limbo Date: Mon, 13 Jul 1998 02:14:22 GMT Message-ID: <35a96b44.3721574@news.nni.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <6nh384$4mj$1@nnrp1.dejanews.com> <359C16EA.5AF0 <6o69rd$og4$1@supernews.com> <slrn6qhgnn.ih4.rog@talisker.ohm.york.ac.uk> Many thanks for calling Limbo to our attention. I never heard of it, and it sounds very intriguing. I do disagree with you 100% about the point you make below. I dislike C and C++ precisely because (among many other things) their syntax forces a distinction between attributes and methods. It seems to me that software development is primarily about managing complexity, and that managing complexity is largely acheived via careful factoring and information hiding. Imho, there is no reason why the code below should care (and lots of reason it shouldn't care) whether procedurename is data or logic. This approach implies, of course, a discipline of avoiding the kinds of hidden side effects you refer to. Have you ever looked at Eiffel? I think its builtin, inheritable preconditions, postconditions, and class invariants help to address many of your concerns. On 12 Jul 1998 14:06:43 GMT, rog@ohm.york.ac.uk (Roger Peppe) wrote: >for an example of what i mean, take Pascal as an example. in pascal, >the way you call a procedure with no arguments is just to name it: > >e.g. > result := procedurename; > >all very well until you come to maintain the code, and you're not quite >sure whether _procedurename_ is a function, or a variable. the >difference is only a minor syntactic one, but this one change makes a >large difference when trying to work out what a piece of code is >doing. >
From: mmh@dcs.qmw.ac.uk (Matthew M. Huntbach) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Followup-To: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Date: 13 Jul 1998 14:23:21 GMT Organization: Queen Mary & Westfield College, London, UK Distribution: world Message-ID: <6od58p$818$2@beta.qmw.ac.uk> References: <6od2arINNq98@RA.DEPT.CS.YALE.EDU> Tom Blenko (blenko-tom@cs.yale.edu) wrote: > |>| Control flow is very much tied to the old fashioned von Neumann model > |>| of the computer. It ought to be the first thing that is thrown away > |>| as peripheral when one is truing to move to a more abstract model. > |> > |>Perhaps I'm overreading your comments, but it's far from clear to me > |>that control flow can be "thrown away", and I also don't understand > |>what is "old-fashioned" about the von Neumann model -- do you know of > |>some alternative model that haven't been throughly discredited at > |>this point? > |Well you write from a CS dept, so... Surely you don't think lazy FP > |has been "throughly discredited" at this point. > I haved worked with non-strict functional languages for the past 13 > years. It never crossed my mind that anyone would cite them in a > response. I have worked with concurrent logic languages for the past 12 years, and am happy with a language where you don't have a control flow giving you a precise order in which every computation is done. If it doesn't matter to the computation which order its data is computed, then it can be computed in any order or concurrently. > The claim that functional programs can be written thinking about what > is to be computed but not how -- a claim which was popular in the FP > community in the 1980's -- has been strongly discredited. That is not what I was claiming. Sure you need to have an idea of the subcomputations that will make the full computation. But you don't need a precise ordering on them in the way that is implied by the idea of "control flow". Matthew Huntbach
From: mmh@dcs.qmw.ac.uk (Matthew M. Huntbach) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Followup-To: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Date: 13 Jul 1998 14:28:16 GMT Organization: Queen Mary & Westfield College, London, UK Distribution: world Message-ID: <6od5i0$818$3@beta.qmw.ac.uk> References: <6od4ceINNros@RA.DEPT.CS.YALE.EDU> Tom Blenko (blenko-tom@cs.yale.edu) wrote: > Matthew Huntbach wrote: > |We are now in the era of networks and distributed computers. Computing > |is no longer about a big standalone mainframe... > |Why impose control flow when it is not necessary? Why not do some > |computations simultaneously with others? There is no need to have a > |separate concept of control flow - if a computation requires data from > |another suspend it until that data is ready. Data flow is sufficient. > I guess the brief answer, based on what I understand from people > who have worked on dataflow architectures, is that their communication > requirements, e.g., for resolving units of computation across a > bus/network, cause the system to saturate. So if you look at the > bus/network as an extended memory, you're back to the von Neumann > bottleneck. > Perhaps you had something more specific in mind. Well it wasn't dataflow architectures. I wasn't talking about hardware. Concurrent logic languages don't require special hardware. They could be mapped onto networks of distributed processors and execute in real parallel, or they could be implemnted on standard von Neumann machines in which case while there's no real parallelism the pseudo-parallelism of concurrency can be helpful in making programs clearer - what after all is object-oriented programming but pseudo-parallelism - lots of objects running in parallel. It's only messed up by the unnecessary control flow imposed on it by von Neumann oriented languages. Matthew Huntbach
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: Java 'bundles'? Newsgroups: comp.sys.next.programmer Message-ID: <35aa1cb0.0@news.depaul.edu> Date: 13 Jul 98 14:41:52 GMT Folks, Is there a bundle-like concept in Java, which works with locally stored files, rather than net-downloaded files? I've heard of JAR files, and java beans, but I haven't really kept up to date on Java since 1996. Thanks, Jon -- "... and subpoenas for all." - Ken Starr
From: blenko-tom@cs.yale.edu (Tom Blenko) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 13 Jul 1998 10:08:14 -0400 Organization: Yale University Computer Science Dept., New Haven, CT 06520-2158 Distribution: world Message-ID: <6od4ceINNros@RA.DEPT.CS.YALE.EDU> |>| Control flow is very much tied to the old fashioned von Neumann |>| model of the computer. It ought to be the first thing that is thrown |>| away as peripheral when one is truing to move to a more abstract model. | |> Perhaps I'm overreading your comments, but it's far from clear to me |> that control flow can be "thrown away", and I also don't understand |> what is "old-fashioned" about the von Neumann model -- do you know of |> some alternative model that haven't been throughly discredited at |> this point? | |We are now in the era of networks and distributed computers. Computing |is no longer about a big standalone mainframe... |Why impose control flow when it is not necessary? Why not do some |computations simultaneously with others? There is no need to have a |separate concept of control flow - if a computation requires data from |another suspend it until that data is ready. Data flow is sufficient. I guess the brief answer, based on what I understand from people who have worked on dataflow architectures, is that their communication requirements, e.g., for resolving units of computation across a bus/network, cause the system to saturate. So if you look at the bus/network as an extended memory, you're back to the von Neumann bottleneck. Perhaps you had something more specific in mind. Tom
From: gmgraves@slip.net (George Graves) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.system,comp.sys.next.programmer,comp.sys.mac.programmer.codewarrior Subject: Re: Apple heading into obscurity? (the Twighlight Zone) Date: Mon, 13 Jul 1998 10:27:28 -0700 Organization: Graves Associates Message-ID: <gmgraves-1307981027280001@sf-pm5-24-88.dialup.slip.net> References: <macghod-1207981812330001@sdn-ar-001casbarp113.dialsprint.net> <6obs8b$hnk$1@morgoth.sfu.ca> <35A980D1.2371@globaldialog.com> In article <35A980D1.2371@globaldialog.com>, webnik@globaldialog.com wrote: > The Mac is already obscure. The real fear is about the Mac disappearing > completely. This is a scarey thought for me. > > Can you imagine? This newsgroup would only have Edwin and Joe Ragosta > posting in it? They would be the last Mac users left on the whole > planet. > > > Their Posts might go something like this: > ********************************************************************** > JR Post: "Let's play the Rizzo impersonator game. You pretend your > Farkny and I'll defend our beloved platform with a scathing attack on > your mindless trolls. Then I get to call you strawman and stuff like > that". > > Edwin Post: "Golly, Joe, why am I always pretending to be Rizzo. When do > I get to call YOU strawman? I wanna defend the Mac! Can I?... huh, Joe? > ....can I"? > *********************************************************************** > ....and so it would go on forever....in the Mac Advocacy Zone....... > > doo doo doo dee...doo doo doo dee...doo doo doo dee...doo doo doo dee... Better expand that list to include me. Long after Joe and Edwin switch over to "the dark side", I'll still be using and advocating Macintosh.. I'll stop when death overtakes me, not until. George Graves
From: "heidi" <heidi@gestalt.org> Newsgroups: comp.sys.next.programmer Subject: connecting to sybase Date: Mon, 13 Jul 1998 12:18:40 -0600 Organization: Canada Connect Corp. Message-ID: <6odj0g$nvf$1@cleavage.canuck.com> We get the following error when trying to connect to Sybase running on NT host using EOModeler on Rhapsody for PC compatibles, DR2: The context allocation routine failed when it tried to load localization files!! One or more following problems may caused the failure Cannot access the sybase home directory, please check environment variable SYBASE or ~sybase Your sybase home directory is . Check the environment variable SYBASE if it is not the one you want! Cannot access file *‚^C Jul 10 16:19:22 Workspace: EOModeler was killed by signal 10 We have also set the environment variable with the following command: setenv SYBASE /usr/sybase The same environment variable is passed to the app from the Launcher window settings. Our Rhapsody root directory has folder /usr/sybase, with the interfaces file sitting there and pointing to the Sybase server. We have no problem connecting to the NT-based Sybase server from NEXTSTEP 3.3 environment, using DB-Library and the same interfaces file, and from another NT host using CT-Library. thanks, heidi
From: schefflr@news.msus.edu (Bill Scheffler) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.system,comp.sys.next.programmer,comp.sys.mac.programmer.codewarrior Subject: Re: Apple heading into obscurity? (the Twighlight Zone) Followup-To: comp.sys.mac.advocacy,comp.sys.mac.system,comp.sys.next.programmer,comp.sys.mac.programmer.codewarrior Date: 13 Jul 1998 19:04:49 GMT Organization: Minesota Colleges and Universities Message-ID: <6odloh$b62$1@Urvile.MSUS.EDU> References: <macghod-1207981812330001@sdn-ar-001casbarp113.dialsprint.net> <6obs8b$hnk$1@morgoth.sfu.ca> <35A980D1.2371@globaldialog.com> Troll. Get a clue. Jim Mueller (webnik@globaldialog.com) wrote: : The Mac is already obscure. The real fear is about the Mac disappearing : completely. This is a scarey thought for me. : Can you imagine? This newsgroup would only have Edwin and Joe Ragosta : posting in it? They would be the last Mac users left on the whole : planet. : Their Posts might go something like this: : ********************************************************************** : JR Post: "Let's play the Rizzo impersonator game. You pretend your : Farkny and I'll defend our beloved platform with a scathing attack on : your mindless trolls. Then I get to call you strawman and stuff like : that". : Edwin Post: "Golly, Joe, why am I always pretending to be Rizzo. When do : I get to call YOU strawman? I wanna defend the Mac! Can I?... huh, Joe? : ....can I"? : *********************************************************************** : ....and so it would go on forever....in the Mac Advocacy Zone....... : doo doo doo dee...doo doo doo dee...doo doo doo dee...doo doo doo dee... -- -------------------------------------------------------------------------- Bill Scheffler | "Good...Bad...I'm the guy with 163 East Snarr Hall -- (218) 236-3285 | the gun." - Ash Moorhead State University | (Army of Darkness) schefflr@mhd1.moorhead.msus.edu |------------------------------- http://dragon.moorhead.msus.edu/~schefflr |-------- ICQ# 11985553 -------- --------------------------------------------------------------------------
From: Jim Mueller <webnik@globaldialog.com> Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.system,comp.sys.next.programmer,comp.sys.mac.programmer.codewarrior Subject: Re: Apple heading into obscurity? (the Twighlight Zone) Date: Mon, 13 Jul 1998 15:42:34 -0500 Organization: http://www.globaldialog.com/~webnik/ Message-ID: <35AA713A.2A48@globaldialog.com> References: <macghod-1207981812330001@sdn-ar-001casbarp113.dialsprint.net> <6obs8b$hnk$1@morgoth.sfu.ca> <35A980D1.2371@globaldialog.com> <6odloh$b62$1@Urvile.MSUS.EDU> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Bill Scheffler wrote: > > Troll. Get a clue. > Advocate. Get a sense of humor! =:-o
From: msb@plexare.com (Michael S. Barthelemy) Newsgroups: comp.sys.next.marketplace,comp.sys.next.programmer Subject: Looking for OPENSTEP/WebObjects/Rhapsody Jobs Date: 13 Jul 1998 22:44:45 GMT Organization: A-Link Network Services, Inc. Message-ID: <6oe2kt$r0p@ns2.alink.net> My first post was munged, this is a repost. If anyone knows of any projects looking for OPENSTEP/WebObjects/Rhapsody developers in the Bay Area (Preferably the southern half.) I would appreciate knowing about them. Thanks, Mike Barthelemy msb@idiom.com
From: schefflr@news.msus.edu (Bill Scheffler) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.system,comp.sys.next.programmer,comp.sys.mac.programmer.codewarrior Subject: Re: Apple heading into obscurity? (the Twighlight Zone) Followup-To: comp.sys.mac.advocacy,comp.sys.mac.system,comp.sys.next.programmer,comp.sys.mac.programmer.codewarrior Date: 13 Jul 1998 21:56:02 GMT Organization: Minesota Colleges and Universities Message-ID: <6odvpi$gna$1@Urvile.MSUS.EDU> References: <macghod-1207981812330001@sdn-ar-001casbarp113.dialsprint.net> <6obs8b$hnk$1@morgoth.sfu.ca> <35A980D1.2371@globaldialog.com> <6odloh$b62$1@Urvile.MSUS.EDU> <35AA713A.2A48@globaldialog.com> I am by no means an advocate, I just thought you were serious. Sorry about that. Jim Mueller (webnik@globaldialog.com) wrote: : Bill Scheffler wrote: : > : > Troll. Get a clue. : > : Advocate. Get a sense of humor! : =:-o -- -------------------------------------------------------------------------- Bill Scheffler | "Good...Bad...I'm the guy with 163 East Snarr Hall -- (218) 236-3285 | the gun." - Ash Moorhead State University | (Army of Darkness) schefflr@mhd1.moorhead.msus.edu |------------------------------- http://dragon.moorhead.msus.edu/~schefflr |-------- ICQ# 11985553 -------- --------------------------------------------------------------------------
From: msb@plexare.com (Michael S. Barthelemy) Newsgroups: comp.sys.next.marketplace,comp.sys.next.programmer Subject: Looking for OPENSTEP/WebObjects/Rhapsody Jobs Date: 13 Jul 1998 22:39:09 GMT Organization: A-Link Network Services, Inc. Message-ID: <6oe2ad$qs2@ns2.alink.net> If anyone knows of
From: "Mike Anderson" <mikea@knowmed.com> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: Mon, 13 Jul 1998 22:27:57 -0700 Organization: The World's Usenet -- http://www.Supernews.com Message-ID: <6oeqfr$keh$1@supernews.com> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> <rex-0807980230180001@cc497300-a.wlgrv1.pa.home.com> <slrn6q81lj.j00.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu> <rex-0907980206430001@cc497300-a.wlgrv1.pa.home.com> <6o52fj$tvg$1@nnrp1.dejanews.com> <6o5r90$1a6$1@supernews.com> <6oacti$5o4$1@nnrp1.dejanews.com> tj_sam@my-dejanews.com wrote in message <6oacti$5o4$1@nnrp1.dejanews.com>... >In article <6o5r90$1a6$1@supernews.com>, > "Mike Anderson" <mikea@knowmed.com> wrote: > >>>would be represented as 2D (X,Y) wire frame objects. If subroutine >>>or function calls are represented as Z plane bi-directional vector >>>(recursive function are outlawed) > >>mikea@knowmed.com wrote in message >>The most notably missing thing is abstraction. GUIs are impaired >>because oftheir inability to handle abstraction, but it's not much of >>a problem because people don't expect much of them. Programming >>without abstraction would be a very tedious excersise (but it would in >>some sense be simpler). >>Abstraction is a powerful tool, and it should be refined, not ignored >>or outlawed. >I would agree, in programming abstractions create a simplification. >However it must be understood (assuming OOP) that abstractions only mask >an internal complex, with or without abstraction (classes) the overall >complexity of any representation remains the same. By "abstraction" I don't mean the the encapsulation and hiding of details (which itself is useful), I mean more like "generalization" or "not concrete". To go back to the GUI analogy, GUIs tend to be oriented towards concreteness. A GUI allows you to manipulate a folder that contains some documents. There are GUI elements for each of the documents that can be dragged and dropped and "opened". But the entity representing "all documents created after 98/01/01" is not available in most GUIs. The collection of documents created after 98/0101 can be grouped and manipulated, but this is not the same as the abstract notion of "all documents created after 98/01/01". The lack of abstraction makes GUIs easier to understand, but it excludes a wide range of functionality that must be cobbled up in ways that are ultimately more complicated than would be a solution that involved abstraction. In programming, if I'm writting a procedure or something that is supplied wi th parameters, those parameters can be defined to varying degrees of abstraction. It can be defined as an array of sequential numbers, or some kind of collection of numbers, or some kind of collection, or something that understands the message 'size'. The more abstract that definition, the more possible uses for the procedure. There are other forms of abstraction in programming, such as objects that represent a sequence of actions. Invoking the actions of such an object is an abstract invocation of action. These abstractions are a kind of leverage in dealing with complexity and I thought it was the kind of thing that would be disallowed by your VR programming system. ...Mike
From: stgeorge@amug.org (Steve George) Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.system,comp.sys.next.programmer,comp.sys.mac.programmer.codewarrior Subject: Re: Apple heading into obscurity? (the Twighlight Zone) Date: 14 Jul 1998 04:47:25 GMT Organization: amug.org Message-ID: <stgeorge-1307982145100001@d27-ts06.amug.org> References: <macghod-1207981812330001@sdn-ar-001casbarp113.dialsprint.net> <6obs8b$hnk$1@morgoth.sfu.ca> <35A980D1.2371@globaldialog.com> <gmgraves-1307981027280001@sf-pm5-24-88.dialup.slip.net> In article <gmgraves-1307981027280001@sf-pm5-24-88.dialup.slip.net>, gmgraves@slip.net (George Graves) wrote: > Better expand that list to include me. Long after Joe and Edwin switch > over to "the dark side", I'll still be using and advocating Macintosh.. I'll > stop when death overtakes me, not until. > > George Graves Jes' pry it on outa' ma col' daid fingees.. ;] Hate most lists, but guess I'd spring for that one too. Think it will ever do us any good..? The Mac yes, the lists, verrrry seldom. Modern man fights the machine age, I'd say (somethin' to "do" you can con some buckies with, that's about it most cases). Jes' "sellin' that soap." :] Steve George
From: "Scott Hildebrand" <sch@roadnet.ups.com> Newsgroups: comp.sys.next.programmer Subject: Re: Objective C woes Date: Tue, 14 Jul 1998 07:34:44 -0400 Organization: UPS IS-Maryland Message-ID: <6offgr$d81@innsrv.roadnet.ups.com> References: <leffert.900164607@cs.uchicago.edu> Jonathan B. Leffert wrote in message ... ............................... >-- foo.m -- ......................... >- (void) setBar: (id) b >{ > [bar autorelease]; > bar = [b copy]; >} > >@end > >Everything, I think, looks ok. Then, I created x.m to test the foo class: > >-- x.m -- >#import "foo.h" > >int main() >{ > id s = @"hello"; > id f = [foo init]; > [f setBar: s]; > printf("%s\n",[[f getBar] cString]); > return 0; >} > One subtle point to remember is that in x.m, this line > [f setBar: s]; causes setBar to mark "bar" for auto-release. At this point bar doesn't point to anything. I believe the compiler set's it to NIL, which is why there is no negative side effect of this. Sending a message to a NIL p-obj does nothing. I would still flag it as an issue in a code review. Scott Hildebrand sch@roadnet.ups.com
From: Bill Bumgarner <bbum@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: Objective C woes Date: Tue, 14 Jul 1998 10:49:27 -0400 Organization: Spacelab.net Internet Access Message-ID: <Pine.NXT.3.96.980714103331.3427D-100000@pathos> References: <leffert.900164607@cs.uchicago.edu> <6offgr$d81@innsrv.roadnet.ups.com> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII To: Scott Hildebrand <sch@roadnet.ups.com> In-Reply-To: <6offgr$d81@innsrv.roadnet.ups.com> See below for specifics. First some general comments. (1) Always name classes starting with a capital letter. I.e. Foo, not foo. Likewise, always name variables starting with a lowercase letter and capitalizing subsequent words.... i.e. aString, anArray, theCow, etc... (2) Whenever possible, use specific pointer types. Instead of: id f; Use: Foo *f; On Tue, 14 Jul 1998, Scott Hildebrand wrote: > > Jonathan B. Leffert wrote in message ... > ............................... > >-- foo.m -- > ......................... > >- (void) setBar: (id) b > >{ > > [bar autorelease]; > > bar = [b copy]; Depending on what you really want to do, it is likely that you don't want to copy b. There are reasons to do so.... basic rules: If (b) is immutable, then just do "bar = [b retain]". If b is mutable, you need to decide if you want your foo instance to see changes as they happen... if so, just retain. If not, copy. Regardless, you should always use -release instead of -autorelease, though it requires a couple of extra lines of code to prevent badness. -release is more effecient in that it will release the object immediately instead of adding yet-another-operation-during-autoreleasepool-death. Something like: - (void) setBar: (SomeClass *) b { if (bar != b) { [bar release]; bar = [b retain]; } } The above code works regardless of whether or not bar or b are nil. > >} > > > >@end > > > >Everything, I think, looks ok. Then, I created x.m to test the foo class: > > > >-- x.m -- > >#import "foo.h" > > > >int main() > >{ > > id s = @"hello"; > > id f = [foo init]; This is the crus of the problem. [foo init] sends the method -init to the CLASS foo. You never allocated an instance of foo! Something like: Foo *f = [[Foo alloc] init]; Is probably much closer to what you wanted.... > > [f setBar: s]; Because f is not an instance of Foo, the above line fails with an unrecognized method exception.... > > printf("%s\n",[[f getBar] cString]); The above could be written as: NSLog(@"%@", [f getBar]); The runtime will take care of converting whatever getBar returns by calling -description. > > return 0; > >} > > > > > > > One subtle point to remember is that in x.m, this line > > > [f setBar: s]; > > causes setBar to mark "bar" for auto-release. At this point bar doesn't > point to anything. I believe the compiler set's it to NIL, which is why > there is no negative side effect of this. Sending a message to a NIL p-obj > does nothing. I would still flag it as an issue in a code review. Instance variables always start with a value of nil/0/NO/0.0 whatever. Basically, teh objc runtime uses something like calloc() to allocate the object structure. It is a defined standard that a message to nil is a no-op; in this particular case [setting an ivar-- either when the ivar is nil and/or the value it is to be set to is nil], this behaviour is probably a good thing. In general, you are quite correct in that it is a good idea to flag reliance on messaging to nil as a potential issue in a code review. Not because it is incorrect but because it relies on a behaviour of ObjC that is not immediately apparent by reading the code-- it should be documented OR simply do something like: - (void) someMethod: (NSButtonCell *) aButtonCell if ( (aButtonCell != nil) && ([aButtonCell isEnabled] == YES) ) { ... } Yes, it is slightly more verbose. Yes, the "== YES" is unnecessary. But the above if test leaves no question in the programmers mind as to what is going on. Likewise, it gives the compiler the maximum number of hints as to what is going on and, as such, the compiler will generate warnings indicative of potential problems. I started using the above style of extremely anally retentive programming about four years ago [after reading Steve McConnell's awesome +Code Complete+] and it has been tremendously rewarding: - code handoffs are less painful... - greatly reduces compile time errors and runtime bugs - i spend a hell of a lot less time fixing problems - the amount of code i write on a per-day basis has skyrockted... Of course, your mileage may vary. [oh, and finally, go read +Objective-C and Object Oriented Programming+ -- it is a truly excellent book that is included with OpenStep or Rhapsody. Wonderful read regarding dynamic OO design and development. It uses ObjC as the language of demonstration, but is by no means limited to being useful in the objc realm] b.bum
From: jwdb@fygir.nl (Jan-Willem de Bruijn) Newsgroups: comp.sys.next.programmer Subject: [Q] NSComboBoxCell in NSTableColumn Date: 14 Jul 1998 16:08:54 GMT Organization: NLnet Message-ID: <6ofvqm$bk1$1@news.Leiden.NL.net> A quick question: I am trying to use NSComboBoxCells in an NSTableView. First I have told the NSComboBoxCell to use a dataSource (and set it to my controller). Then I call -setDataCell: on an NSTableColumn. -numberOfItemsInComboBoxCell: is being called, but not -comboBoxCell:objectValueForItemAtIndex: The cell that appears has an empty (5 item) popup list. Am I forgetting something? Is this a bug? Or is it just not possible? Thanks in advance for any comments, Jan-Willem -- Jan-Willem de Bruijn - F Y G I R logistic information systems http://www.fygir.com/
From: Mark Bessey <"mark_bessey"@apple.com (no spam, please)> Newsgroups: comp.sys.next.programmer Subject: Re: Objective C woes Date: Tue, 14 Jul 1998 10:47:26 -0700 Organization: Apple Computer, Inc. Message-ID: <6og608$k0k$1@news.apple.com> References: <leffert.900164607@cs.uchicago.edu> <6offgr$d81@innsrv.roadnet.ups.com> <Pine.NXT.3.96.980714103331.3427D-100000@pathos> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit > It is a defined standard that a message to nil is a no-op; in this > particular case [setting an ivar-- either when the ivar is nil and/or the > value it is to be set to is nil], this behaviour is probably a good thing. One little clarification: a message to nil returns nil. This actually matters in many cases, for instance: - (void) setBar: (SomeClass *) b { if (bar != b) { [bar release]; bar = [b retain]; } } This code will actually set bar to nil if b is nil (unless they're both nil...). Another case where it really matters is when the message you send to nil returns a type other than (id). In this case, the results can be really unexpected, depending on the compiler/runtime that you use. For instance, the following will return a totally bogus answer on OPENSTEP (whatever happens to be in the appropriate register). What's worse is that the answer isn't even consistent - it can vary from run to run... #include <Foundation/Foundation.h> @interface Foo: NSObject { float myVal; } - (float) getFloatValue; - (void) setFloatValue: (float) newValue; @end @implementation Foo - (float) getFloatValue { return myVal; } - (void) setFloatValue: (float) newValue { myVal = newValue; } @end int main(void) { Foo *f = [[Foo alloc] init]; [f setFloatValue: 5]; NSLog(@"[f getFloatValue] == %f", [f getFloatValue]); NSLog(@"[nil getFloatValue] == %f", [nil getFloatValue]); }
From: ajmas@bigfoot.com Newsgroups: comp.sys.next.programmer Subject: [Q] Palette libraries - sources? Date: Tue, 14 Jul 1998 18:16:16 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6og79g$cl2$1@nnrp1.dejanews.com> Can anyone tell me where I can get hold of supplemental palettes for Project Builder? Thanks AJ -- http://www.bigfoot.com/~ajmas/ -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: ajmas@bigfoot.com Newsgroups: comp.sys.next.programmer Subject: How about a JAR style file for OpenStep Apps? Date: Tue, 14 Jul 1998 18:14:48 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6og76p$c3i$1@nnrp1.dejanews.com> Would it be at all unreasonable to look at implenting the OpenStep .app folder as a uncompressed ZIP file that contains the same stuff, much in the same way as is done for the Java JAR file? AJ -- http://www.bigfoot.com/~ajmas/ -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: "Daniel T. Fahey" <DanFahey@DanSources.com> Newsgroups: comp.sys.next.marketplace,comp.sys.next.advocacy,comp.sys.next.programmer Subject: Job: NYC EOF and WOF Date: Tue, 14 Jul 1998 15:56:58 -0700 Organization: DanSources Technical Services, Inc. Message-ID: <35ABE239.47DD@DanSources.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NNTP-Posting-Date: 14 Jul 1998 19:58:27 GMT I have a short term contract assignment in NYC, NT.. About 2 months..Need EoF and WoF experience. Could go longer. If you know anyone interested, please have them call Dan Fahey 301-217-0425 www.DanSources.com
From: pete@ohm.york.ac.uk (-bat.) Newsgroups: comp.sys.next.programmer Subject: Re: Objective C woes Date: 14 Jul 1998 19:16:55 GMT Organization: The University of York, UK Sender: pcf1@york.ac.uk Message-ID: <6ogar7$a1d$1@pump1.york.ac.uk> > int main() > { > id s = @"hello"; > id f = [foo init]; > [f setBar: s]; > printf("%s\n",[[f getBar] cString]); > return 0; > } Change that from "[foo init]" into "[foo new]" and it works fine. > Jul 11 08:31:10 x[3130] *** _NSAutoreleaseNoPool(): Object 0xc64c of class > NSInlineCString autoreleased with no pool in place - just leaking You can get rid of them by adding a line at the start: NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init]; -bat.
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: 4.2 DriverKit Newsgroups: comp.sys.next.programmer Message-ID: <35abf21c.0@news.depaul.edu> Date: 15 Jul 98 00:04:44 GMT Recently Apple made available a package of OpenStep 4.2 DriverKit materials. I downloaded it, but it doesn't seem to be anything I didn't already have. Further, I wasn't able to build one of the examples. Has anyone been able to use this? -- "... and subpoenas for all." - Ken Starr
From: arti@lava.DOTnet (Art Isbell - remove "DOT") Newsgroups: comp.sys.next.programmer Subject: Minimized windows can't be maximized on OSE 4.2 apps on Win 95 Date: 15 Jul 1998 04:47:37 GMT Organization: LavaNet, Inc. Distribution: world Message-ID: <6ohc99$ril@mochi.lava.net> We have encountered a serious problem with our OSE 4.2 app under Windows 95 (but not NT). After a window has been minimized either by clicking the minimize window button or by selecting a menu item whose target is the window and whose action is minimize:, the window cannot be maximized again. Clicking the window's task bar button causes only a narrow section of the window's top frame to appear - no window contents appear at all! Does anyone have a workaround for this problem short of preventing window minimization? -- Art Isbell NeXT/MIME Mail: arti@lavaDOTnet IDX Systems Corporation Voice/Fax: +1 808 394 0511 (for whom I don't speak) Voice Mail: +1 808 394 0495 Healthcare Info Technology US Mail: Honolulu, HI 96825-2638
From: blenko-tom@cs.yale.edu (Tom Blenko) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 15 Jul 1998 07:20:56 -0400 Organization: Yale University Computer Science Dept., New Haven, CT 06520-2158 Distribution: world Message-ID: <6oi3aoINNq4j@RA.DEPT.CS.YALE.EDU> |> The claim that functional programs can be written thinking about what |> is to be computed but not how -- a claim which was popular in the FP |> community in the 1980's -- has been strongly discredited. | |That is not what I was claiming. Sure you need to have an idea of the |subcomputations that will make the full computation. But you don't need |a precise ordering on them in the way that is implied by the idea of |"control flow". I think I was overreading you. I would regard this as a weakening of control flow, not a case of throwing it away. |... I wasn't talking about hardware. Concurrent logic languages don't |require special hardware. They could be mapped onto networks of |distributed processors and execute in real parallel, or they could be |implemnted on standard von Neumann machines in which case while there's |no real parallelism the pseudo-parallelism of concurrency can be |helpful in making programs clearer - what after all is object-oriented |programming but pseudo-parallelism - lots of objects running in |parallel. It's only messed up by the unnecessary control flow imposed |on it by von Neumann oriented languages. So are there solutions to the problems of (non-local) resource management and, e.g., race conditions which arise from committed-choice non-determinism? Because control flow in the strong sense may limit your ability to abstract (in this fashion, anyway), but the weakening you mention seems to introduce problems that are very difficult to solve, highly non-modular, etc. (hence my earlier remark about models that have been discredited). (I assume you are leaving the problem of granularity to the programmer -- though I wouldn't be willing to agree that that is something a programmer can necessarily handle very well). I always thought it was telling that the inventor of Concurrent Prolog didn't himself realize the race conditions existed, initially, for example. What, by the way, happens to assert- and retract-like operations in such a language? Because I might think that, again, these would make it very difficult for a programmer to program in the style you suggest. I don't understand the comment about about OO programs -- I would think that Actors might be about as radical as you could get in parallel, object-oriented languages, but it's much, much harder to program in (I've done it) than a conventional OO language. So the latter don't seem pseudo-parallel to me. Tom
From: scott@doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: Objective C woes Date: 14 Jul 98 12:14:33 Organization: Is a sign of weakness Message-ID: <SCOTT.98Jul14121433@slave.doubleu.com> References: <leffert.900164607@cs.uchicago.edu> <6offgr$d81@innsrv.roadnet.ups.com> <Pine.NXT.3.96.980714103331.3427D-100000@pathos> In-reply-to: Bill Bumgarner's message of Tue, 14 Jul 1998 10:49:27 -0400 In article <Pine.NXT.3.96.980714103331.3427D-100000@pathos>, Bill Bumgarner <bbum@codefab.com> writes: In general, you are quite correct in that it is a good idea to flag reliance on messaging to nil as a potential issue in a code review. Not because it is incorrect but because it relies on a behaviour of ObjC that is not immediately apparent by reading the code-- it should be documented OR simply do something like: - (void) someMethod: (NSButtonCell *) aButtonCell if ( (aButtonCell != nil) && ([aButtonCell isEnabled] == YES) ) { ... } Yes, it is slightly more verbose. Yes, the "== YES" is unnecessary. But the above if test leaves no question in the programmers mind as to what is going on. I'd like to emphasize this point (I periodically come out of the woodwork and do this :-). A more common template for me is: - init { self=[super init]; >>> if( self!=nil) { <<< } return self; } In the marked line, you could say "if( self)". But, a decent C programmer will always fill in the blanks when reading the code and say "if self is non-zero." They do this because they've been bitten so many times in the past... in any case, when you read it as I phrased it, you're saying "if self isn't nil", which is generally easier to scan, because you don't have to invoke all those damned special case processing routines that you've so painfully aquired. I call this the "Say what you mean" clause. If you mean "if self isn't nil", then _say_ "if( self!=nil)". Don't elide the obvious points. Likewise, if you rely on instance variables being initialized to nil, then _initialize_ them to nil. Treat all uninitialized instance variables as undefined. My single biggest gripe about the foundation kit is: -(void)release; You used to use -(id)free, which led to the excellent coding style of: obj=[obj free]; A compact way of saying "free the object and forget about it." Now, you have to say "Free the object. Forget about the object.": [obj release]; obj=nil; Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: blenko-tom@cs.yale.edu (Tom Blenko) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 15 Jul 1998 07:50:37 -0400 Organization: Yale University Computer Science Dept., New Haven, CT 06520-2158 Distribution: world Message-ID: <6oi52dINNqnd@RA.DEPT.CS.YALE.EDU> ||> I haved worked with non-strict functional languages for the past 13 ||> years. It never crossed my mind that anyone would cite them in a ||> response. | |Why not? We use Erlang (a strict functional language) to program |several very large distributed control systems (> 300 KLOCS). If you're saying that Erlang is strict and represents an abandonment of control flow -- which was the topic for the exchange above -- I'd like to hear more (because I don't get it!). Regarding the other claims, I'm afraid I don't know much about Erlang. I do know a bit about ML. 1. Is Erlang different from ML in some way that, e.g., affects control flow? 2. I'm interested in the software engineering claims. How does one argue that a large program requires fewer lines of code in one language than in another? Even if you provide nearly-identical programs in each language, I'd probably want to know which one was implemented first, who implemented each, how they were qualified. 3. Your claim that a line of Erlang code requires the same time to develop as a line of C++ code is interesting (the claim has been made around the functional programming community for years, with no credible supporting evidence that I've been able to discover -- maybe it's a chance for you to become famous :-). Again, I'd like to know what methodology was used. 4. One missing topic was... performance. What can you tell us about performance? 5. I'm not a great expert on C++, but I have been programming in C for a long time. As I understand it, C is more or less a subset of C++. So here's the question: if I can write your program in Erlang, can't I write it in a functional subset of C? Admittedly, it's not higher-order, but a lot of higher-orderness can be faked, and generally it's something to used sparingly because it's expensive. What do I give up? Typing? Well, there is some weak typing available for C. I have to do my own storage management, but that doesn't necessarily add much code to a large program (it might add more to a smaller program). You could convince me that a functional program might be shorter than a C program providing similar functionality. But I think the C programmer probably wouldn't write in the functional style -- which is, after all, freely available -- because he thinks the non-functional style will run faster, and he can always do a tighter and faster job of storage management (even though it might well take longer to develop). I would expect the development times to converge for a longer program, and the C programmer would be left with a faster, lighter program. 6. There are, of course, other issues to be addressed wrt software engineering, e.g., maintainability. Tom
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: [Q] Creating a HICON from a NSImage Date: Wed, 15 Jul 1998 11:12:26 +0200 Organization: Square B.V. Message-ID: <35AC727A.6D00FB4@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Is there a way to create a WIN32 icon handle (HICON) from a NSImage or a NSBitmapRep, or a way to load a .ico file from the resources and make a HICON from that file? I know it is very platform specific, but that is allright. Maurice. -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto:mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+ Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" and "to trick into believing or accepting as genuine something false and often preposterous"
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: The reason why Date: Wed, 15 Jul 1998 11:25:51 +0200 Organization: Square B.V. Message-ID: <35AC759F.A797422F@Square.nl> References: <35AC727A.6D00FB4@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To clarify my request let me tell you why. I want to change the Icon of a NSWindow. OpenStep uses the default application window, but I want another image to be used. NSWindow-setMiniWindowImage does not work (?). Now I use the following code: - CODE ---------------------- HWND hwnd; HICON hicon; hicon = LoadIcon((HINSTANCE)[ [NSApplication sharedApplication] applicationHandle], MAKEINTRESOURCE(3)); hwnd = (HWND)[window windowHandle]; DefWindowProc(hwnd,WM_SETICON, ICON_SMALL, hicon); - END CODE ------------------ The problem is that I use a hard coded '3' in the MAKEINTRESOURCE call. If I change the number of icons recognized by the application I have to change the call. This is very inconvienient. Regards, Maurice. -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto:mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+ Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" and "to trick into believing or accepting as genuine something false and often preposterous"
From: mike@erix.ericsson.se (Mike Williams) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 15 Jul 1998 13:12:44 GMT Organization: Ericsson Telecom, Stockholm, Sweden Sender: mike@morgan (Mike Williams) Distribution: world Message-ID: <6oi9sc$beq$1@news.du.etx.ericsson.se> References: <6oi52dINNqnd@RA.DEPT.CS.YALE.EDU> NNTP-Posting-User: mike In article <6oi52dINNqnd@RA.DEPT.CS.YALE.EDU>, blenko-tom@cs.yale.edu (Tom Blenko) writes: |> 1. Is Erlang different from ML in some way that, e.g., affects control |> flow? No. Except in that Erlang provides concurrent processes and asynchronous communication between the processes. We use the term "process" rather than "thread" since they have logically disjunct address spaces. |> 2. I'm interested in the software engineering claims. How does one |> argue that a large program requires fewer lines of code in one language |> than in another? Even if you provide nearly-identical programs in each |> language, I'd probably want to know which one was implemented first, |> who implemented each, how they were qualified. Of course no such claims can ever be made exactly. But we have examined many nearly identical programs written in Erlang and C/C++ and found that the Erlang programs were (in terms of non commented lines of code) shorter than the C/C++ by factors of between 2.5 and 10. On average we find that a factor of four seems realistic. You are quite right in saying that there are other factors, such as which programs were written first, how the programmers were qualified etc etc. As I said in my previous posting our metrics are based on HUGE systems. |> 3. Your claim that a line of Erlang code requires the same time to |> develop as a line of C++ code is interesting (the claim has been made |> around the functional programming community for years, with no credible |> supporting evidence that I've been able to discover -- maybe it's a |> chance for you to become famous :-). Again, I'd like to know what |> methodology was used. Ericsson keeps detailed metrics of all large software projects. This claim is substantiated by measurements on the design of nearly one million lines of Erlang code and several million lines of C/C++ code. |> 4. One missing topic was... performance. What can you tell us about |> performance? Just as with metrics on design productivity, there are a number of factors here. Execution of Erlang code is of course slower that execution of well written C code. But Erlang certainly executes fast to be commercially viable for use in building real time switching system products. You possibly have already used Ericsson switches developed using Erlang without knowing it. |> 5. I'm not a great expert on C++, but I have been programming in C for |> a long time. As I understand it, C is more or less a subset of C++. |> So here's the question: if I can write your program in Erlang, can't I |> write it in a functional subset of C? Admittedly, it's not |> higher-order, but a lot of higher-orderness can be faked, and generally |> it's something to used sparingly because it's expensive. You could do that, but the software would be a lot longer and contain much more extraneous detail. Also C has no simple primitives for modeling concurrency, message passing in distributed systems, distributed error handling etc etc. |> What do I give up? Typing? Well, there is some weak typing available |> for C. I have to do my own storage management, but that doesn't |> necessarily add much code to a large program (it might add more to a |> smaller program). It is more a question of what you *gain*. Pattern matching to start with.... |> You could convince me that a functional program might be shorter than a |> C program providing similar functionality. But I think the C |> programmer probably wouldn't write in the functional style -- which is, |> after all, freely available -- because he thinks the non-functional |> style will run faster, and he can always do a tighter and faster job of |> storage management (even though it might well take longer to develop). Most of the people who have developed our systems are designers with previous experience from using imperative languages. Most of them are able to convert to writing in a functional style without difficulty. |> I would expect the development times to converge for a longer program, |> and the C programmer would be left with a faster, lighter program. Our systems often comprise in the order of of 500 000 lines of Erlang code. The equivalent of 2 000 000 lines of C code. I am not taking about "a programmer", I am talking about teams of 50 or more programmers, systems designers, testers, integrators, project managers. This isn't an academic experiment, this is big time industrial product development! |> 6. There are, of course, other issues to be addressed wrt software |> engineering, e.g., maintainability. We have metrics about that too. About the same fault frequency (or better) per line of Erlang code as for C/C++. As we do have much fewer lines of code than C/C++ this also means far fewer faults. --- To summarize: 1) We *do* use Erlang to build products. Ericsson is one of the largest manufactures of telecom equipment in the world with a revenue of about 24 Billion USD in 1998. We use Erlang, C, C++, Java and lots of other languages as well. 2) Projects using Erlang have progressed very well. Several of them have resulted in successful commercial products. Many new products based on systems designed using Erlang are in the pipes. Both the "gut feelings" of experienced designers and detailed metrics taken from many projects confirm that the use of Erlang radically reduces the total software design work required. A rough average is by a factor of four. 3) The quality and maintainability of the software we produce has also been confirmed in several product developments. 4) Our experience from using Erlang comes from observation of the design of several million lines of code written by large teams of programmers. /Mike
From: "Daniel C. Wang" <danwang+news@cs.princeton.edu> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 12 Jul 1998 11:17:33 -0400 Organization: Princeton University Sender: danwang@vista.CS.Princeton.EDU Message-ID: <r8tsok7yu7m.fsf@vista.CS.Princeton.EDU> References: <6oahqrINNrkm@RA.DEPT.CS.YALE.EDU> <james-ya023180001207981011070001@news> james@unspam.edu (James McCartney) writes: > >Perhaps I'm overreading your comments, but it's far from clear to me > >that control flow can be "thrown away", and I also don't understand > >what is "old-fashioned" about the von Neumann model -- do you know of > >some alternative model that haven't been throughly discredited at this > >point? > > Well you write from a CS dept, so I'd assume you know the following anyway. > Surely you don't think lazy FP has been "throughly discredited" at this > point. > > Lazy functional languages order of evaluation is demand driven. The > order of operations is not written explicitly by the programmer but > is determined by the runtime structure of the problem. This allows one > to program by specifying the problem and not how to do it step by step, > which is quite liberating. It has been reported that programs written > in lazy FPs look, to those unfamiliar with how they work, like only a > description of the problem and not a solution. Laziness is oversold. What you win in control flow abstraction you lose in space/memory abstraction. Debugging a lazy program with a space problem is a pain. Lazyiness is useful but it just solves a particular kind of problem and gets in the way for other kinds.
From: Malcolm Steel <malcolm.steel@gecm.com> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: Wed, 15 Jul 1998 14:48:16 +0100 Organization: Own thoughts Message-ID: <35ACB320.4128@gecm.com> References: <6oi52dINNqnd@RA.DEPT.CS.YALE.EDU> <6oi9sc$beq$1@news.du.etx.ericsson.se> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Mike Williams wrote: <snip> > But we have > examined many nearly identical programs written in Erlang and C/C++ > and found that the Erlang programs were (in terms of non commented > lines of code) shorter than the C/C++ by factors of between 2.5 and > 10. On average we find that a factor of four seems realistic. I guess the comparison was based on source code and not object code ? <snip> > Execution of Erlang code is of course slower that > execution of well written C code. But Erlang certainly executes fast > to be commercially viable for use in building real time switching > system products. This I find interesting. A piece of software four times shorter can take longer to run ? Why ? <snip> > We have metrics about that too. About the same fault frequency (or > better) per line of Erlang code as for C/C++. As we do have much fewer > lines of code than C/C++ this also means far fewer faults. You may have fewer faults, but what kind of faults are they and how long do they take to spot and correct. Is the whole lifecycle 4 times shorter ? What kind of testing is performed and what coverage to your achieve (statement, branch or path)? These questions are out of interest not critisism. Regards Malcolm Steel
From: mmh@dcs.qmw.ac.uk (Matthew M. Huntbach) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Followup-To: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Date: 15 Jul 1998 13:01:13 GMT Organization: Queen Mary & Westfield College, London, UK Distribution: world Message-ID: <6oi96p$68d$1@beta.qmw.ac.uk> References: <6oi3aoINNq4j@RA.DEPT.CS.YALE.EDU> Tom Blenko (blenko-tom@cs.yale.edu) wrote: > |... I wasn't talking about hardware. Concurrent logic languages don't > |require special hardware. They could be mapped onto networks of > |distributed processors and execute in real parallel, or they could be > |implemnted on standard von Neumann machines > So are there solutions to the problems of (non-local) resource > management and, e.g., race conditions which arise from committed-choice > non-determinism? Not quite sure what race conditions you are referring to. Logic languages by definition don't have rewriteable variables, so the standard race condition examples won't apply. Were you thinking of something more specialised? > What, by the way, happens to assert- and retract-like operations in > such a language? Because I might think that, again, these would make it > very difficult for a programmer to program in the style you suggest. They don't have them. Good logic programmers shouldn't need to use them. Generally when people use them in Prolog it means they haven't got out of thinking in an imperative state-changing style. > I don't understand the comment about about OO programs -- I would think > that Actors might be about as radical as you could get in parallel, > object-oriented languages, but it's much, much harder to program in > (I've done it) than a conventional OO language. So the latter don't > seem pseudo-parallel to me. The concurrent logic languages can be seen as a form of Actors. Indeed, I feel it's better to program with them in an Actors style than think of them as "parallel Prologs". What is it you find hard about Actors? My current research work is based on building a language which is like Actors but compiles to a concurrent logic language, thus providing it with a syntax and easy implementation. I've also added a few extra constructs which may overcome the problems you feel Actors have, basically extra control-like mechanisms, implemented underneath by the "recursive argument as states" mechanism used to map Actor-like programming into concurrent logic, related also to the "continuation passing" style that is fashionable in the functional programming community. Matthew Huntbach
From: Bill Bumgarner <bbum@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: Objective C woes Date: Wed, 15 Jul 1998 10:49:35 -0400 Organization: Spacelab.net Internet Access Message-ID: <Pine.NXT.3.96.980715104632.9122E-100000@pathos> References: <leffert.900164607@cs.uchicago.edu> <6offgr$d81@innsrv.roadnet.ups.com><Pine.NXT.3.96.980714103331.3427D-100000@pathos> <SCOTT.98Jul14121433@slave.doubleu.com> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII To: Scott Hess <scott@doubleu.com> In-Reply-To: <SCOTT.98Jul14121433@slave.doubleu.com> Or you can say: CFRSetVariableToobject(obj, nil); Where CFRSetVariableToObject() is #defined as: #define CFRSetVariableToObject(_v_,_o_) do {\ id _object_ = (_o_);\ if (_v_ == _object_)\ break;\ if (_v_ != nil)\ [_v_ release];\ _v_ = _object_;\ if (_v_ != nil)\ [_v_ retain];\ } while(0) b.bum > My single biggest gripe about the foundation kit is: > > -(void)release; > > You used to use -(id)free, which led to the excellent coding style of: > > obj=[obj free]; > > A compact way of saying "free the object and forget about it." Now, > you have to say "Free the object. Forget about the object.": > > [obj release]; > obj=nil; > > Later, > -- > scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ > <Favorite unused computer book title: The Compleat Demystified Idiots > Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed> > >
From: mike@erix.ericsson.se (Mike Williams) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 15 Jul 1998 15:08:12 GMT Organization: Ericsson Telecom, Stockholm, Sweden Sender: mike@morgan (Mike Williams) Message-ID: <6oigks$beq$2@news.du.etx.ericsson.se> References: <6oi52dINNqnd@RA.DEPT.CS.YALE.EDU> <6oi9sc$beq$1@news.du.etx.ericsson.se> <35ACB320.4128@gecm.com> NNTP-Posting-User: mike In article <35ACB320.4128@gecm.com>, Malcolm Steel <malcolm.steel@gecm.com> writes: |> > 10. On average we find that a factor of four seems realistic. |> |> I guess the comparison was based on source code and not object code ? Yes, on non commented, non blank lines of source code. |> This I find interesting. A piece of software four times shorter can |> take longer to run ? Why ? Because Erlang is compiled to code for a virtual machine (like Java). This has the same advantages as for Java, portability, garbage collection (on a per process ("thread") for Erlang), safe memory addressing etc. In addition Erlang's virtual machine handles asynchronous message passing between Erlang processes, even if the processes are on different machines and distributed error handling. |> You may have fewer faults, but what kind of faults are they and how long |> do they take to spot and correct. Is the whole lifecycle 4 times |> shorter ? Yes, the whole life cycle is four times faster. We have only got metrics for the whole life cycle of projects as our time reports don't differentiate between specification, coding, test, integration, project management, documentation etc etc. I only have a subjective answer to the time taken to spot and correct bugs, again due to lack of detailed statistics in our time reporting. My subjective answer is that bugs probably take the same time to spot but are faster to locate and correct since there is less code. |> What kind of testing is performed and what coverage to your achieve |> (statement, branch or path)? Different projects using Erlang have performed different degrees of coverage during testing. In general, testing can be hard in the sort of systems we design due to the fact that control systems control things (hardware :-). |> These questions are out of interest not criticism. Ask away! /Mike
From: csaldanh@mae.carleton.ca (Chris Saldanha) Newsgroups: comp.sys.next.programmer Subject: NSTextField updating Date: 15 Jul 1998 15:15:31 GMT Organization: Carleton University, Ottawa, Canada Message-ID: <6oih2j$l5g$1@bertrand.ccs.carleton.ca> OK, so how does one trigger an update of an NSTextField inside a controller's action method? (OpenStep/Mach 4.2) If I call [myTextField setStringValue: @"FOO!"] and then call [myTextField display] it doesn't work. The updates queue up until the action method finishes, and then it does the updates. I found that if you call [myTextField display] 3 or 4 times in a row, it will do it. Is display the wrong method to call? --Chris Chris Saldanha | Computers are useless. Carleton University (Comp. Sci) | They can only give you answers. csaldanh@mae.carleton.ca (NeXT/MIME) | http://www.mae.carleton.ca/~csaldanh | -Pablo Picasso
Newsgroups: comp.sys.next.programmer Subject: 4.2 DriverKit? Message-ID: <k8kevJEMXQIC@cc.usu.edu> From: root@127.0.0.1 Date: 15 Jul 98 09:21:29 MDT Distribution: world MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I've looked in NextAnswers with no success. Could someone please post a pointer to the 4.2 DriverKit? TIA.
From: boehring@biomed.ruhr-uni-bochum.de (Daniel Boehringer) Newsgroups: comp.sys.next.programmer Subject: Re: NSTextField updating Date: 15 Jul 1998 17:30:55 GMT Organization: Ruhr-Universitaet Bochum, Rechenzentrum Message-ID: <6oip0f$t4u$1@sun579.rz.ruhr-uni-bochum.de> References: <6oih2j$l5g$1@bertrand.ccs.carleton.ca> csaldanh@mae.carleton.ca (Chris Saldanha) wrote: >OK, so how does one trigger an update of an NSTextField inside a >controller's action method? (OpenStep/Mach 4.2) > >If I call [myTextField setStringValue: @"FOO!"] and then call >[myTextField display] it doesn't work. The updates queue up until the >action method finishes, and then it does the updates. > you could try flushWindowIfNeeded (guessing). daniel
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: Re: 4.2 DriverKit? Newsgroups: comp.sys.next.programmer References: <k8kevJEMXQIC@cc.usu.edu> Distribution: world Message-ID: <35acf1fe.0@news.depaul.edu> Date: 15 Jul 98 18:16:30 GMT root@127.0.0.1 wrote: > I've looked in NextAnswers with no success. > Could someone please post a pointer to the 4.2 DriverKit? ftp://dev.apple.com/devworld/Rhapsody/UsefulStuff/4.2DriverKit.tar.gz -- "... and subpoenas for all." - Ken Starr
From: csaldanh@mae.carleton.ca (Chris Saldanha) Newsgroups: comp.sys.next.programmer Subject: Re: NSTextField updating Date: 15 Jul 1998 19:22:55 GMT Organization: Carleton University, Ottawa, Canada Message-ID: <6oivif$i7p$1@bertrand.ccs.carleton.ca> References: <6oih2j$l5g$1@bertrand.ccs.carleton.ca> <6oip0f$t4u$1@sun579.rz.ruhr-uni-bochum.de> Daniel Boehringer (boehring@biomed.ruhr-uni-bochum.de) wrote: : csaldanh@mae.carleton.ca (Chris Saldanha) wrote: : >If I call [myTextField setStringValue: @"FOO!"] and then call : >[myTextField display] it doesn't work. The updates queue up until the : >action method finishes, and then it does the updates. : > : you could try flushWindowIfNeeded (guessing). Well, I just tried that and it doesn't do anything at all, despite the documentation. The docs explain that display will happen automatically when the eventloop runs. I tried adding [mainWindow display] instead after each change in the textfield. This works more frequently, but still dosn't work properly (ie it sometimes updates and sometimes not) --Chris Chris Saldanha | Computers are useless. Carleton University (Comp. Sci) | They can only give you answers. csaldanh@mae.carleton.ca (NeXT/MIME) | http://www.mae.carleton.ca/~csaldanh | -Pablo Picasso
From: umfrage@gppg.com Newsgroups: comp.sys.next.programmer Subject: Web-consultants, -developers, -programmers wanted Date: Wednesday, 15 Jul 1998 22:59:04 -0600 Organization: Nacamar Data Communications Message-ID: <15079822.5904@gppg.com> We are constantly seeking contacts to consultants, programmers, designers which can help us making our site the number 1 within the adult industry. Are you interested ? Then please visit our site, http://www.internet-sexshop.de/gpeg/umfrage/memhome_e.htm tell us what you think and send suggestions and quotations !
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <15079822.5904@gppg.com> Control: cancel <15079822.5904@gppg.com> Date: 15 Jul 1998 21:34:29 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.15079822.5904@gppg.com> Sender: umfrage@gppg.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Steve Dekorte <dekorte@slip.net> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 15 Jul 1998 22:31:34 GMT Organization: Slip.Net (http://www.slip.net) Message-ID: <6ojak6$gnh$1@owl.slip.net> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> In comp.sys.next.programmer Christopher Browne <cbbrowne@news.hex.net> wrote: > The answers are likely the same as those to the question: > "Why do engineers still write out equations rather than ``simply'' > drawing the relationships?" Actually, most engineers and scientists make extensive use of drawings/diagrams to understand problems. This usually happens even before symbols are assigned to the elements of the problem. > The written language of mathematics has been, and still is, symbolic > equations as opposed to graphical pictures. Actually, math uses alot of visual elements. The over-under divider bar, parenthetic groupings, etc. > I would suggest that in the same way, the notion of some sort of written > language is critical to computing, and that while visual representations > can provide some insight, they are not so fundamentally important as the > symbolic representations. I think it's true that symbols are inescapable. But visual metaphors are powerfull way of representing relationships between symbols. Unfortunately most visual programming efforts are wasted on trying to replace symbols instead of making sybmols easier to understand and manage. [argument that math = computing] Traditional mathematics is about avoiding computation. Computing is about _doing_ computation. These would seem to have similiar goals, but it turns out that very few interesting problem can actually be solved in the mathematical sense. (The n-body problem, higher order polynomial roots, etc) So the techniques and perspectives of traditional mathematics are of limited value. > VR still leaves you with the need to determine an appropriate set of > symbols and ways to manipulate those symbols. I agree completely. --- Steve Dekorte
From: Steve Dekorte <dekorte@slip.net> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 15 Jul 1998 22:45:52 GMT Organization: Slip.Net (http://www.slip.net) Message-ID: <6ojbf0$gnh$2@owl.slip.net> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> <rex-0807980230180001@cc497300-a.wlgrv1.pa.home.com> <slrn6q81lj.j00.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu> <james-ya023180000807982343080001@news> In comp.sys.next.programmer James McCartney <james@unspam.edu> wrote: > I try to convince people sometimes that text based programming > actually is a graphical programming environment, and a very dense and > sophisticated one. But they usually don't buy it. > I've yet to see a graphical language that could express something in less > space than a text equivalent. You're implying less space = more understandable. There is something to that, but only something. Compilers are happy with no return characters (and less space) while people often aren't. >Graphical languages I've seen are one of > graphs with nodes and edges, (easy to translate to text) > containment diagrams, (even easier -> trees) > tiles. ( just a variation of nodes and edges) > But then I'm not really up on what else might be out there. > (I'm not counting graphical languages whose end product is the graphics > themselves.) There is little else. Unfortunately, most visual programing research has been more interested in making things "graphical" than using graphics to make programing easier. > All of them can be translated pretty easily to a lisp like syntax. Or machine language. <grin> > I think graphical browsers that can display relationships in code > are probably more beneficial and might be a good VR direction. > Some advantages to graphical environments is that they can > provide default arguments, and popup help for their arguments and usage. > However this is not impossible for text environments, just not usually > implemented as well. Good ideas. But text environments don't lend themselve's to this sort of use. --- Steve Dekorte
Newsgroups: comp.sys.next.programmer From: jmvallat-nospam-@nexty.fdn.fr (Jean-Michel Vallat) Subject: Re: NSTextField updating Message-ID: <1998Jul15.235245.14497@nexty.fdn.org> Sender: jmvallat@nexty.fdn.org Organization: Individual References: <6oivif$i7p$1@bertrand.ccs.carleton.ca> Date: Wed, 15 Jul 1998 23:52:45 GMT What your looking for is: 1. [myTextField setStringValue:@"FOO!"]; 2. [myTextField display]; 3. PSWait(); 1. The AppKit automatically sends a 'setNeedsDisplay:YES' to 'myTextField' view. This is recorded by the window, and (at the end of the current event-loop) all the views marked as 'needsDisplay' will be ... redisplayed (in an optimized way). 2. Force the view 'myTextField' to be displayed right now. The PostScript code (drawRect:) is sent to the WindowServer. 3. Ask to the WindowServer to perform all its pending drawing immediately. These comments can be wrong, but it's how I understand the mechanism ;-) Cheers, JeanMi. -- In article <6oivif$i7p$1@bertrand.ccs.carleton.ca> csaldanh@mae.carleton.ca (Chris Saldanha) writes: > Daniel Boehringer (boehring@biomed.ruhr-uni-bochum.de) wrote: > : csaldanh@mae.carleton.ca (Chris Saldanha) wrote: > : >If I call [myTextField setStringValue: @"FOO!"] and then call > : >[myTextField display] it doesn't work. The updates queue up until the > : >action method finishes, and then it does the updates. > : > > > : you could try flushWindowIfNeeded (guessing). > > Well, I just tried that and it doesn't do anything at all, despite the > documentation. The docs explain that display will happen automatically > when the eventloop runs. I tried adding [mainWindow display] instead > after each change in the textfield. This works more frequently, but > still dosn't work properly (ie it sometimes updates and sometimes not)
From: Darrell Scott Mockus <darrellm@earthlink.net> Newsgroups: comp.sys.next.programmer Subject: Oracle, Java, EOF, and primary key sequences.... Date: Wed, 15 Jul 1998 17:18:10 -0700 Organization: Mockus Consulting Message-ID: <35AD46C2.6D6FBE19@earthlink.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Has anybody out there worked with generating primary keys for an Oracle database in Java? I have done this in Obj C for SQL server but and having some trouble with the Java classes and Oracle. It seems you have to go about it in a different way. I was trying something like this but I am a little off, can somebody share some Java code they got working that gets this done? Thanks a ton! ----------------------------------------- public void saveChanges() throws Exception { ImmutableHashtable newPK; FetchSpecification fspec; ObjectStoreCoordinator rootStore; Adaptor anAdaptor; DatabaseContext dbContext; AdaptorContext adContext; AdaptorChannel adChannel; Entity entity; Model myModel; try { fspec = new FetchSpecification("PARTNERPROFILE_SEQ", null, null); rootStore = (ObjectStoreCoordinator)this.session().defaultEditingContext().rootObjectStore(); dbContext = (DatabaseContext)rootStore.objectStoreForFetchSpecification(fspec); anAdaptor = dbContext.database().adaptor(); adContext = dbContext.adaptorContext(); adChannel = adContext.createAdaptorChannel(); adChannel.evaluateExpression(SQLExpression.expressionForString("create SOMETABLE_SEQ")); myModel = ModelGroup.defaultGroup().modelNamed("Model"); newPK= adChannel.primaryKeyForNewRowWithEntity(myModel.entityNamed("SOMETABLE_SEQ")); bla bla bla , get values from the ImmutableHashtable, set PK, insert object bla bla... this.session().defaultEditingContext().saveChanges(); } catch (Exception exception) { // An error occurred during the save. bla bla handle it. } Thanks! -- =-=-=-=-=-=-=-=-=-=-=-===-=-=-=-=-=-=-=-=-=-=-=-=-==-=-=-=-=-=-= "To err is human, to really foul up you need a machine." =-=-=-=-=-=-=-=-=-=-=-=--=-=-=-=--=-=-=-=-=-=-=-=-==-=-=-=-=-=-= darrellm@earthlink.net
From: cbbrowne@news.brownes.org (Christopher B. Browne) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 16 Jul 1998 03:40:49 GMT Organization: Hex.Net Superhighway, DFW Metroplex 817-329-3182 Message-ID: <slrn6qqsp4.8ub.cbbrowne@knuth.brownes.org> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> <rex-0807980230180001@cc497300-a.wlgrv1.pa.home.com> <slrn6q81lj.j00.NOSPAMmbkennelNOSPAM@lyapunov.ucsd.edu> <james-ya023180000807982343080001@news> <6ojbf0$gnh$2@owl.slip.net> On 15 Jul 1998 22:45:52 GMT, Steve Dekorte <dekorte@slip.net> posted: >In comp.sys.next.programmer James McCartney <james@unspam.edu> wrote: > >> I try to convince people sometimes that text based programming >> actually is a graphical programming environment, and a very dense and >> sophisticated one. But they usually don't buy it. > >> I've yet to see a graphical language that could express something in less >> space than a text equivalent. > >You're implying less space = more understandable. >There is something to that, but only something. Compilers are happy with >no return characters (and less space) while people often aren't. I'm not seeking the most cryptic representations; scrunching code up together may be economical of space, but those "obfuscated Perl" contests are certainly not economical of the reader's understanding. There may be some graphical system out there that does provide better expressiveness than text; I haven't seen it yet. For the most part, I see visual systems where instead of having a word to describe a language element, they have some sort of pictogram. And seldom do these pictograms represent the sorts of pictures that are worth the proverbial thousand words. >>Graphical languages I've seen are one of >> graphs with nodes and edges, (easy to translate to text) >> containment diagrams, (even easier -> trees) >> tiles. ( just a variation of nodes and edges) >> But then I'm not really up on what else might be out there. >> (I'm not counting graphical languages whose end product is the graphics >> themselves.) > >There is little else. Unfortunately, most visual programing research >has been more interested in making things "graphical" than using graphics >to make programing easier. Perhaps there's something out there that does better; I'm quite skeptical, and there has been little out there in terms of evidence of "really useful" visual systems. >> All of them can be translated pretty easily to a lisp like syntax. > >Or machine language. <grin> The point is that LISP has as part of its design the intent that you can take the code and manipulate it as data. A "pretty-printer" tends to be just a few pages of code. The code may even be able to rewrite itself. This lends flexibility in doing things with the code. That's generally harder with machine language. >> I think graphical browsers that can display relationships in code >> are probably more beneficial and might be a good VR direction. >> Some advantages to graphical environments is that they can >> provide default arguments, and popup help for their arguments and usage. >> However this is not impossible for text environments, just not usually >> implemented as well. > >Good ideas. But text environments don't lend themselve's to this sort of use. That's fair. -- Those who do not understand Unix are condemned to reinvent it, poorly. -- Henry Spencer <http://www.hex.net/~cbbrowne/lsf.html> cbbrowne@hex.net - "What have you contributed to Linux today?..."
From: macghod@concentric.net (Steve Sullivan) Newsgroups: comp.sys.next.programmer,comp.sys.mac.programmer.codewarrior Subject: Newby needs help with macsbug Date: Wed, 15 Jul 1998 22:38:21 -0700 Organization: Great till Apple got rid of the newton Message-ID: <macghod-1507982238210001@sdn-ar-001casbarp257.dialsprint.net> I am having daily crashes with my internet apps (claris emailer, newswatcher mt, and netscape). It is a bad crash, that requires a restart. RATHER than going over what inits I have etc, hopefully someone can help me use this to become more proficient with macsbug. I know some of the basics, such as stdlog, I can tell what app it was in, if it was in 68k mode and other small things. From the bellow stdlog, I would guess the problem is with the toolbox call IconDispatch ? Or maybe since it is in 68k mode that something went wrong with a modeswitch? Does register d0 = 000000000 show something wrong? I also noticed at the very bottom it says: Displaying memory from 0 00000000 FFC1 0000 FFC1 0000 0007 5C16 0007 5C18 o¡ooo¡oooo\ooo\o 00000010 0007 5C1A 0007 5C1C FFC0 3048 FFC0 304A oo\ooo\oo¿0Ho¿0J Memory 0 is not a good memory spot is it? Isnt that a sign that you tried to write to a null pointer? What are some commands to further find out the toolbox calls that were called before the crash? I do know a bit of mac toolbox. MacsBug 6.5.4a4, Copyright Apple Computer, Inc. 1981-98 Bus Error at FFC6024A _IconDispatch+015FA while reading long word from BCFFA0A8 in Supervisor data space 15-Jul-1998 5:20:03 PM (since boot = 22 hours, 59 minutes) Current application is "MT-NewsWatcher 2.4.4" Machine = 510 (PowerMacG3), System $0800, sysu = $01008000 ROM version $077D, $40F2, $0001 (ROMBase $FFC00000) VM is off NIL^ = $FFC10000 Stack space used = +1906 Address FFC6024A is in the ROM at _IconDispatch+015FA 68020 Registers D0 = 00000000 A0 = 000034D8 USP = 00000000 D1 = 00000010 A1 = FFC5F830 MSP = 00000000 D2 = 00000010 A2 = 0006EF50 ISP = 045B6758 D3 = 00000001 A3 = BCFFA0A8 VBR = 00000000 D4 = 00000000 A4 = 045B67AA CACR = 00000001 SFC = 0 D5 = 00000002 A5 = 045B6F98 CAAR = 00000000 DFC = 0 D6 = 00000003 A6 = 045B6778 PC = FFC6024A D7 = 00000010 A7 = 045B6758 SR = Smxnzvc Int = 0 Disassembling from FFC60236 _IconDispatch +015E6 FFC60236 MOVE.W D6,D7 | 3E06 +015E8 FFC60238 ADDQ.W #$4,A7 | 584F +015EA FFC6023A BRA.S _IconDispatch+01604 ; FFC60254 | 6018 +015EC FFC6023C EXT.L D7 | 48C7 +015EE FFC6023E MOVE.L D7,-(A7) | 2F07 +015F0 FFC60240 MOVE.L $0072(A4),-(A7) | 2F2C 0072 +015F4 FFC60244 MOVEA.L $0076(A4),A0 | 206C 0076 +015F8 FFC60248 JSR (A0) | 4E90 +015FA FFC6024A *MOVEA.L D0,A3 | 2640 +015FC FFC6024C MOVE.L A3,D0 | 200B +015FE FFC6024E ADDQ.W #$8,A7 | 504F +01600 FFC60250 BNE.S _IconDispatch+01632 ; FFC60282 | 6630 +01602 FFC60252 ADDQ.W #$3,D7 | 5647 +01604 FFC60254 CMPI.W #$0006,D7 | 0C47 0006 +01608 FFC60258 BLE.S _IconDispatch+015EC ; FFC6023C | 6FE2 +0160A FFC6025A MOVE.W D6,D7 | 3E06 +0160C FFC6025C SUBQ.W #$3,D7 | 5747 +0160E FFC6025E BRA.S _IconDispatch+01628 ; FFC60278 | 6018 +01610 FFC60260 EXT.L D7 | 48C7 +01612 FFC60262 MOVE.L D7,-(A7) | 2F07 Heap zones #1 Mod 10101K 00002800 to 009DFEFF SysZone^ #2 Mod 6K 00010690 to 000121AF ROM read-only zone #3 Mod 48K 0003A9D0 to 000469CF #4 Mod 83019K 009DFF00 to 05AF2E8F Process Manager zone #5 Mod 11404K 03A7EEA0 to 045A209F "MT-NewsWatcher 2.4.4" ApplZone^ TheZone^ TargetZone #6 Mod 4430K 045B6FD0 to 04A0A8BF "Claris Emailer" #7 Mod 7310K 04A1A9E0 to 0513E38F "Netscape Navigator" #8 Mod 19K 055FA590 to 055FF54F #9 Mod 32K 055FF570 to 0560769F #10 Mod 15K 056076C0 to 0560B69F #11 Mod 24K 0560B6C0 to 05611A1F #12 Mod 16K 05611A40 to 05615D5F #13 Mod 47K 05615D80 to 05621B2F #14 Mod 59K 05621B50 to 05630A1F #15 Mod 27K 05630A40 to 0563796F #16 Mod 509K 0563B6F0 to 056BABEF "PPP" #17 Mod 183K 056D6720 to 0570439F "File Sharing Extension" #18 Mod 4K 0570CEE0 to 0570E21F #19 Mod 895K 0592BFE0 to 05A0BE8F "Finder" #20 Mod 727K 05A1C720 to 05AD246F " Instant Organizer" Checking all heaps The System heap at 00002800 is ok The ROM read-only heap at 00010690 is ok The heap at 0003A9D0 is ok The Process Manager heap at 009DFF00 is ok The "MT-NewsWatcher 2.4.4" heap at 03A7EEA0 is ok Totaling the "MT-NewsWatcher 2.4.4" heap at 03A7EEA0 Total Blocks Total of Block Sizes Free 009B #155 006EE5E0 #7267808 Nonrelocatable 0105 #261 0018604C #1597516 Relocatable 0D81 #3457 002AEB90 #2812816 Locked 0039 #57 00099A90 #629392 Purgeable and not locked 0027 #39 0000C2B0 #49840 Heap size 0F21 #3873 00B231BC #11678140 The target heap is the System heap at 00002800 Totaling the System heap at 00002800 Total Blocks Total of Block Sizes Free 0026 #38 0002B950 #178512 Nonrelocatable 0832 #2098 0049E83C #4843580 Relocatable 0827 #2087 00513530 #5322032 Locked 0141 #321 0025F4D0 #2487504 Purgeable and not locked 01B0 #432 001A7400 #1733632 Heap size 107F #4223 009DD6BC #10344124 The target heap is the "MT-NewsWatcher 2.4.4" heap at 03A7EEA0 Displaying File Control Blocks fRef File Vol Type Fl Fork LEof 0002 System g3 ide zsys dW rsrc #5498848 0060 g3 ide oooo dw data #1010688 00BE g3 ide oooo dw data #2021376 011C System Enabler 770 g3 ide gbly dW rsrc #494843 0930 Appearance Extension g3 ide INIT dw rsrc #611315 098E Appearance Extension g3 ide INIT dw rsrc #611315 09EC Open Tpt AppleTalk LiŠ g3 ide libr dw rsrc #541141 0A4A Open Transport Library g3 ide libr dw rsrc #583729 0AA8 Shared Library ManageŠ g3 ide INIT dw rsrc #211694 0B06 Open Transport Library g3 ide libr dw rsrc #583729 0B64 Open Tpt AppleTalk LiŠ g3 ide libr dw rsrc #541141 0BC2 Open Tpt Internet LibŠ g3 ide libr dw rsrc #481470 0C20 OpenTpt Remote Access g3 ide libr dw rsrc #541432 0C7E OpenTpt Modem g3 ide libr dw rsrc #85766 0CDC OpenTpt Remote Access g3 ide libr dw rsrc #541432 0D3A Remote Access Log g3 ide lzlg dW data #192512 0D98 Open Transport Library g3 ide libr dw rsrc #583729 0DF6 Serial (Built-in) g3 ide libr dw rsrc #52797 0E54 OpenTpt Serial ArbitrŠ g3 ide libr dw rsrc #7638 0EB2 OpenTpt Serial ArbitrŠ g3 ide libr dw rsrc #7638 0F10 Users & Groups Data FŠ g3 ide BTFL dW data #192512 0F6E QuickTime g3 ide INIT dw data #298076 0FCC Mac OS Easy Open g3 ide cdev dw rsrc #132666 102A Instant Organizer g3 ide appe dW rsrc #437813 1088 Finder g3 ide FNDR dW rsrc #496519 10E6 File Sharing Library g3 ide shlb dw rsrc #3483 1144 Finder Preferences g3 ide pref dW rsrc #622 11A2 Desktop DB g3 ide BTFL dW data #385024 1200 Desktop DF g3 ide DTFL dW data #1920242 125E Users & Groups Data FŠ g3 ide BTFL dW data #192512 12BC AppleShare PDS g3 ide BTFL dW data #385024 131A File Sharing Extension g3 ide INIT dW rsrc #194076 1378 PPP g3 ide cdev dW rsrc #285347 13D6 Netscape Navigator g3 ide APPL dW rsrc #772461 1434 OpenTpt Remote Access g3 ide libr dw rsrc #541432 1492 Netscape Resources g3 ide NSPL dW rsrc #634729 14F0 Global History g3 ide DBMG dW data #331776 154E CCache log g3 ide DBMC dW data #167936 15AC Certificates7 g3 ide CERT dW data #98304 160A Key Database3 g3 ide TEXT dW data #16384 1668 Claris Emailer g3 ide APPL dW rsrc #1712909 16C6 Claris Emailer Lite PŠ g3 ide PREF dW data #10496 1724 MODEM.DB g3 ide MoDB dw data #22469 1782 Connections g3 ide MMCN dW data #2304 17E0 Schedules g3 ide MMCN dW data #2304 183E Services g3 ide MMSE dW data #4352 189C Signatures g3 ide MMSG dW data #2304 18FA Mail Database g3 ide MMDB dW data #1229056 1958 Mail Index g3 ide MMDB dW data #246016 19B6 Destinations g3 ide MMDE dW data #6400 1A14 Address Book g3 ide MMAB dW data #2304 1A72 Locations g3 ide MMLC dW data #2304 1AD0 Mail Actions g3 ide MMMA dW data #2304 1B2E StdLog g3 ide TEXT dW data #124118 1B8C MT-NewsWatcher 2.4.4 g3 ide APPL dW rsrc #1156950 #181 FCBs, #76 in use (including #21 fonts not listed), #105 free Displaying resource information: > Map $03A7F028, flags $0000, file $1B8C = MT-NewsWatcher 2.4.4 + Map $0000332C, flags $001E, file $0930 = Appearance Extension + Map $000032B0, flags $001E, file $011C = System Enabler 770 + Map $000032A4, flags $801E, file $0003 = oROM resources that override Systemo S Map $00003330, flags $000D, file $0002 = System Map $00554C34, flags $0000, file $0FCC = Mac OS Easy Open [Skipped $0015 maps belonging to font files] Calling chain using A6/R1 links Back chain ISA Caller 00000000 PPC 03BD6568 045B6CF0 PPC 03B15C40 main+00034 045B6CB0 PPC 03B14560 MainEvent+00198 045B6C30 PPC FFD51FB8 WaitNextEvent+00028 045B6BF0 PPC 00637104 045B6AF8 68K 000775AE 'scod BFB1 0002'+019EE 045B6ACC 68K 00633CD6 045B6AA8 68K 004C1E60 045B6A64 68K 00077636 'scod BFB1 0002'+01A76 045B6A52 68K 0007E374 'scod BFB1 0002'+087B4 045B6A04 68K FFC35E38 _Fix2Frac+001E4 045B69EC 68K FFC62886 _IconDispatch+03C36 045B69B0 68K 008D63E2 'MBDF 0000 0930'+00172 045B6946 68K 001D1DCC 'MBDF 0000 0002'+011CC Return addresses on the stack Stack Addr Frame Addr ISA Caller 045B6B34 68K 00624FFE 045B6B18 045B6B14 68K 00624F18 045B6AD0 045B6ACC 68K 000775AE 'scod BFB1 0002'+019EE 045B6AC8 68K 00624C72 045B6AAC 045B6AA8 68K 00633CD6 045B6A94 045B6A90 68K FFC3E8E6 _DateToSeconds+030CE 045B6A68 045B6A64 68K 004C1E60 045B6A56 045B6A52 68K 00077636 'scod BFB1 0002'+01A76 045B6A08 045B6A04 68K 0007E374 'scod BFB1 0002'+087B4 045B69F4 68K 00076C76 'scod BFB1 0002'+010B6 045B69F0 045B69EC 68K FFC35E38 _Fix2Frac+001E4 045B69B4 045B69B0 68K FFC62886 _IconDispatch+03C36 045B6998 68K FFC620FE _IconDispatch+034AE 045B698A 68K FFC62116 _IconDispatch+034C6 045B694A 045B6946 68K 008D63E2 'MBDF 0000 0930'+00172 045B6946 68K 045B69AE 045B693C PPC 001D6120 045B6904 045B6900 68K FFC3C71E _DateToSeconds+00F06 045B68C8 68K 001D0C78 'MBDF 0000 0002'+00078 045B68B0 045B68AC 68K 001D1DCC 'MBDF 0000 0002'+011CC 045B68A2 68K 001D15F6 'MBDF 0000 0002'+009F6 045B689E 045B689A 68K 001D15F6 'MBDF 0000 0002'+009F6 045B6848 PPC 00166808 EmToNatEndMoveParams+00014 045B6838 68K 001A43BE 045B6824 68K FFC620EC _IconDispatch+0349C 045B6804 045B6800 68K 0012FFFE 045B67D4 68K 001D1D42 'MBDF 0000 0002'+01142 045B67B8 045B67B0 PPC FFD5F5F4 StdTxMeas+00030 045B67A8 68K 001CA8FE 045B677C 045B6778 68K FFC604C4 _IconDispatch+01874 Displaying memory from 0 00000000 FFC1 0000 FFC1 0000 0007 5C16 0007 5C18 o¡ooo¡oooo\ooo\o 00000010 0007 5C1A 0007 5C1C FFC0 3048 FFC0 304A oo\ooo\oo¿0Ho¿0J Closing log -- So many pedestrians, so little time.
From: gvandyk@icon.co.za Newsgroups: comp.sys.next.programmer Subject: Re: NSTextField updating Date: 16 Jul 1998 05:43:09 GMT Organization: Technologies Domain Message-ID: <6ok3td$f5f$1@hermes.is.co.za> References: <6oih2j$l5g$1@bertrand.ccs.carleton.ca> <6oip0f$t4u$1@sun579.rz.ruhr-uni-bochum.de> <6oivif$i7p$1@bertrand.ccs.carleton.ca> In-Reply-To: <6oivif$i7p$1@bertrand.ccs.carleton.ca> On 07/15/98, Chris Saldanha wrote: >Well, I just tried that and it doesn't do anything at all, despite the >documentation. The docs explain that display will happen automatically >when the eventloop runs. I tried adding [mainWindow display] instead after >each change in the textfield. This works more frequently, but still dosn't >work properly (ie it sometimes updates and sometimes not) > Try using [[NSDPSContext currentContext] flush] after you display the field. This will flush all pending postscript to the windowserver. -- Regards, Gerrit van Dyk email: gvandyk@icon.co.za (NeXTMail welcome) Technologies Domain (Pty) Ltd The OBJECT is the ADVANTAGE
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 16 Jul 98 00:13:57 Organization: Is a sign of weakness Message-ID: <SCOTT.98Jul16001357@slave.doubleu.com> References: <6nu3gi$97i$1@nnrp1.dejanews.com> <6nube1$c8n$6@blue.hex.net> <6ojak6$gnh$1@owl.slip.net> In-reply-to: Steve Dekorte's message of 15 Jul 1998 22:31:34 GMT In article <6ojak6$gnh$1@owl.slip.net>, Steve Dekorte <dekorte@slip.net> writes: In comp.sys.next.programmer Christopher Browne, <cbbrowne@news.hex.net> wrote: > The answers are likely the same as those to the question: > > "Why do engineers still write out equations rather than > ``simply'' drawing the relationships?" Actually, most engineers and scientists make extensive use of drawings/diagrams to understand problems. This usually happens even before symbols are assigned to the elements of the problem. <...> visual metaphors are powerfull way of representing relationships between symbols. While I find drawings/diagrams _essential_ when explaining solutions to problems to clients, and sometimes in understanding the solutions myself, invariably the diagram is just a high-level description of what's going to happen. I use the diagram to understand how things will work or interact, and then I flesh in the code that implements what the diagram describes. When it comes down to it, engineers _do_ write out the equations. They might use diagrams en route to writing out the equations, but diagrams seldom replace the equations. Sometimes a picture is worth a thousand words - but what's it worth if what's you're describing is actually 500,000 words? Unfortunately most visual programming efforts are wasted on trying to replace symbols instead of making sybmols easier to understand and manage. Which may be related to this point, though I confess that I'm not sure whether what you meant is how I took it. [I took it as saying that replacing a for() loop with a visual for() loop construct doesn't help much.] Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
Newsgroups: comp.sys.next.programmer Subject: GNUstep Foundation Message-ID: <NZ+JJSQ$4GH9@cc.usu.edu> From: root@127.0.0.1 Date: 15 Jul 98 10:12:51 MDT Distribution: world MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Anyone familiar with GNUstep developer software: Does the GNUstep FoundationKit support loadable bundles (ie, is NSBundle implemented)? I'd like to port some OpenStep/ObjC to Linux and really need to know how mature the FoundationKit is. TIA
Newsgroups: comp.sys.next.programmer Subject: GNUstep Foundation Message-ID: <$B4AS769ktuJ@cc.usu.edu> From: root@127.0.0.1 Date: 15 Jul 98 10:02:02 MDT Distribution: world MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Anyone familiar with GNUstep developer software: Does the GNUstep FoundationKit support loadable bundles (ie, is NSBundle implemented)? I'd like to port some OpenStep/ObjC to Linux and really need to know how mature the FoundationKit is. TIA
From: "Ivan RENOUD-GRAPPIN" <albert.deloin@wanadoo.fr> Newsgroups: comp.sys.next.programmer Subject: Trouble with preview on NT Date: Thu, 16 Jul 1998 15:35:16 +0200 Organization: Wanadoo - (Client of French Internet Provider) Message-ID: <6okvl3$fdr$1@platane.wanadoo.fr> I have got a trouble with OpenStep on NT I try to preview the PS result of a printing (using NSPrintInfo setJobDisposition: method, with the NSPrintPreviewJob option), but it doesn't work. I suspect it only works with OpenStep on Mach. Is this right ? So, i have tried to save the PS result in a temporary file (using NSPrintSaveJob instead of NSPriontPreviewJob option), and launched the preview application, by using openTempFile: method on the PS file just created. It works, but the file isn't destroyed after Preview has terminated. And, now, my disk is fool of PS files. Can anybody help me ? ====================== Ivan RENOUD-GRAPPIN - Albert DELOIN SA albert.deloin@wanadoo.fr
From: far_no@spam.ix.netcom.com(Felipe A. Rodriguez) Newsgroups: comp.sys.next.programmer Subject: Re: GNUstep Foundation Date: 16 Jul 1998 13:50:32 GMT Organization: ICGNetcom Distribution: world Message-ID: <6ol0f8$5u8@sjx-ixn10.ix.netcom.com> References: <$B4AS769ktuJ@cc.usu.edu> In article <$B4AS769ktuJ@cc.usu.edu> writes: >Anyone familiar with GNUstep developer software: > >Does the GNUstep FoundationKit support loadable bundles >(ie, is NSBundle implemented)? I'd like to port some >OpenStep/ObjC to Linux and really need to know how mature >the FoundationKit is. > >TIA I just took a quick look and it looks like it has been implemented. I haven't used it though so it's difficult to say. In general the GNUstep Foundation stuff is pretty far along. -- Felipe A. Rodriguez # Francesco Sforza became Duke of Milan from Agoura Hills, CA # being a private citizen because he was # armed; his successors, since they avoided far@ix.netcom.com # the inconveniences of arms, became private (NeXTmail preferred) # citizens after having been dukes. (MIMEmail welcome) # --Nicolo Machiavelli
From: "Stu" <stu@grayechnologies.com> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Program99, was (Re: Programming and the colossal failure to advance) Date: 16 Jul 1998 14:17:24 GMT Organization: Gray Technologies, Inc. Message-ID: <01bdb0c4$801859c0$0fd29cce@kiki.graytechnologies.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <35A0A008.41C6@pc-plus.de.remove.this.bit><01bdac0f$f6d84f20$0fd29cce@kiki.graytechnologies.com> <SCOTT.98Jul10125032@slave.doubleu.com> Scott Hess <scott@doubleu.com> wrote in article <SCOTT.98Jul10125032@slave.doubleu.com>... > In article <01bdac0f$f6d84f20$0fd29cce@kiki.graytechnologies.com>, > "Stu" <stu@grayechnologies.com> writes: > Stephen Riehm <sr@pc-plus.de.remove.this.bit>, > wrote in article <35A0A008.41C6@pc-plus.de.remove.this.bit>... > > The same should be true for programming. The representation on > > the screen doesn't have to be the same as the representation on > > disk. If we had code processors, programs could be treated like > > hypertext documents, interactive - living - systems. > > I concur with this, but remember (assuming your old enough to > remember) the issues that brought us to todays word processors, > principally the incompatibilities. Today using a word processor on > a document is pretty mindless. Just open the document with your > favorite word processor and, in general, viola - you can edit your > document. The reason we can do this is because the word processor > makers have gone to the effort of supporting the majority of > document formats in use today. > > The above paragraph pretty much destroys your credibility, in my eyes. > You can only swap word processors with ease for the simplest > documents. I've been involved for about five years with a project > who's goal is to make it easy to build custom documents quickly, and > one way you get the document is as a word processor format file. > [snip ... diatribe on word processors doc formats] The point being is that years ago you couldn't get one word processor to even READ another word processors format. No committee came up and got the market moving. Yes, there are a lot of common document usages out there now but it has really been the market that has caused this. Everyone wants to be compatible to everyone else (well, maybe with the exception of Micro$oft). I'm not claiming that these document formats are all standard or even well done, I'm sayiong the market has pushed this product the the state it is in - where one word processor can READ another word processors file format. The kind of "file format" compatibility needed for a "code processor" is similiar. And I'm not talking the nut's and bolts of the file format syntax or semantics, I'm talking what the user sees when he calls up the tool and reads in a file. If the "high level" of abstraction was incorporated into the file when is was created by a "code processor" than when it is read by a different companies "code processor" than that "high level" of abstraction better be there or the product is useless. Yeah, yeah ... I can imagine these file formats are real dogs when it comes to content, syntax, symantics, whatever. And maybe trying to comes to some compatibility is an issue in what your trying to do with them ... but you missed the point (and thus lose credibility, in my eyes). The whole idea of the thread was that there should be some "Code processor" that allowed us to program at a higher level of abstraction and you could read in the output of such a system and see the code at that "high level of abstraction". I am saying that the incompatibilities in "format" are what are stopping that from happening, NOT the language issues, but the output format. One "Code processor" is not going read/represent the information (i.e. program structure and hierarchy) the same as the next "Code processor". JUST like it was when word processors hit the market. At least now you CAN read in document formats from several word processor vendors, sure you may have to mess with the formatting etc, but you can do it. -- Stu Potter Principal Engineer Gray Technologies, Inc -- to reply delete the REMOVETHIS from the e-mail address
From: Mark Trombino <mtrombin@ix.netcom.com> Newsgroups: comp.sys.next.programmer Subject: Re: NSTextField updating Date: Thu, 16 Jul 1998 10:24:43 +0000 Organization: ICGNetcom Message-ID: <35ADD4EA.84E116F4@ix.netcom.com> References: <6oih2j$l5g$1@bertrand.ccs.carleton.ca> <6oip0f$t4u$1@sun579.rz.ruhr-uni-bochum.de> <6oivif$i7p$1@bertrand.ccs.carleton.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Chris Saldanha wrote: > > Daniel Boehringer (boehring@biomed.ruhr-uni-bochum.de) wrote: > : csaldanh@mae.carleton.ca (Chris Saldanha) wrote: > : >If I call [myTextField setStringValue: @"FOO!"] and then call > : >[myTextField display] it doesn't work. The updates queue up until the > : >action method finishes, and then it does the updates. > : > > > : you could try flushWindowIfNeeded (guessing). > > Well, I just tried that and it doesn't do anything at all, despite the > documentation. The docs explain that display will happen automatically > when the eventloop runs. I tried adding [mainWindow display] instead after > each change in the textfield. This works more frequently, but still dosn't > work properly (ie it sometimes updates and sometimes not) I've experience similar problems and found no workable solution. I try and update a text field before spawning a new process and I can't for the life of me get the text field to display before the new process starts. flushWindowIfNeeded doesn't work, calling PSWait (or it's OpenStep counterpart) after a display call didn't work, and neither did using [[NSDPSContext currentContext] flush]. If you find something that works for you Chris, would you please post it to the group? -- ============================================== Mark Trombino <mtrombin@ix.netcom.com> ==============================================
From: Bill Bumgarner <bbum@codefab.com> Newsgroups: comp.unix.programmer,comp.sys.next.programmer Subject: Solaris: ioctl() causing app lock-up Date: Thu, 16 Jul 1998 10:08:37 -0400 Organization: Spacelab.net Internet Access Message-ID: <Pine.NXT.3.96.980716095816.1293A-100000@pathos> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII I walked into the middle of a particularly nasty mess of socket code while maintaining a client's application. Symptoms: Occasionally, the production applications (Apple's WebObjects middleware apps served via the web) "spin". That is, their CPU usage goes to 100% and the app stops responding to all requests. Unfortunately, core files are useless in the current environment [on my todo list, trust me] and gdb is not installed on the production servers. "pstack" indicates that the app is typically spinning in ioctl(). Details: - on almost every request, the app opens yet-another-TCP/IP stream or two or three or four or five to grab more information. All streams SHOULD be closed by the time the response is generated. - some of the read-data-from-the-stream code is really ineffecient. For example, some of it looks like: do { read(socket, ... args to read ONE BYTE AT A TIME ...); while (bufPtr != \n); i.e. some of the read code actually reads a single byte at a time from the receive buffer. I'm not totally sure of the internals of how read() works, but this strikes me as being asoundingly ineffecient. - The ioctl() call that caused the most recent spin was: if (ioctl(curQCore->socket, I_NREAD, &avail) < 0) return NULL; Solutions: I don't know-- that's why I'm asking the community at large... :-) To give an idea of the scope of an answer that I could believe, some details: - it may be that a patch is required? Where there any patches released in the last year that fix problems with the socket stuff on Solaris? - could it simply be that all the opening/closing of the data streams is enough to cause the socket substrate to explode under high load situations? - could an additional stress factor be the presence of the read-a-byte-at-atime style of grabbing data out of the receive buffer? thank you, b.bum
From: Dan Wellman <wellman@students.uiuc.edu> Newsgroups: comp.sys.next.programmer Subject: Profiling ObjC in NT Date: 16 Jul 1998 18:59:08 GMT Organization: University of Illinois at Urbana-Champaign Message-ID: <6olihs$dm2$1@vixen.cso.uiuc.edu> I'm using openstep for windows NT and would like to run a profiler such as gdb on the code generated with the gcc compiler that came with Openstep 4.2. Where can I get a 32 bit NT profiler such as gdb? (preferrably that works with OpenStep) Thanks, dan -- Dan Wellman <> wellman@uiuc.edu <> http://www.cen.uiuc.edu/~wellman/ "A million thoughts in one night can't be wrong" - Cause & Effect
Newsgroups: comp.sys.next.programmer,comp.sys.mac.programmer.codewarrior From: markm@bolanos.ana.bna.boeing.com (Mark Miller) Subject: Re: Newby needs help with macsbug In-Reply-To: <macghod-1507982238210001@sdn-ar-001casbarp257.dialsprint.net> Message-ID: <Ew7CJC.FBA@news.boeing.com> Sender: nntp@news.boeing.com (Boeing NNTP News Access) Organization: The Boeing Company References: <macghod-1507982238210001@sdn-ar-001casbarp257.dialsprint.net> Date: Thu, 16 Jul 1998 19:23:36 GMT On 07/15/98, Steve Sullivan wrote: >I am having daily crashes with my internet apps (claris emailer, >newswatcher mt, and netscape). It is a bad crash, that requires a >restart. RATHER than going over what inits I have etc, hopefully someone >can help me use this to become more proficient with macsbug. [ snip ] This most certainly does NOT belong in comp.sys.next.programmer. Please keep your postings in their relevant news groups. >-- >So many pedestrians, so little time. > --
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Program99, was (Re: Programming and the colossal failure to advance) Date: 16 Jul 98 10:51:28 Organization: Is a sign of weakness Message-ID: <SCOTT.98Jul16105128@slave.doubleu.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <35A0A008.41C6@pc-plus.de.remove.this.bit><01bdac0f$f6d84f20$0fd29cce@kiki.graytechnologies.com> <SCOTT.98Jul10125032@slave.doubleu.com> <01bdb0c4$801859c0$0fd29cce@kiki.graytechnologies.com> In-reply-to: "Stu"'s message of 16 Jul 1998 14:17:24 GMT In article <01bdb0c4$801859c0$0fd29cce@kiki.graytechnologies.com>, "Stu" <stu@grayechnologies.com> writes: Scott Hess <scott@doubleu.com> wrote in article <SCOTT.98Jul10125032@slave.doubleu.com>... > In article <01bdac0f$f6d84f20$0fd29cce@kiki.graytechnologies.com>, > "Stu" <stu@grayechnologies.com> writes: > Today using a word processor on a document is pretty mindless. > Just open the document with your favorite word processor and, > in general, viola - you can edit your document. The reason we > can do this is because the word processor makers have gone to > the effort of supporting the majority of document formats in > use today. > > The above paragraph pretty much destroys your credibility, in my > eyes. You can only swap word processors with ease for the > simplest documents. [snip ... diatribe on word processors doc formats] The point being is that years ago you couldn't get one word processor to even READ another word processors format. Hmm. On reading my portion in the light of day, obviously I'm sensitive on this subject. Probably because users keep saying "It should be simple enough to...". No committee came up and got the market moving. Yes, there are a lot of common document usages out there now but it has really been the market that has caused this. Everyone wants to be compatible to everyone else (well, maybe with the exception of Micro$oft). The problem is that Microsoft is a moving target, and everyone wants to be compatible with Microsoft. The fact that Microsoft's formats change with the wind is _not_ necessary. If they wanted to, they could use a format which would be verbose enough to let them describe anything they want, without being so terse that minor changes break everyone else's products. Part of what annoys me about all this is that there is no technical reason why things are the way they are - it's all competitiveness and marketting. I'm not claiming that these document formats are all standard or even well done, I'm sayiong the market has pushed this product the the state it is in - where one word processor can READ another word processors file format. The kind of "file format" compatibility needed for a "code processor" is similiar. And I'm not talking the nut's and bolts of the file format syntax or semantics, I'm talking what the user sees when he calls up the tool and reads in a file. That's exactly the problem we have. The user doesn't always see what we intend them to see in our output files. This is not solely a formatting issue. Sometimes content also gets messed up because of lack of transparency in the tools involved. The content is all _there_, in the file, it's just not making it to the user's level. but you missed the point (and thus lose credibility, in my eyes). The whole idea of the thread was that there should be some "Code processor" that allowed us to program at a higher level of abstraction and you could read in the output of such a system and see the code at that "high level of abstraction". Naturally, I was responding to a single piece of your point. Your argument seems to be that we need format compatibility like word processors currently have. I'm suggesting that the assumption is faulty. It's bad enough when word processors make "reasonable estimates" in an attempt to read in a foreign file. If a code processor does that, you're in for trouble. The problem is that if you open a file with your code processor, and immediately save it back out, you should UNDER NO CIRCUMSTANCES have a result which does not contain all of the components of the original file. The components might be rearranged (whitespace moved, etc), but when you process the file into an executable, it must retain all of the same functionality. This is not _remotely_ true with word processors today. Significant functional elements can be lost if you import from X and immediately export back to X. I am saying that the incompatibilities in "format" are what are stopping that from happening, NOT the language issues, but the output format. The problem is in designing a format which both allows flexibility, and doesn't enforce uselessness. As an example, I'd say SGML allows flexibility, but makes it hard to really _do_ certain things, while RTF doesn't allow much flexibility but is pretty capable within it's own field. XML is a pretty good profile of SGML in that it allows much flexibility, but reduces the range in areas that are hard to programmatically do things with. [I've become somewhat of an XML advocate because it at least offers the possibility that a processor built on an XML model could read a file, ignore the parts it doesn't understand, process the parts it does, and at least _potentially_ write out a new file which still retains the pieces that it didn't understand. Ignore!=discard, in other words.] Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: theisen@akaMail.com (Dirk Theisen) Newsgroups: comp.sys.next.programmer Subject: Use of Delegation Date: Fri, 17 Jul 1998 12:08:24 +0200 Organization: University of Bonn, Germany / www.joint.org Message-ID: <1dcap3y.14ssodz12gkhq8N@rhrz-isdn3-p52.rhrz.uni-bonn.de> Hello! I'm looking for people who have successfully employed delegation as a means to dynamically adapt the behavior of their software at runtime. I.E. change the delegate to switch between different behaviors... Please mail... Regards, Dirk -- Student of Computer Science, University of Bonn, Germany No RISC - No fun
From: blenko-tom@cs.yale.edu (Tom Blenko) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 17 Jul 1998 07:21:22 -0400 Organization: Yale University Computer Science Dept., New Haven, CT 06520-2158 Distribution: world Message-ID: <6onc3iINN8i@RA.DEPT.CS.YALE.EDU> Matthew M. Huntbach wrote: |> Tom Blenko (blenko-tom@cs.yale.edu) wrote: |> So are there solutions to the problems of (non-local) resource |> management and, e.g., race conditions which arise from committed-choice |> non-determinism? | |Not quite sure what race conditions you are referring to. Logic |languages by definition don't have rewriteable variables, so the |standard race condition examples won't apply. Were you thinking of |something more specialised? Suppose you have multiple clauses for the same goal. Each clause has a guard, and the semantics call for the clauses to be executed in parallel, with the first clause to reach the guard winning out and the others being abandoned. I believe this has been called "committed-choice" non-determinism, and that it has been included in several proposed concurrent/parallel logic programming languages (although I'm several years from the terminology and studying the languages). There is a race condition among the different clauses, and I claim that, for example, even by requiring a "non-overlap" property for the guards, the programmer is left with a difficult problem. The race condition is a fairness issue -- can the programmer manageably write programs that give their intended results. In the kind of languages you are describing the programmer must establish properties of a program that allow intended results to be computed even when an adversary schedules tasks on the processor(s). And the programmer must, e.g., assure that the properties are maintained across modifications to the program, etc. (There might be a liveness issue here, as well). The resource management question is a liveness issue. Can the programmer manageably write programs which won't saturate the resource(s)? I point these out because they seem to pop up repeatedly in the proposals I've seen for relaxing or eliminating control flow. Until these questions are addressed I find it difficult to accept the proposals as credible substitutes for conventional control flow because they burden the programmer unmanageable problems. |> I don't understand the comment about about OO programs -- I would thi |> that Actors might be about as radical as you could get in parallel, |> object-oriented languages, but it's much, much harder to program in |> (I've done it) than a conventional OO language. So the latter don't |> seem pseudo-parallel to me. | |The concurrent logic languages can be seen as a form of Actors. Indeed, |I feel it's better to program with them in an Actors style than think |of them as "parallel Prologs". What is it you find hard about Actors? In my experience there are two difficult things about programming in Actors. The first is making sure that all possible control flows produced the desired result (Actors has a very weak fairness policy, as you know). This is a global, distinctly non-modular problem of the sort I cite above as being unmanageable for programmers. The second is the problem of inserting control flow where necessary. For example, I implemented a concurrent logic programming language in Actors. There were, in general, race conditions on binding logic variables, i.e., two unifications in different tasks might be attempting to bind the same variable. Atomicity wasn't difficult to enforce, the problem was that the control flow for the task that failed had to be built before the synch and then unwound/restarted afterward. I'm afraid I'm fuzzy on the details, it was several years ago, but the same problem arises in implementing any language with synchronization/control flow in Actors: you must explicitly add structure to provide the control flow you expect in the implemented language. |My current research work is based on building a language which is like |Actors but compiles to a concurrent logic language, thus providing it |with a syntax and easy implementation. I've also added a few extra |constructs which may overcome the problems you feel Actors have, |basically extra control-like mechanisms, implemented underneath by the |"recursive argument as states" mechanism used to map Actor-like |programming into concurrent logic, related also to the "continuation |passing" style that is fashionable in the functional programming |community. I'd be interested in hearing more, and specifically how your language addresses these two seemingly general problems that I've listed above. (Anothe poster has raised the issue of debugging, which I think is also a serious challenge for these kinds of languages). Tom
From: blenko-tom@cs.yale.edu (Tom Blenko) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 17 Jul 1998 08:32:17 -0400 Organization: Yale University Computer Science Dept., New Haven, CT 06520-2158 Distribution: world Message-ID: <6ong8hINN1gs@RA.DEPT.CS.YALE.EDU> mike@erix.ericsson.se (Mike Williams) writes: |blenko-tom@cs.yale.edu (Tom Blenko) writes: ||> 1. Is Erlang different from ML in some way that, e.g., affects control ||> flow? | |No. Except in that Erlang provides concurrent processes and |asynchronous communication between the processes. We use the term |"process" rather than "thread" since they have logically disjunct |address spaces. Ah, so how do your programmers manage message queues, what happens when the resource is exhausted, and how do you debug such a situation? |To summarize: | |1) We *do* use Erlang to build products. Ericsson is one of the largest |manufactures of telecom equipment in the world with a revenue of about |24 Billion USD in 1998. We use Erlang, C, C++, Java and lots of other |languages as well. | |2) Projects using Erlang have progressed very well. Several of them |have resulted in successful commercial products. Many new products |based on systems designed using Erlang are in the pipes. Both the "gut |feelings" of experienced designers and detailed metrics taken from many |projects confirm that the use of Erlang radically reduces the total |software design work required. A rough average is by a factor of four. | |3) The quality and maintainability of the software we produce has also |been confirmed in several product developments. | |4) Our experience from using Erlang comes from observation of the |design of several million lines of code written by large teams of |programmers. I must say that I find your results very interesting (and singular in the functional programming community)! Do you have references to any publications? Tom
From: mmh@dcs.qmw.ac.uk (Matthew M. Huntbach) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Followup-To: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Date: 17 Jul 1998 13:06:33 GMT Organization: Queen Mary & Westfield College, London, UK Distribution: world Message-ID: <6oni8p$72a$1@beta.qmw.ac.uk> References: <6onc3iINN8i@RA.DEPT.CS.YALE.EDU> Tom Blenko (blenko-tom@cs.yale.edu) wrote: > Matthew M. Huntbach wrote: > |> Tom Blenko (blenko-tom@cs.yale.edu) wrote: > |> So are there solutions to the problems of (non-local) resource > |> management and, e.g., race conditions which arise from committed-choice > |> non-determinism? > | > |Not quite sure what race conditions you are referring to. Logic > |languages by definition don't have rewriteable variables, so the > |standard race condition examples won't apply. Were you thinking of > |something more specialised? > Suppose you have multiple clauses for the same goal. Each clause has a > guard, and the semantics call for the clauses to be executed in > parallel, with the first clause to reach the guard winning out and the > others being abandoned. I believe this has been called > "committed-choice" non-determinism, and that it has been included in > several proposed concurrent/parallel logic programming languages > (although I'm several years from the terminology and studying the > languages). There is a race condition among the different clauses, and > I claim that, for example, even by requiring a "non-overlap" property > for the guards, the programmer is left with a difficult problem. Fine. Early versions of the committed choice logic languages allowed arbitary predicates in the guard portion. However, the problems this caused led these to be dropped, and all development of the languages has concentrated on the "flat" versions where the clause to which a goal commits is chosen only by the variables in the goal becoming bound to match the pattern in the clause head and to meet the guard conditions (which are just simple comparison primitives like > and <). Thus the non-determinism is only the same sort you would find in notations like CSP. I've never found it a problem in using these languages - you can generally tell at a glance when a predicate has been written such that there are goals which could commit to more than one clause. It's no problem at all to write programs where this non-determinism occurs only when you want to use it as a feature e.g. in merging streams that are being produced in parallel so you want a merger that takes from one or the other stream depending on which produces data first. > The race condition is a fairness issue -- can the programmer manageably > write programs that give their intended results. As I said, yes. > In the kind of > languages you are describing the programmer must establish properties > of a program that allow intended results to be computed even when an > adversary schedules tasks on the processor(s). And the programmer > must, e.g., assure that the properties are maintained across > modifications to the program, etc. (There might be a liveness issue > here, as well). In general, the only parallelism is divide-and-conquer parallelism - the solution comes only when all subgoals have finished commputation, so the exact scheduling of the subgoals doesn't matter. In fact this is not exactly the case, and it is possible to write programs which in effect give speculative computation. I discuss this in my paper in the November 1995 issue of IEEE Software. > The resource management question is a liveness issue. Can the > programmer manageably write programs which won't saturate the > resource(s)? Both myself (working on previous experience with something similar in parallel functional languages) and ICOT in Japan came up with the idea of a priority notation to control this, which seems to work ok. > I point these out because they seem to pop up repeatedly in the > proposals I've seen for relaxing or eliminating control flow. Until > these questions are addressed I find it difficult to accept the > proposals as credible substitutes for conventional control flow because > they burden the programmer unmanageable problems. I've fairly extensive experience of use of these languages, and I haven't found it a problem. What I like about the concurrent logic languages is the way they unburden the programmer of many of the problems associated with concurrent programming in more conventional languages. > |My current research work is based on building a language which is like > |Actors but compiles to a concurrent logic language, thus providing it > |with a syntax and easy implementation. I've also added a few extra > |constructs which may overcome the problems you feel Actors have, > |basically extra control-like mechanisms, implemented underneath by the > |"recursive argument as states" mechanism used to map Actor-like > |programming into concurrent logic, related also to the "continuation > |passing" style that is fashionable in the functional programming > |community. > I'd be interested in hearing more, and specifically how your language > addresses these two seemingly general problems that I've listed above. A very early version was published in the 1995 ACM Symposium on Applied Computing. Since then the syntax has been improved and a lot more features have been added. but it gives a general idea of what I'm doing. If you can get hold of that volume, you'll also find another paper in it from me which discusses the idea of speculative computation and priorities as mentioned above in more detail. Matthew Huntbach
From: mike@erix.ericsson.se (Mike Williams) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance - VR Date: 17 Jul 1998 13:28:28 GMT Organization: Ericsson Telecom, Stockholm, Sweden Sender: mike@morgan (Mike Williams) Distribution: world Message-ID: <6onjhs$1oi$1@news.du.etx.ericsson.se> References: <6ong8hINN1gs@RA.DEPT.CS.YALE.EDU> NNTP-Posting-User: mike In article <6ong8hINN1gs@RA.DEPT.CS.YALE.EDU>, blenko-tom@cs.yale.edu (Tom Blenko) writes: |> Ah, so how do your programmers manage message queues, what happens when |> the resource is exhausted, and how do you debug such a situation? This isn't really the place for a long discussion about the details of Erlang. See "Concurrent Programming in Erlang" Armstong, Virding, Wikström and Williams, Prentice Hall 1993, ISBN 0-13-508301-X for more exact details. This book is a bit out of date as it was written before Erlang had higher order functions. To answer you question briefly, message are send to individual processes, not mailboxes. A message can be any Erlang term and need not be specially declared in any way in advance. The syntax is Process ! Message where "Process" can be anywhere in a distributed network of machines. Reception is selective i.e. a processes can wait for a specific message or messages and leave others to be receive later receive Pattern1 -> Action1; Pattern2 -> Action2; Pattern3 -> Action3; .... after TimeOutTime -> TimeOutActions end Basically queues are unbounded, all sizes are dynamic, i.e. queues grow and shrink as required, processes have separate stacks and heaps which also grow and shrink as required. We have a number of special tools which allow developers to monitor memory use, measure the size of queues, processes stacks and heaps etc etc. |> I must say that I find your results very interesting (and singular in |> the functional programming community)! Do you have references to any |> publications? We have published a number of papers over the years. A good one about the complete Erlang development environment and associated middle-ware is: http://www.ericsson.se/Review/er1_97/art_2/art2.html and you will find more in: http://www.ericsson.se:800/cslab/publications.shtml We are a commercial organization where the main focus is on profit and products, not on publications. We are indeed singular in the FP community. I attribute this to the fact that the pressures on us are to get high quality software written in as short time as possible with as few people as possible. Ericsson management (which includes myself) are not that interested in lazy evaluation, monads, purity etc etc. But we most certainly bitch about late deliveries, overrun budgets, bugs in delivered products etc.
From: "Justin Morgan" <jmorgan@objectronics.com> Newsgroups: comp.sys.next.programmer Subject: mmap() on Rhapsody? Date: Fri, 17 Jul 1998 09:57:56 -0700 Organization: Objectronics Consulting Message-ID: <35af824d.0@blushng.jps.net> Hi, There is no man page for mmap() on Rhapsody DR2. Is the mmap() system call supported under Rhapsody DR2? If not, then will it be supported for CR1 ("Mac OS X Server")? Why do we need mmap()? Because mmap() is a significant part of Unix 4.4BSD. There are a plenty of Unix programs that depend on mmap() since that call is available for both the latest Sys V Unixes and also Berkeley 4.4. We won't be able to compile these programs on Rhapsody without mmap(). Without mmap(), Rhapsody is not fully 4.4BSD compatible. (I hope it's there on Rhapsody but merely undocumented.) Thanks for any info, - Justin
From: dbk@mcs.com (Dan "Bud" Keith) Newsgroups: comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Will CodeWarrior (including C++ and MSL) exist for Rhapsody/MacOSXServer? Date: Fri, 17 Jul 1998 14:39:51 -0500 Organization: Emergent Systems Message-ID: <1dcbewy.d0tm6gw4npkkN@dbklap.bb.opentext.com> Ever since Apple announced their Carbon strategy, Metrowerks has been silent about whether they will be supporting Rhapsody (now called MacOS X Server) with their excellent CodeWarrior product. I think that there is a command-line version of the compiler, mwcc, which can be used in conjunction with Project Builder. I am more interested, however, in their MSL libraries for C and C++. I am getting the uncomfortable feeling that Metrowerks is going to simply wait until MacOSX is available before they involve themselves in Apple's NeXT-based technology. This means no C++ and STL will be available for Rhapsody, at least as far as Metrowerks is concerned. Did I miss a critical press release or other announcement that clarified this position? Thanks for any advice, bud ----- Dan Keith (dbk@mcs.com) -----
From: holger@_REMOVE_THIS_.wizards.de (Holger Hoffstaette) Newsgroups: comp.sys.next.programmer Subject: Re: Use of Delegation Date: 17 Jul 1998 21:38:02 GMT Organization: The secret circle of the NSRC Message-ID: <6oog7q$5vc@ragnarok.en.uunet.de> References: <1dcap3y.14ssodz12gkhq8N@rhrz-isdn3-p52.rhrz.uni-bonn.de> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit NNTP-Posting-Date: 17 Jul 1998 21:40:19 GMT Dirk Theisen wrote: > I'm looking for people who have successfully employed delegation as a > means to dynamically adapt the behavior of their software at runtime. > I.E. change the delegate to switch between different behaviors... Basically everybody who has ever programmed NS/OS is familiar with this concept. What questions do you have in particular? How the process works, or if there are certain usage patterns? You'll have to be a bit more precise :-) Holger
Newsgroups: comp.sys.next.programmer Subject: Re: GNUstep Foundation Message-ID: <$kmpzEUpSq19@cc.usu.edu> From: root@127.0.0.1 Date: 17 Jul 98 15:31:30 MDT References: <$B4AS769ktuJ@cc.usu.edu> <6ol0f8$5u8@sjx-ixn10.ix.netcom.com> Distribution: world MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit In <6ol0f8$5u8@sjx-ixn10.ix.netcom.com> Felipe A. Rodriguez wrote: > In article <$B4AS769ktuJ@cc.usu.edu> writes: > >Anyone familiar with GNUstep developer software: > > > >Does the GNUstep FoundationKit support loadable bundles > >(ie, is NSBundle implemented)? I'd like to port some > >OpenStep/ObjC to Linux and really need to know how mature > >the FoundationKit is. > > > >TIA > > > I just took a quick look and it looks like it has been implemented. > I haven't used it though so it's difficult to say. In general the > GNUstep Foundation stuff is pretty far along. Excellent! Thank you, kind sir, for the information.
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: How about a JAR style file for OpenStep Apps? Date: 18 Jul 1998 09:49:31 GMT Organization: http://www.supernews.com, The World's Usenet: Discussions Start Here Message-ID: <6opr3b$phm$1@supernews.com> References: <6og76p$c3i$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: ajmas@bigfoot.com ajmas@bigfoot.com may or may not have said: -> Would it be at all unreasonable to look at implenting the OpenStep .app -> folder as a uncompressed ZIP file that contains the same stuff, much in -> the same way as is done for the Java JAR file? Yes. Been there, done that (pretty much). It used to be the case that NeXTSTEP apps could either keep their resources in the .app directory or in extra segments of the executable (Mach-O) file. Keeping the resources embedded in the executable a' la' Mac resource forks makes it a pain to edit said resources. (Actually, you can still build the monolithic apps, but we don't 'cause it's a bad idea.) -jcr
From: leigh@NOSPAMcs.uwa.edu.au (Leigh Smith) Newsgroups: comp.sys.next.programmer Subject: Re: 4.2 DriverKit Date: 18 Jul 1998 11:19:40 GMT Organization: The University of Western Australia Message-ID: <6oq0cc$pmm$1@enyo.uwa.edu.au> References: <35abf21c.0@news.depaul.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: jhendry@shrike.depaul.edu In <35abf21c.0@news.depaul.edu> Jonathan W Hendry wrote: > Recently Apple made available a package of OpenStep 4.2 > DriverKit materials. I downloaded it, but it doesn't > seem to be anything I didn't already have. > > Further, I wasn't able to build one of the examples. > > Has anyone been able to use this? It's incomplete, the Driver.framework (OpenStep library) is missing. Most of the stuff in the tar file is already in OS4.2 developer distribution. Leigh -- Leigh Computer Music Lab, Computer Science Dept, Smith University of Western Australia +61-8-9380-2279 leigh@cs.uwa.edu.au (NeXTMail/MIME) C++ is to C, as Lung Cancer is to Lung - John C. Randolph
Message-ID: <35B0AA6D.29C0@bellsouth.net> From: Tom Marchand <unix_wiz@bellsouth.net> MIME-Version: 1.0 Newsgroups: comp.sys.mac.advocacy,comp.sys.mac.system,comp.sys.next.programmer,comp.sys.mac.programmer.codewarrior Subject: Re: Apple heading into obscurity? (the Twighlight Zone) References: <macghod-1207981812330001@sdn-ar-001casbarp113.dialsprint.net> <6obs8b$hnk$1@morgoth.sfu.ca> <35A980D1.2371@globaldialog.com> <gmgraves-1307981027280001@sf-pm5-24-88.dialup.slip.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Sat, 18 Jul 1998 13:00:19 GMT NNTP-Posting-Date: Sat, 18 Jul 1998 09:00:19 EST Hey, I just bought a G3! Add me to the list tooooo.
Message-ID: <35B0A597.5977@his.com> Date: Sat, 18 Jul 1998 13:39:43 +0000 From: Royce Priem <priem@his.com> MIME-Version: 1.0 Newsgroups: comp.sys.next.software,comp.sys.next.hardware,comp.sys.next.programmer Subject: RenderManager Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Organization: Heller Information Services Hey - I noticed that NS 3.3 has a demo called RenderManager. Does this allow me to use more than one NeXT to render images? If so, how do I set it up to use it? Thanks in advance... Royce
From: arti@lava.DOTnet (Art Isbell - remove "DOT") Newsgroups: comp.sys.next.programmer Subject: Re: Objective C woes Date: 18 Jul 1998 18:37:00 GMT Organization: LavaNet, Inc. Distribution: world Message-ID: <6oqq0c$ha1@mochi.lava.net> References: <leffert.900164607@cs.uchicago.edu> <6offgr$d81@innsrv.roadnet.ups.com> <Pine.NXT.3.96.980714103331.3427D-100000@pathos> <SCOTT.98Jul14121433@slave.doubleu.com> scott@doubleu.com (Scott Hess) wrote: > My single biggest gripe about the foundation kit is: > > -(void)release; > > You used to use -(id)free, which led to the excellent coding style of: > > obj=[obj free]; > > A compact way of saying "free the object and forget about it." Now, > you have to say "Free the object. Forget about the object.": > > [obj release]; > obj=nil; One can take advantage of a standard C construct to emulate the NEXTSTEP semantics under OPENSTEP: obj = ([obj release], nil); If one integrates this pattern into a set of common Objective-C patterns, then the probability of accessing a freed object is reduced. -- Art Isbell NeXT/MIME Mail: arti@lavaDOTnet IDX Systems Corporation Voice/Fax: +1 808 394 0511 (for whom I don't speak) Voice Mail: +1 808 394 0495 Healthcare Info Technology US Mail: Honolulu, HI 96825-2638
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <26220900216033@digifix.com> Date: 19 Jul 1998 03:48:46 GMT Organization: Digital Fix Development Message-ID: <17553900820823@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1993. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: cms@macisp.net Newsgroups: comp.sys.next.programmer Subject: Wanted: Next GUI for Web/Database Setup ** Date: Sun, 19 Jul 1998 11:08:58 -0400 Organization: Crystal Wind Communications, Inc. Message-ID: <cms-1907981109030001@209.26.71.138> NNTP-Posting-Date: 19 Jul 1998 15:08:56 GMT We are looking for a Next Gui for Web/Database Setup ** 1. Must Know Next System OS.. Setup or Adjustment to the System 2. Must Know of Server Software for Web/Database access 3. Must Know of Sybase or if there is a better choice. tell us. 4. Must be able to complete project quickly. Bottom line: To have a Next Color Turbo be a Database Server for the Web. Interested: Send E-Mail: cms@macisp.net Attn: Richard
From: ajmas@bigfoot.com Newsgroups: comp.sys.next.programmer Subject: [Q] When should we expect a more polished YB for Windows? Date: Mon, 20 Jul 1998 14:34:36 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6ovkht$lki$1@nnrp1.dejanews.com> I have checked out the Yellow Box for Windows installation and found that it had a few rough edges. Has their been any work put improving the YB for Windows or will this not happen until MacOS X server is released? AJ -- http://www.bigfoot.com/~ajmas/ -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <200798003912@hotmail.com> Control: cancel <200798003912@hotmail.com> Date: 20 Jul 1998 15:15:12 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.200798003912@hotmail.com> Sender: <soothsayer5@hotmail.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <200798003916@hotmail.com> Control: cancel <200798003916@hotmail.com> Date: 20 Jul 1998 15:15:40 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.200798003916@hotmail.com> Sender: <soothsayer5@hotmail.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <200798003914@hotmail.com> Control: cancel <200798003914@hotmail.com> Date: 20 Jul 1998 15:15:40 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.200798003914@hotmail.com> Sender: <soothsayer5@hotmail.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <200798003917@hotmail.com> Control: cancel <200798003917@hotmail.com> Date: 20 Jul 1998 15:15:40 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.200798003917@hotmail.com> Sender: <soothsayer5@hotmail.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <200798003923@hotmail.com> Control: cancel <200798003923@hotmail.com> Date: 20 Jul 1998 15:15:40 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.200798003923@hotmail.com> Sender: <soothsayer5@hotmail.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Christian Neuss <neuss.@informatik.th-darmstadt.de.nos-pam> Newsgroups: comp.sys.next.programmer Subject: Re: Wanted: Next GUI for Web/Database Setup ** Date: 20 Jul 1998 15:47:50 GMT Organization: Technische Universitaet Darmstadt Message-ID: <6ovor6$3qk$2@sun27.hrz.tu-darmstadt.de> References: <cms-1907981109030001@209.26.71.138> cms@macisp.net wrote: > >We are looking for a Next Gui for Web/Database Setup ** > >1. Must Know Next System OS.. Setup or Adjustment to the System >2. Must Know of Server Software for Web/Database access >3. Must Know of Sybase or if there is a better choice. tell us. >4. Must be able to complete project quickly. Have tried OPENBASE? Comes with EOF adaptor and good graphical setup utilities. Cheaper, too, and supported. HTH; Chris -- // Christian Neuss "static typing? how quaint.." // http://www.nexttoyou.de/~neuss/ // fax: (+49) 6151 16 5472
From: Sigthor Hrafnsson <siffi@islandia.is> Newsgroups: comp.sys.next.marketplace,comp.sys.next.programmer Subject: Re: Looking for OPENSTEP/WebObjects/Rhapsody Jobs Date: Sun, 19 Jul 1998 16:39:11 +0100 Organization: Islandia ehf. Message-ID: <35B2131F.28D69FA8@islandia.is> References: <6oe2kt$r0p@ns2.alink.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: msb@idiom.com Check with lozinski@openstepnews.com at: http://www.openstepnews.com/jobmarket.html Michael S. Barthelemy wrote: > My first post was munged, this is a repost. > > If anyone knows of any projects looking for OPENSTEP/WebObjects/Rhapsody developers in the Bay Area (Preferably the > southern half.) I would appreciate knowing about them. > > Thanks, > > Mike Barthelemy > msb@idiom.com
Newsgroups: comp.sys.next.programmer Subject: Re: 4.2 DriverKit? Message-ID: <H4QRL0Pl7i+K@cc.usu.edu> From: root@127.0.0.1 Date: 19 Jul 98 22:06:36 MDT References: <k8kevJEMXQIC@cc.usu.edu> <35acf1fe.0@news.depaul.edu> Distribution: world MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit In <35acf1fe.0@news.depaul.edu> Jonathan W Hendry wrote: > root@127.0.0.1 wrote: > > I've looked in NextAnswers with no success. > > Could someone please post a pointer to the 4.2 DriverKit? > > ftp://dev.apple.com/devworld/Rhapsody/UsefulStuff/4.2DriverKit.tar.gz Thank you sir for the pointer.
From: MWRon@metrowerks.com (MW Ron) Newsgroups: comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: Will CodeWarrior (including C++ and MSL) exist for Rhapsody/MacOSXServer? Date: Mon, 20 Jul 1998 13:01:07 -0400 Organization: Metrowerks Corporation Message-ID: <MWRon-ya02408000R2007981301070001@enews.newsguy.com> References: <1dcbewy.d0tm6gw4npkkN@dbklap.bb.opentext.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <1dcbewy.d0tm6gw4npkkN@dbklap.bb.opentext.com>, dbk@mcs.com (Dan "Bud" Keith) wrote: >Ever since Apple announced their Carbon strategy, Metrowerks has been >silent about whether they will be supporting Rhapsody (now called MacOS >X Server) with their excellent CodeWarrior product. I think that there >is a command-line version of the compiler, mwcc, which can be used in >conjunction with Project Builder. This is correct. >I am more interested, however, in their MSL libraries for C and C++. I >am getting the uncomfortable feeling that Metrowerks is going to simply >wait until MacOSX is available before they involve themselves in Apple's >NeXT-based technology. This means no C++ and STL will be available for >Rhapsody, at least as far as Metrowerks is concerned. I'm afraid all I can say is.... We're working with Apple to get the final plan of rollout of our tools intermixed with their tools before we announce anything Ron -- The CodeWarrior FAQ is now open to the public. <http://cw-faq.mit.edu/cw-faq/faq> METROWERKS Ron Liechty "Software at Work" MWRon@metrowerks.com
From: blob@invalid.email.address (BLoB) Newsgroups: comp.sys.next.programmer,comp.sys.mac.programmer.codewarrior Subject: Re: Newby needs help with macsbug Date: Mon, 20 Jul 1998 10:48:32 -0700 Organization: none Message-ID: <blob-2007981048320001@menudo.apple.com> References: <macghod-1507982238210001@sdn-ar-001casbarp257.dialsprint.net> In article <macghod-1507982238210001@sdn-ar-001casbarp257.dialsprint.net>, macghod@concentric.net (Steve Sullivan) wrote: > I am having daily crashes with my internet apps (claris emailer, > newswatcher mt, and netscape). It is a bad crash, that requires a > restart. RATHER than going over what inits I have etc, hopefully someone > can help me use this to become more proficient with macsbug. I know some > of the basics, such as stdlog, I can tell what app it was in, if it was in > 68k mode and other small things. > From the bellow stdlog, I would guess the problem is with the toolbox call > IconDispatch ? No. The offset is _IconDispatch+015FA. 015FA hexadecimal equals 5626 decimal bytes. It's very unlikely that the _IconDispatch call takes up five thousand bytes of code. What you are encountering is some unnamed routine in the system; Macsbug simply reports the last symbol it could find with a very large offset. > Or maybe since it is in 68k mode that something went wrong > with a modeswitch? No, I don't see anything indicating this. > Does register d0 = 000000000 show something wrong? No. What has failed is the instruction at _IconDispatch+015FA. It says: MOVEA.L D0,A3 which moves the address pointed to by register A3 into register D0. The current contents of D0 are not relevant; they will be replaced by the value pointed to by register A3. But look at the address found in A3: it's BCFFA0A8, or between 2 and 3 gigabytes. This is an invalid address on your machine. That's why you're getting a bus error; you have tried to access memory which doesn't exist. > I also noticed at the very bottom it says: > Displaying memory from 0 > 00000000 FFC1 0000 FFC1 0000 0007 5C16 0007 5C18 o¡ooo¡oooo\ooo\o > 00000010 0007 5C1A 0007 5C1C FFC0 3048 FFC0 304A oo\ooo\oo¿0Ho¿0J > > Memory 0 is not a good memory spot is it? Isnt that a sign that you tried > to write to a null pointer? No, this happens to be irrelvant to this bug. This part of the STDLOG dump is a dump of the first few bytes of low memory so that an experienced debugger can tell whether it is likely that you have written over low memory. It has nothing to do with your crash; it's just more information that Macsbug's STDLOG dcmd is putting out to (hopefully) help you. > > What are some commands to further find out the toolbox calls that were > called before the crash? I do know a bit of mac toolbox. Look at the stack crawl to figure out where you came from. It's the part of the STDLOG dump which starts with the phrase "Calling chain using A6/R1 links". > Calling chain using A6/R1 links > Back chain ISA Caller > 00000000 PPC 03BD6568 > 045B6CF0 PPC 03B15C40 main+00034 > 045B6CB0 PPC 03B14560 MainEvent+00198 > 045B6C30 PPC FFD51FB8 WaitNextEvent+00028 > 045B6BF0 PPC 00637104 > 045B6AF8 68K 000775AE 'scod BFB1 0002'+019EE > 045B6ACC 68K 00633CD6 > 045B6AA8 68K 004C1E60 > 045B6A64 68K 00077636 'scod BFB1 0002'+01A76 > 045B6A52 68K 0007E374 'scod BFB1 0002'+087B4 > 045B6A04 68K FFC35E38 _Fix2Frac+001E4 > 045B69EC 68K FFC62886 _IconDispatch+03C36 > 045B69B0 68K 008D63E2 'MBDF 0000 0930'+00172 > 045B6946 68K 001D1DCC 'MBDF 0000 0002'+011CC The last call made, i.e. the code which called the failing routine, was 'MBDF 0000 0002'+011CC. The 'MBDF 0000 0002' tells you that the call was from a resource of type 'MBDF' (a menu bar drawing code resource) which came from file 0002 (the System file) and resource ID 0. (I can tell that 0002 is the system file by looking at the output of the files dmcd included as part of the STDLOG.) > Displaying File Control Blocks > fRef File Vol Type Fl Fork LEof > 0002 System g3 ide zsys dW rsrc #5498848 This MBDF was called from another MBDF in the Appearance Extension (again, the files dcmd tells me where I was). It was resource ID 0. > Displaying File Control Blocks > fRef File Vol Type Fl Fork LEof > 0930 Appearance Extension g3 ide INIT dw rsrc #611315 You can keep going up the chain like that, but I don't think it's going to be very helpful. I _would_ suggest you discover what things are going on in the 63xxxx range (you have two calls in the calling chain at that address range) and the 4C1E60 address. You probably have some extension or control panel which is patching a system call, such as WaitNextEvent or a system routine which WaitNextEvent calls, and is doing something incorrectly. Use the "wh" command to tell where this code is. It will probably just tell you you're at some offset from the beginning of a block of memory. Dump the beginning of that block of memory and see if there are any clues which may help you. Ultimately, from what the STDLOG is telling you, you're crashing several layers deep in menu drawing code. From my experience, this is probably because you have something installed which is patching a toolbox call incorrectly, but I can't tell much more than that from what is given. Good luck. Helpful web sites for Macsbug include: <http://www.macfixit.com/reports/MacsBug.shtml> "Macsbug for non-programmers" by Matt Deatherage. This fine article from the "Macintosh Weekly Journal" was republished by MacFixit <http://www.macfixit.com> <http://antioch-college.edu/~stefan/howtousemacsbug.html> How to Use MacsBug: A Tutorial by Stefan Anthony <http://www.scruznet.com/~crawford/Computers/macsbug.html> Secrets of Debug Meisters by Michael Crawford <http://www.mactech.com/articles/mactech/Vol.07/07.09/MacsBug-results/text.html> <http://www.mactech.com/articles/mactech/Vol.07/07.09/MacsBug-results-2/text.html> Getting Results With Macsbug by Jeff Turnbull in MacTech magazine. (This article is in two parts.) <http://developer.apple.com/dev/techsupport/develop/bysubject/testing_debugging.html> Develop Issue 8: "Macintosh Debugging: A Weird Journey into the Belly of the Beast" by Bo3b Johnson and Fred Huxham (In Adobe Acrobat PDF only) <http://developer.apple.com/dev/techsupport/develop/bysubject/testing_debugging.html> Develop Issue 13: "Macintosh Debugging:The Belly of the Best Revisited" by Fred Huxham and Greg Marriott (In Adobe Acrobat PDF only) <http://developer.apple.com/dev/techsupport/develop/issue22/balance.html> Develop Issue 22: "Balance of Power: MacsBug for PowerPC" by Dave Evans and Jim Murphy <http://developer.apple.com/dev/techsupport/develop/issue26/balance.html> develop issue 26: "Sleuthing Through Your Code" by Dave Evans. (For those who do not have English as your first language, a "Sleuth" is a detective. The title means "being a detective through your code.") <http://developer.apple.com/dev/techsupport/develop/issue27/balance.html> develop issue 27: "Balance of Power: Stalking the Wild Defect" by Dave Evans. -- blob at ricochet dot net
From: csaldanh@mae.carleton.ca (Chris Saldanha) Newsgroups: comp.sys.next.programmer Subject: Re: NSTextField updating Date: 20 Jul 1998 17:51:21 GMT Organization: computerActive Inc. Message-ID: <6p002p$jk8$1@bertrand.ccs.carleton.ca> References: <6oih2j$l5g$1@bertrand.ccs.carleton.ca> <6oip0f$t4u$1@sun579.rz.ruhr-uni-bochum.de> <6oivif$i7p$1@bertrand.ccs.carleton.ca> <35ADD4EA.84E116F4@ix.netcom.com> Mark Trombino (mtrombin@ix.netcom.com) wrote: : I've experience similar problems and found no workable solution. I try and : update a text field before spawning a new process and I can't for the life of : me get the text field to display before the new process starts. : flushWindowIfNeeded doesn't work, calling PSWait (or it's OpenStep : counterpart) after a display call didn't work, and neither did using : [[NSDPSContext currentContext] flush]. If you find something that works for : you Chris, would you please post it to the group? OK, so this is what I'm doing that works (thanks for the ideas everyone!): [statusText setStringValue: @"Processing data."]; [mainWindow display]; [[NSDPSContext currentContext] flush]; This seems to work properly and update the textfield right away. --Chris Chris Saldanha, Software Analyst -------------------------------------- computerActive, Inc |"The telephone was not invented by | csaldanh@computerActive.com (NeXT/MIME)| Alexander Graham Unitel" -Bell Ad | http://www.mae.carleton.ca/~csaldanh --------------------------------------
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: Odd behaviour when looking for modifiers Date: 20 Jul 1998 20:59:21 GMT Organization: none Message-ID: <6p0b39$62u$1@server.signat.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit In an inspector I'm building I'm trying to use the modifier keys to make a certain button mean something slightly different. So to look if the shiftKey is down I do a... NSEvent *theEvent = [[NSApplication sharedApplication] currentEvent]; and then I look at it using... ([theEvent modifierFlags] & NSShiftKeyMask) != 0) Everything works fine if I single click on the button in question, however when I hold down the button so that it posts multiple events, I only get the shift key action at the end. Now I think I know what's going on here, clearly the target's being called on the mouseUp (as you'd expect with a button I suppose). And since you hold down the mouse button on a auto, you only get one mouseUp. But this still strikes me as very strange, what events are being posted to make the button repeat - and why don't those events have the shift key recorded? I mean if the system goes into an internal tracking loop when the auto is fired, shouldn't the last event be the mouseDown and that have shift recorded? Maury
From: gnyswjgu@nounwantedemail.com Newsgroups: comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.pen Subject: ACCEPT CHECKS ONLINE $19.99 SPECIAL OFFER Date: 21 Jul 1998 05:07:20 GMT Organization: Vision Marketing Association Distribution: inet Message-ID: <6p17m8$4gu$704@nw001t.infi.net> INCREASE YOUR SALES IMMEDIATELY ACCEPT CHECKS ONLINE! Fax Email Phone ************************* CheckWriter 4 for Windows 95 http://www.vmacom.com/checksonline Million People Don't Have Credit Cards... But They Do Have CHECKING ACCOUNTS! ***DON'T LOOSE THESE SALES*** SPECIAL: THIS SOFTWARE REGULARLY SALES FOR $69.99 BUT THE 300 PURCHASERS GET AN INSTANT($50.00 DISCOUNT) ORDER AND DOWNLOAD TODAY FOR ONLY $19.99 *Increase Your Sales Instantly By Allowing Anyone to BUY - You'll get the sales your competitors loose *Never Risk Sending COD Again, Collect COD Fees Before You Ship - avoid the time to change the mind *Cut the "check in the mail" delay - you know, the one that never comes *Customers Save On Credit Card Finance Charges - consumers are more than every economically aware, show them you are too *Increase The Number of Impulse Purchases - Until now only those with credit cards could buy on impulse but now, you'll get the extra sales *Use For The Convenience Of Monthly Payments - Reduce those Late Payments by issuing Monthly Drafts Easy to use Windows 95 interface To order online today http://www.vmacom.com/checksonline ORDER TODAY ----------------------------------------------------------------- VISION MARKETING ASSOCIATION "Internet Commerce Solutions" Visit http://www.vmacom.com call 757-623-6769 Fax 757-623-2577 ----------------------------------------------------
From: MWRon@metrowerks.com (MW Ron) Newsgroups: comp.sys.mac.programmer.codewarrior,comp.sys.next.programmer Subject: Re: Will CodeWarrior (including C++ and MSL) exist for Rhapsody/MacOSXServer? Date: Mon, 20 Jul 1998 13:03:19 -0400 Organization: Metrowerks Corporation Message-ID: <MWRon-ya02408000R2007981303190001@enews.newsguy.com> References: <1dcbewy.d0tm6gw4npkkN@dbklap.bb.opentext.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <1dcbewy.d0tm6gw4npkkN@dbklap.bb.opentext.com>, dbk@mcs.com (Dan "Bud" Keith) wrote: >Ever since Apple announced their Carbon strategy, Metrowerks has been >silent about whether they will be supporting Rhapsody (now called MacOS >X Server) with their excellent CodeWarrior product. I think that there >is a command-line version of the compiler, mwcc, which can be used in >conjunction with Project Builder. This is correct. >I am more interested, however, in their MSL libraries for C and C++. I >am getting the uncomfortable feeling that Metrowerks is going to simply >wait until MacOSX is available before they involve themselves in Apple's >NeXT-based technology. This means no C++ and STL will be available for >Rhapsody, at least as far as Metrowerks is concerned. I'm afraid all I can say is.... We're working with Apple to get the final plan of rollout of our tools intermixed with their tools before we announce anything Ron -- The CodeWarrior FAQ is now open to the public. <http://cw-faq.mit.edu/cw-faq/faq> METROWERKS Ron Liechty "Software at Work" MWRon@metrowerks.com
From: pete@ohm.york.ac.uk (-bat.) Newsgroups: comp.sys.next.programmer Subject: awakeFromNib odd behaviour Date: 21 Jul 1998 11:19:43 GMT Organization: The University of York, UK Sender: pcf1@york.ac.uk Message-ID: <6p1tgf$97m$1@pump1.york.ac.uk> No, correct me if I am wrong, but isn't an object being unarchived from a .nib file supposed to get awakeFromNib before it gets anything else ? I have a custom view being unarchived from sucha .nib file, but it would appear that it is being sent a drawInrect mechind, followed by the awakeFromNib and then a second drawInRect method ! This is a right pain as it means I cannot simply set-up the object in awakeFromNib as it is called upon to draw itself before this happens. Grrr... bug or feature ? -bat.
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: NSView setBounds: method - does it work properly? Date: 21 Jul 1998 11:09:51 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6r8to2.100.rog@talisker.ohm.york.ac.uk> this is under openstep 4.2 on intel; i've come across some slightly odd behaviour in the NSView setBounds: method. the documentation for setBounds says: -(void)setBounds:(NSRect)boundsRect Sets the receiver's bounds rectangle to boundsRect. The bounds rectangle determines the origin and scale of the receiver's coordinate system within its frame rectangle. so i would assume that if i did: [view setBounds:r]; r = [view bounds]; then r would not have changed significantly (beyond rounding errors). however, this ain't the case. in the code snippet attached below, the rectangle passed to setBounds: is [(256, 0), (128, 128)], but the bounds actually change to [(85.333336, 0), (127.999992, 128)] is this the proper behaviour, or a bug? if the latter, then how much faith can i have in the rest of the bounds modification methods? how many more fundamental appkit methods are horribly broken? :-( cheers, rog. demo code: #import <AppKit/AppKit.h> const char *rect2str(NSRect r); int main(int argc, char **argv) { NSAutoreleasePool *pool; NSWindow *win; NSView *content; NSRect frame, bounds; pool = [[NSAutoreleasePool alloc] init]; [NSApplication sharedApplication]; frame.origin.x = frame.origin.y = 0; frame.size.width = frame.size.height = 384; bounds.origin.x = 256; bounds.origin.y = 0; bounds.size.width = bounds.size.height = 128; win = [[NSWindow alloc] initWithContentRect:frame styleMask:NSBorderlessWindowMask backing:NSBackingStoreRetained defer:NO]; content = [win contentView]; [content setBounds:bounds]; printf("set view bounds to: %s\n", rect2str(bounds)); printf("view bounds are actually: %s\n", rect2str([content bounds])); [win release]; [pool release]; return 0; } const char *rect2str(const NSRect r) { static char buf[80]; sprintf(buf, "[(%g, %g), (%g, %g)]", r.origin.x, r.origin.y, r.size.width, r.size.height); return buf; }
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: How about a JAR style file for OpenStep Apps? Date: 20 Jul 98 10:10:08 Organization: Is a sign of weakness Message-ID: <SCOTT.98Jul20101008@slave.doubleu.com> References: <6og76p$c3i$1@nnrp1.dejanews.com> <6opr3b$phm$1@supernews.com> In-reply-to: jcr.remove@this.phrase.idiom.com's message of 18 Jul 1998 09:49:31 GMT In article <6opr3b$phm$1@supernews.com>, jcr.remove@this.phrase.idiom.com (John C. Randolph) writes: ajmas@bigfoot.com may or may not have said: -> Would it be at all unreasonable to look at implenting the -> OpenStep .app folder as a uncompressed ZIP file that contains -> the same stuff, much in the same way as is done for the Java JAR -> file? Yes. Been there, done that (pretty much). It used to be the case that NeXTSTEP apps could either keep their resources in the .app directory or in extra segments of the executable (Mach-O) file. Keeping the resources embedded in the executable a' la' Mac resource forks makes it a pain to edit said resources. (Actually, you can still build the monolithic apps, but we don't 'cause it's a bad idea.) You can? What do you do about .nib wrappers? Actually, it wasn't that hard to muck with resources in Mach-O files. SegHoarker let you get at them, and by now I'm sure there would be a couple slicker ways to do it, if things hadn't changed. I actually suspect the driving force behind the change was not problems with resources in Mach-O files so much as how to prevent duplication in "fat" Mach-O files? So far as I recall, fat executables were literally a group of thin executables bundled together, so if resources were stored in the executable, you'd have gotten seperate duplicated resources for each type of executable. Of course, this would have been a non-starter for Windows, in any case. BTW, how many people realize that you can put both a Mach-O executable and a OS/Windows .exe file in the same .app directory and have them use the same resources? Now if we could only get Windows to work with directory wrappers... [Actually, I sort of wish they'd managed another way to fix the problem. Monolithic executables were more efficient, both in inode usage, and in general filesystem overhead. You didn't have to keep opening and closing files to read resources, you already had everything memory mapped. Just because this is hidden from the programmer doesn't mean it isn't happening ...] Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: Objective C woes Date: 20 Jul 98 10:00:00 Organization: Is a sign of weakness Distribution: world Message-ID: <SCOTT.98Jul20100000@slave.doubleu.com> References: <leffert.900164607@cs.uchicago.edu> <6offgr$d81@innsrv.roadnet.ups.com> <Pine.NXT.3.96.980714103331.3427D-100000@pathos> <SCOTT.98Jul14121433@slave.doubleu.com> <6oqq0c$ha1@mochi.lava.net> In-reply-to: arti@lava.DOTnet's message of 18 Jul 1998 18:37:00 GMT In article <6oqq0c$ha1@mochi.lava.net>, arti@lava.DOTnet (Art Isbell - remove "DOT") writes: scott@doubleu.com (Scott Hess) wrote: > My single biggest gripe about the foundation kit is: > > -(void)release; > > You used to use -(id)free, which led to the excellent coding > style of: > > obj=[obj free]; > > A compact way of saying "free the object and forget about it." > Now, you have to say "Free the object. Forget about the > object.": > > [obj release]; > obj=nil; One can take advantage of a standard C construct to emulate the NEXTSTEP semantics under OPENSTEP: obj = ([obj release], nil); If one integrates this pattern into a set of common Objective-C patterns, then the probability of accessing a freed object is reduced. Jeez, I was thinking you could have a define like: #define releaseObject( anObject) ([anObject release], nil) but I didn't really like that, because then readers will say "What in heck is releaseObject()?" But doing it _without_ the define seems sensible enough. Of course, now you're using the comma operator, which I suspect wouldn't be understood by a goodly chunk of your C readership :-), -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: Patrick.dot.Stein.@.eko.dot.de Newsgroups: comp.sys.next.programmer Subject: Fast 3D Graphics won't show up on MacOS X ? Date: 21 Jul 1998 12:51:34 GMT Organization: Institut fuer Informatik der Universitaet Muenchen Message-ID: <6p22sm$nb4$2@union.informatik.uni-muenchen.de> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit There is no special support for 3d-graphic cards in the driverkit. ( not even linedrawing or blitting ) So - How do we get decent 3d-hardware support in MacOS X ? -- bye bye - jolly =================================================================== Electronic mail IS insecure - please use PGP when sending a mail !! -----BEGIN PGP PUBLIC KEY BLOCK----- Version: 2.6.i mQCNAjJknRQAAAEEAMj65mWo/OICXpA6uzkwHCFhhlt4Z7O9coGEOhk/CCbhkFc5 iRFq/wDOtDHF88OxQeXMspE8uemLLoby45c7HDv7S6K05eZV2bqf0zqlPFmIPPD3 acdbBMKurc/bq5G0PDY6Q6+dFsSoNo6xDSA5K1e8kti1w8eJzBzVPjEoJOFJAAUR tDNQYXRyaWNrIFN0ZWluIDxqb2xseUBqb2tlci5wcHAuY2lzLnVuaS1tdWVuY2hl bi5kZT4= =fRKZ -----END PGP PUBLIC KEY BLOCK----- ===================================================================
From: Uwe Hees <Uwe.Hees@rz-online.de> Newsgroups: comp.sys.next.programmer Subject: Remote NSFont object Date: Tue, 21 Jul 1998 15:32:52 +0200 Organization: Rhein-Zeitung Message-ID: <35B49884.6700DD29@rz-online.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi all, has anybody ever tried to distribute an NSFont instance using the distributed objects mechanism? If I try, I can ask the remote object for name and size but using the proxy as an argument to -[NSText setFont:] makes the app hang while waiting for a response. The stack is in -[NSLayoutManager _fillAttributeCache].
From: "Jeremy Bettis" <jeremy@hksys.com> Newsgroups: comp.sys.next.programmer Subject: Re: NSTextField updating Date: Tue, 21 Jul 1998 09:31:17 -0500 Organization: Internet Nebraska Message-ID: <6p28nn$9ua$1@owl.inetnebr.com> References: <6oih2j$l5g$1@bertrand.ccs.carleton.ca> <6oip0f$t4u$1@sun579.rz.ruhr-uni-bochum.de> <6oivif$i7p$1@bertrand.ccs.carleton.ca> <35ADD4EA.84E116F4@ix.netcom.com> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 What works for me is this: [titleField setStringValue:string]; [titleField display]; [panel display]; PSFlush(); If fact you can make all windows redraw with this little titbit: [NSApp nextEventMatchingMask:NSAnyEventMask untilDate:[NSDate date] inMode:NSEventTrackingRunLoopMode dequeue:NO]; >I've experience similar problems and found no workable solution. I try and >update a text field before spawning a new process and I can't for the life of >me get the text field to display before the new process starts. >flushWindowIfNeeded doesn't work, calling PSWait (or it's OpenStep >counterpart) after a display call didn't work, and neither did using >[[NSDPSContext currentContext] flush]. If you find something that works for >you Chris, would you please post it to the group? -----BEGIN PGP SIGNATURE----- Version: PGP 5.5.5 iQA/AwUBNbSmNQwyFOHbBpZpEQKUsQCfUrQYWsDUz0NmsEHGiQqMNO4K02oAoKsz Sr4ltN4UcpZqX5c6FzxI9Y2B =yEpE -----END PGP SIGNATURE-----
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: Dead-key processing on NT Date: Tue, 21 Jul 1998 16:30:02 +0200 Organization: Square B.V. Message-ID: <35B4A5EA.2FD495C0@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit How do I tell the OpenStep toolkit to process the dead-keys? For example, if I change my keyboard to US-International the ' becomes a dead key and can be combined with characters like i and e. NeXT displays however 'i and not the single glyph I expected. Dutch doesn't use that much of dead keys, but many of our German customers do! Maurice. -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto:mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+ Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" and "to trick into believing or accepting as genuine something false and often preposterous"
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: How about a JAR style file for OpenStep Apps? Date: 21 Jul 1998 18:19:35 GMT Organization: http://www.supernews.com, The World's Usenet: Discussions Start Here Message-ID: <6p2m3n$gqt$1@supernews.com> References: <6og76p$c3i$1@nnrp1.dejanews.com> <6opr3b$phm$1@supernews.com> <SCOTT.98Jul20101008@slave.doubleu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: scott@nospam.doubleu.com Scott Hess may or may not have said: -> In article <6opr3b$phm$1@supernews.com>, -> jcr.remove@this.phrase.idiom.com (John C. Randolph) writes: -> ajmas@bigfoot.com may or may not have said: -> -> Would it be at all unreasonable to look at implenting the -> -> OpenStep .app folder as a uncompressed ZIP file that contains -> -> the same stuff, much in the same way as is done for the Java JAR -> -> file? -> -> Yes. -> -> Been there, done that (pretty much). -> -> It used to be the case that NeXTSTEP apps could either keep their -> resources in the .app directory or in extra segments of the -> executable (Mach-O) file. Keeping the resources embedded in the -> executable a' la' Mac resource forks makes it a pain to edit said -> resources. (Actually, you can still build the monolithic apps, but -> we don't 'cause it's a bad idea.) -> -> You can? What do you do about .nib wrappers? You create your nib files with a very old version of IB, and include them in the Mach-O file! ->Monolithic executables were more efficient, both in inode -> usage, and in general filesystem overhead. That's what Bud Tribble said, or so I've heard. -jcr
From: Jamie Hogg <jwh100@york.ac.uk> Newsgroups: comp.sys.mac.programmer.tools,comp.sys.newton.programmer,comp.sys.next.programmer,comp.unix.programmer Subject: Computer music Survey, U of York. Date: Tue, 21 Jul 1998 19:28:44 +0100 Organization: University of York Sender: jwh100@york.ac.uk Message-ID: <35B4DDDC.78AD1C83@york.ac.uk> Mime-Version: 1.0 Content-Type: multipart/alternative; boundary="------------07AFC5E331D76E35CF2FD6B3" --------------07AFC5E331D76E35CF2FD6B3 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi, my name is Jamie Hogg. I am researching a project for an MSc in Music Technology at York University. If you have time, I'd really appreciate it if you could spare me 10 minutes to visit my website and answer a few questions - the subject is intuitive computer music symbols. Thanks very much! http://www.york.ac.uk/~jwh100/Survey/SFront.htm --------------07AFC5E331D76E35CF2FD6B3 Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit <HTML> <B>Hi, my name is Jamie Hogg</B>. <P>I am researching a project for an MSc in Music Technology <BR>at York University. <BR>If you have time, I'd really appreciate it if you could spare me 10 <BR>minutes to visit my website and answer a few questions - the subject is <BR>intuitive computer music symbols. <BR>Thanks very much! <P> <A HREF="http://www.york.ac.uk/~jwh100/Survey/SFront.htm">http://www.york.ac.uk/~jwh100/Survey/SFront.htm</A> <BR> <BR> </HTML> --------------07AFC5E331D76E35CF2FD6B3--
From: pete@ohm.york.ac.uk (-bat.) Newsgroups: comp.sys.next.programmer Subject: Re: Fast 3D Graphics won't show up on MacOS X ? Date: 21 Jul 1998 18:00:02 GMT Organization: The University of York, UK Sender: pcf1@york.ac.uk Message-ID: <6p2kv2$4lg$1@pump1.york.ac.uk> References: <6p22sm$nb4$2@union.informatik.uni-muenchen.de> Patrick.dot.Stein.@.eko.dot.de writes: > > There is no special support for 3d-graphic cards in the driverkit. ( not even > linedrawing or blitting ) > So - How do we get decent 3d-hardware support in MacOS X ? There is a GLIDE driverkit availble for OpenStep 4.2, and Mesa has within it support for using GLIDE to do rendering which then gets copied to a buffer to render windows under Linux. mesa can be used to do offscreen rendering to use OpenGL on OpenStep so it should, in theory, be possible to get mesa to use GLIDE to render stuff into a buffer for displaying in an OpenStep window. Not that I've tried it yet... -bat.
From: quinlan@intergate.bc.ca Newsgroups: comp.sys.next.programmer Subject: Re: Fast 3D Graphics won't show up on MacOS X ? Date: Tue, 21 Jul 1998 20:54:26 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6p2v62$arg$1@nnrp1.dejanews.com> References: <6p22sm$nb4$2@union.informatik.uni-muenchen.de> In article <6p22sm$nb4$2@union.informatik.uni-muenchen.de>, Patrick.dot.Stein.@.eko.dot.de wrote: > So - How do we get decent 3d-hardware support in MacOS X ? You pray nightly that Apple implements OpenGL and designs a nice driver architecture for it? Brian Quinlan quinlan@intergate.bc.ca -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: "Darrell S. Mockus" <darrellm@earthlink.net> Newsgroups: comp.sys.next.programmer Subject: Problem accessing DisplayGroup objects.... Date: Tue, 21 Jul 1998 17:24:16 -0700 Organization: Mockus Consulting Message-ID: <35B5312F.A15232B1@earthlink.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I was hoping somebody there could help me. I am working with DisplayGroup objects and I can't seem to find a way to reference the attributes of the the Objects I retrieve. I created a Java app using the WebObjects wizard. The app can access the attributes as they are declared in the .wod file. But when I try to access the attributes in code in the .java file. ( I need to set a PK before insert), the object is assumed to be a java.lang.object and not of the custom class which it is part of. I sent the java.lang.object the getClass() message and it is returning the correct custom class, but I don't know how to set/retrieve those variables in that custom class. Is there something I have to do like return a dictionary and parse it? Thanks, --- Darrell -- =-=-=-=-=-=-=-=-=-=-=-===-=-=-=-=-=-=-=-=-=-=-=-=-==-=-=-=-=-=-= "To err is human, to really foul up you need a machine." =-=-=-=-=-=-=-=-=-=-=-=--=-=-=-=--=-=-=-=-=-=-=-=-==-=-=-=-=-=-= darrellm@earthlink.net
From: spamcancel@wupper.com Newsgroups: comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.pen Subject: cmsg cancel <6p17m8$4gu$704@nw001t.infi.net> Control: cancel <6p17m8$4gu$704@nw001t.infi.net> Date: 21 Jul 1998 08:25:55 GMT Message-ID: <cancel.6p17m8$4gu$704@nw001t.infi.net> Sender: gnyswjgu@nounwantedemail.com Excessive Multi-Posted spam article exceeding a BI of 20 cancelled by spamcancel@wupper.com. From was: gnyswjgu@nounwantedemail.com Subject was: ACCEPT CHECKS ONLINE $19.99 SPECIAL OFFER NNTP-Posting-Host was: pm10-102.orf.infi.net
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: Remote NSFont object Date: 21 Jul 98 10:48:14 Organization: Is a sign of weakness Message-ID: <SCOTT.98Jul21104814@slave.doubleu.com> References: <35B49884.6700DD29@rz-online.de> In-reply-to: Uwe Hees's message of Tue, 21 Jul 1998 15:32:52 +0200 In article <35B49884.6700DD29@rz-online.de>, Uwe Hees <Uwe.Hees@rz-online.de> writes: has anybody ever tried to distribute an NSFont instance using the distributed objects mechanism? If I try, I can ask the remote object for name and size but using the proxy as an argument to -[NSText setFont:] makes the app hang while waiting for a response. The stack is in -[NSLayoutManager _fillAttributeCache]. I suspect this is a non-starter. The problem is that the font object has significant out-of-band state. Put another way, you can distribute the Objective-C wrappers around windowserver operations, but you generally can't use all methods because the wrapped windowserver operations won't distribute. Naturally, you can still distribute the non-windowserver operations. So, you can distribute an NSView, but you can't make a distributed NSView subview or superview of a non-distributed NSView. But you could still send messages to it that will modify it's placement and whatnot within the context of the system it exists on... similar things apply to NSApplication, NSWindow, and generally anything else that isn't a pure data structure. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: Linking other files? Date: Wed, 22 Jul 1998 11:00:57 +0200 Organization: Square B.V. Message-ID: <35B5AA49.A5D52E2@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit How do I link a Windows resource file (.RES) so that the generated application contains the resources in that file? -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto:mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+ Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" and "to trick into believing or accepting as genuine something false and often preposterous"
Message-ID: <35B60704.DC3795E3@fair.nl> Date: Wed, 22 Jul 1998 17:36:36 +0200 From: Bert Oosterveen <bert.oosterveen@fair.nl> Organization: FAIR information services MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Timezone problem with OS4.2 under Dutch NT Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello, I have an English and Dutch version of NT. When running the same (small) application on the Dutch NT-system the following error occurs: Could not find data for default time zone, using GMT. This happens somewhere in the foundation classes. Can anyone give me some pointers what I should alter to prevent this? I've looked in the registry for date/time and also in the Foundation framework but I do not know what to modify. Thanks in advance Bert
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <20W91216.P5Q4WTLF@aslfjalsfjsal.net> Control: cancel <20W91216.P5Q4WTLF@aslfjalsfjsal.net> Date: 22 Jul 1998 12:13:44 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.20W91216.P5Q4WTLF@aslfjalsfjsal.net> Sender: asljalskdjflaks@aslfjalsfjsal.net Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <SM5O8PBB.A0V0FD85@jasljflasd.com> Control: cancel <SM5O8PBB.A0V0FD85@jasljflasd.com> Date: 22 Jul 1998 12:13:44 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.SM5O8PBB.A0V0FD85@jasljflasd.com> Sender: oereow@jasljflasd.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Christian Jensen <cejensen@winternet.com> Newsgroups: comp.sys.next.programmer Subject: Has anyone managed to compile mkhybrid? Date: 24 Jul 1998 01:12:57 GMT Organization: StarNet Communications, Inc. Message-ID: <6p8n2p$2ae$1@blackice.winternet.com> NNTP-Posting-Date: 24 Jul 1998 01:12:57 GMT I have attempted to compile the CD-ROM binary image-creation utility mkhybrid (a mkisofs variant) on my NSFIP box, with no success. If anyone has a binary they'd be willling to share, or some compile tips for this, please drop me a line. Thanks! --Chris ************************** Chris Jensen cejensen@winternet.com MIME, Sun, NeXTMail OK "Sacred cows make the best hamburger." --Mark Twain
From: gnyswjgu@nounwantedemail.com Newsgroups: comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.pen Subject: cmsg cancel <6p17m8$4gu$704@nw001t.infi.net> Control: cancel <6p17m8$4gu$704@nw001t.infi.net> Date: 24 Jul 1998 15:43:06 -0400 Organization: University of Economics and Business Administration, Vienna, Austria Sender: root@cantine.wu-wien.ac.at Distribution: inet Message-ID: <cancel.8.6p17m8$4gu$704@nw001t.infi.net> Spam Cancelled by news-admin@wu-wien.ac.at
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: NSFileWrapper? Newsgroups: comp.sys.next.programmer Message-ID: <35baa558.0@news.depaul.edu> Date: 26 Jul 98 03:41:12 GMT Is there any source code available that demonstrates usage of NSFileWrapper in a non-contrived situation? Or in a contrived situation, for that matter? Thx, Jon -- "... and subpoenas for all." - Ken Starr
From: uli@tallowcross.uni-frankfurt.de (Uli Zappe) Newsgroups: comp.sys.next.programmer Subject: PowerOff - what to do with an App that still has running subprocesses? Date: 27 Jul 1998 12:56:32 GMT Organization: NEXTTOYOU Message-ID: <6phte0$1tm$1@tallowcross.uni-frankfurt.de> NNTP-Posting-Date: 27 Jul 1998 12:56:32 GMT Hi, I'm working on a (NS 3.3) app that has a lot of long-running subprocesses; so it's quite possible such a process is running when the user wants to log out or power off the computer. So far, in this case I bring up an alert panel analogous to the "There are still unsaved documents" panels in the app:powerOffIn:andSave: delegate method. The panel allows the user to quit immidiately or wait till the subproces is finished. In the latter case a modal loop is waiting for the end of the subprocess before the app:powerOffIn:andSave: method returns. This works fine if the user is just logging out. It also seemed to work fine when I tested it with power off. However, when I restarted the computer after this power off, the /dev directory was corrupted. Since I can't think of any other problematic activity, it seems quite reasonable to assume that the modal loop prolonging the power off process by approx. 3 minutes had been the reason for that corruption. Now, my questions are: 1. Are there any known reasonable arguments to assume that indeed prolonging the power off process by not immidiatley returning from the app:powerOffIn:andSave: method could corrupt the filesystem, or, maybe even more specific, the dev directory? (But what would happen in this case if a user doesn't react immidiately to a "There are still unsaved documents" panel?) 2. If so, could this be remedied by sending an extendPowerOffBy: message with a very large number as argument (my app doesn't know in advance the amount of time the subprocess will still need, but my understanding is that the WorkspaceManager will power off anyway as soon as all apps have terminated)? 3. If not, is there any common sense how to deal with this situation? Thanks for your insight! Bye Uli -- _____________________________________________________________________ Uli Zappe E-Mail: uli@tallowcross.uni-frankfurt.de (NeXTMail,Mime,ASCII) PGP on request Lorscher Strasse 5 WWW: - D-60489 Frankfurt Fon: +49 (69) 9784 0007 Germany Fax: +49 (69) 9784 0042 staff member of NEXTTOYOU - the German NEXTSTEP/OPENSTEP magazine _____________________________________________________________________
From: boehring@biomed.ruhr-uni-bochum.de (Daniel Boehringer) Newsgroups: comp.sys.next.programmer Subject: Re: PowerOff - what to do with an App that still has running subprocesses? Date: 27 Jul 1998 17:07:31 GMT Organization: Ruhr-Universitaet Bochum, Rechenzentrum Message-ID: <6pic4j$nn5$1@sun579.rz.ruhr-uni-bochum.de> References: <6phte0$1tm$1@tallowcross.uni-frankfurt.de> NNTP-Posting-Date: 27 Jul 1998 17:07:31 GMT uli@tallowcross.uni-frankfurt.de (Uli Zappe) wrote: >Hi, >[...] >1. Are there any known reasonable arguments to assume that indeed prolonging >the power off process by not immidiatley returning from the >app:powerOffIn:andSave: method could corrupt the filesystem, or, maybe even >more specific, the dev directory? (But what would happen in this case if a >user doesn't react immidiately to a "There are still unsaved documents" >panel?) to my knowledge WM.app accesses the filesystem through the bsd api as any NS application and so i consider it highly unlikely that WM.app (distributes app:powerOffIn:andSave:) can corrupt your filesystem. you should however power down in verbouse mode to see if everything goes fine though. btw: having implemented app:powerOffIn:andSave: myself i have not experienced any FS-damage yet. daniel (http://www.biomed.ruhr-uni-bochum.de/~boehring/)
From: Pascal Bourguignon <pbourgui@afaa.asso.fr> Newsgroups: comp.sys.next.programmer Subject: Re: PowerOff - what to do with an App that still has running subprocesses? Date: 27 Jul 1998 18:54:53 GMT Organization: None Message-ID: <6piidt$mip$1@news.imaginet.fr> References: <6phte0$1tm$1@tallowcross.uni-frankfurt.de> <6pic4j$nn5$1@sun579.rz.ruhr-uni-bochum.de> boehring@biomed.ruhr-uni-bochum.de (Daniel Boehringer) wrote: >uli@tallowcross.uni-frankfurt.de (Uli Zappe) wrote: >>Hi, >>[...] >>1. Are there any known reasonable arguments to assume that indeed prolonging >>the power off process by not immidiatley returning from the >>app:powerOffIn:andSave: method could corrupt the filesystem, or, maybe even >>more specific, the dev directory? (But what would happen in this case if a >>user doesn't react immidiately to a "There are still unsaved documents" >>panel?) Could you save the state of the subprocesses to let them finish at next reboot? As a user, I find it quite irritating when my computer does not go down as soon as I ask. >to my knowledge WM.app accesses the filesystem through the bsd api as any NS application and so >i consider it highly unlikely that WM.app (distributes app:powerOffIn:andSave:) can corrupt your >filesystem. you should however power down in verbouse mode to see if everything goes fine though. How do you do that ? Powering down in verbose mode? >btw: >having implemented app:powerOffIn:andSave: myself i have not experienced any FS-damage yet. > > >daniel (http://www.biomed.ruhr-uni-bochum.de/~boehring/)
From: Ken MacLeod <ken@bitsko.slc.ut.us> Newsgroups: comp.sys.next.programmer Subject: Re: Use of Delegation Date: 27 Jul 1998 09:20:47 -0500 Organization: PSINet Message-ID: <m33ebnv0hs.fsf@biff.bitsko.slc.ut.us> References: <1dcap3y.14ssodz12gkhq8N@rhrz-isdn3-p52.rhrz.uni-bonn.de> <6oog7q$5vc@ragnarok.en.uunet.de> <1dccgr7.2f599p1ufuwvmN@rhrz-isdn3-p24.rhrz.uni-bonn.de> theisen@akaMail.com (Dirk Theisen) writes: > Holger Hoffstaette <holger@_REMOVE_THIS_.wizards.de> wrote: > > What questions do you have in particular? How the process works, > > or if there are certain usage patterns? > > I'm looking for applied examples which show the advantages of this > feature. Cases, where other approaches would have been much more > complicated or less flexible. The Casbah Project <http://www.ntlug.org/casbah/> has been developing a lightweight distributed objects protocol that is a look-alike generalization of Apple's (nee NeXT's) Distributed Objects. Part of the ``lightweight'' piece is that the on-the-wire protocol is a simple, self-describing format of dictionaries, lists, and scalars. To make the protocol work between different languages we made it the responsibility of the client to ensure that class, method, member, and type names are recognizeable to the server. The implementation includes two object serializers, a plain-text and a binary version, that simply convert native objects into the wire protocol. We could have built the logic for name-mapping into the serializers but that would have duplicated code between the serializers or forced them into a superclass that may not have been shareable with other serializers. Instead, we are using a name-mapping delegate that has methods for mapping the class, method, member, and type names as appropriate for the server you are communicating to. The delegate can also be shared with any number of connections to similar servers. You can also leave the delegate undefined if the server is of the same type as the client. The upper-layer module also has an authorization delegate similar to NSConnection's that authorizes messages on an object and method basis per session. [I'll drop a plug in here too and mention that we've got people working on Perl, Java, and a C version but that other languages are still available for adoption.] -- Ken MacLeod ken@bitsko.slc.ut.us
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <35bbe544.0@d2o7.telia.com> Control: cancel <35bbe544.0@d2o7.telia.com> Date: 28 Jul 1998 04:02:39 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.35bbe544.0@d2o7.telia.com> Sender: by_the_lake@jboat.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: boehring@biomed.ruhr-uni-bochum.de (Daniel Boehringer) Newsgroups: comp.sys.next.programmer Subject: Re: PowerOff - what to do with an App that still has running subprocesses? Date: 28 Jul 1998 10:25:39 GMT Organization: Ruhr-Universitaet Bochum, Rechenzentrum Message-ID: <6pk8v3$dla$1@sun579.rz.ruhr-uni-bochum.de> References: <6phte0$1tm$1@tallowcross.uni-frankfurt.de> <6pic4j$nn5$1@sun579.rz.ruhr-uni-bochum.de> <6piidt$mip$1@news.imaginet.fr> NNTP-Posting-Date: 28 Jul 1998 10:25:39 GMT Pascal Bourguignon <pbourgui@afaa.asso.fr> wrote: > [...] >How do you do that ? Powering down in verbose mode? > in /NextLibrary/Devices/System.config/Instance0.table add "Shutdown Graphics" = "No"; and reboot (off course ;-) daniel (http://www.biomed.ruhr-uni-bochum.de/~boehring/)
From: Jodell Bumatai <jbumatai@inprise.com> Newsgroups: comp.sys.next.programmer Subject: Re: Objective C woes Date: Tue, 28 Jul 1998 12:38:31 -0700 Organization: Inprise Corporation Message-ID: <35BE28B6.7511B31F@inprise.com> References: <leffert.900164607@cs.uchicago.edu> <6offgr$d81@innsrv.roadnet.ups.com> <Pine.NXT.3.96.980714103331.3427D-100000@pathos> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello, I just bought an NS 3.3 for Intel/Black. I plan to put it on my Intel machine. I decided to learn Obj-C because I have been doing research about NeXT. I have been using the RAD tools for C/C++/Java which run on Winx only. So, I am jumping right into the NS 3.3 without knowing exactly I am doing here. I am looking for some really good how to program books or tutorials online. I keep reading Obj-C is easier than C++/Java. Any more clues? Why is it easier? jyb Bill Bumgarner wrote: > See below for specifics. First some general comments. > > (1) Always name classes starting with a capital letter. I.e. Foo, not > foo. Likewise, always name variables starting with a lowercase letter and > capitalizing subsequent words.... i.e. aString, anArray, theCow, etc... > > (2) Whenever possible, use specific pointer types. Instead of: > > id f; > > Use: > > Foo *f; > > On Tue, 14 Jul 1998, Scott Hildebrand wrote: > > > > > Jonathan B. Leffert wrote in message ... > > ............................... > > >-- foo.m -- > > ......................... > > >- (void) setBar: (id) b > > >{ > > > [bar autorelease]; > > > bar = [b copy]; > > Depending on what you really want to do, it is likely that you don't want > to copy b. There are reasons to do so.... basic rules: > > If (b) is immutable, then just do "bar = [b retain]". If b is mutable, > you need to decide if you want your foo instance to see changes as they > happen... if so, just retain. If not, copy. > > Regardless, you should always use -release instead of -autorelease, though > it requires a couple of extra lines of code to prevent badness. -release > is more effecient in that it will release the object immediately instead > of adding yet-another-operation-during-autoreleasepool-death. > > Something like: > > - (void) setBar: (SomeClass *) b > { > if (bar != b) { > [bar release]; > bar = [b retain]; > } > } > > The above code works regardless of whether or not bar or b are nil. > > > >} > > > > > >@end > > > > > >Everything, I think, looks ok. Then, I created x.m to test the foo class: > > > > > >-- x.m -- > > >#import "foo.h" > > > > > >int main() > > >{ > > > id s = @"hello"; > > > id f = [foo init]; > > This is the crus of the problem. [foo init] sends the method -init to > the CLASS foo. You never allocated an instance of foo! Something like: > > Foo *f = [[Foo alloc] init]; > > Is probably much closer to what you wanted.... > > > > [f setBar: s]; > > Because f is not an instance of Foo, the above line fails with an > unrecognized method exception.... > > > > printf("%s\n",[[f getBar] cString]); > > The above could be written as: > > NSLog(@"%@", [f getBar]); > > The runtime will take care of converting whatever getBar returns by > calling -description. > > > > return 0; > > >} > > > > > > > > > > > > > One subtle point to remember is that in x.m, this line > > > > > [f setBar: s]; > > > > causes setBar to mark "bar" for auto-release. At this point bar doesn't > > point to anything. I believe the compiler set's it to NIL, which is why > > there is no negative side effect of this. Sending a message to a NIL p-obj > > does nothing. I would still flag it as an issue in a code review. > > Instance variables always start with a value of nil/0/NO/0.0 whatever. > Basically, teh objc runtime uses something like calloc() to allocate the > object structure. > > It is a defined standard that a message to nil is a no-op; in this > particular case [setting an ivar-- either when the ivar is nil and/or the > value it is to be set to is nil], this behaviour is probably a good thing. > > In general, you are quite correct in that it is a good idea to flag > reliance on messaging to nil as a potential issue in a code review. Not > because it is incorrect but because it relies on a behaviour of ObjC that > is not immediately apparent by reading the code-- it should be documented > OR simply do something like: > > - (void) someMethod: (NSButtonCell *) aButtonCell > > if ( (aButtonCell != nil) && ([aButtonCell isEnabled] == YES) ) { > ... > } > > Yes, it is slightly more verbose. Yes, the "== YES" is unnecessary. But > the above if test leaves no question in the programmers mind as to what is > going on. Likewise, it gives the compiler the maximum number of hints as > to what is going on and, as such, the compiler will generate warnings > indicative of potential problems. > > I started using the above style of extremely anally retentive programming > about four years ago [after reading Steve McConnell's awesome +Code > Complete+] and it has been tremendously rewarding: > > - code handoffs are less painful... > > - greatly reduces compile time errors and runtime bugs > > - i spend a hell of a lot less time fixing problems > > - the amount of code i write on a per-day basis has skyrockted... > > Of course, your mileage may vary. > > [oh, and finally, go read +Objective-C and Object Oriented Programming+ -- > it is a truly excellent book that is included with OpenStep or Rhapsody. > Wonderful read regarding dynamic OO design and development. It uses ObjC > as the language of demonstration, but is by no means limited to being > useful in the objc realm] > > b.bum
From: mlchan@ews.uiuc.edu (chan moshen l) Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.lang.objective-c Subject: Details on porting NEXTSTEP app to OPENSTEP Date: 29 Jul 1998 00:10:43 GMT Organization: University of Illinois at Urbana-Champaign Message-ID: <6plpa3$a36$1@vixen.cso.uiuc.edu> I might be porting an application written for NextStep to run on a windows box. Currently, the plan is to port it to OPENSTEP so it may run under YellowBox for for windows. Several difficulties in this matter. 1. I have never developed on NextStep 2. I have never developed using the OPENSTEP APIs 3. I currently do not know the Objective-C superset of C Besides for these hurdles, can someone point me in the right direction on starting this port? I've heard there are some tools put out by NEXT to help with this. Where can I get them? The application is not too large, and I should have plenty of time. I should be able to get all documentation on NEXTSTEP and OPENSTEP APIs. The makefile looks like it was generated by NeXT Project Builder. It has these lines which may be relavent to how the application was made: MAKEFILEDIR = /NextDeveloper/Makefiles/app MAKEFILE = app.make LIBS = -lMedia_s -lNeXT_s -include Makefile.preamble include $(MAKEFILEDIR)/$(MAKEFILE) -include Makefile.postamble -include Makefile.dependencies If anyone could point me in the right direction on getting started, and what info/hurdles I should know about - it would be greatly appreciated! --
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.next.programmer Subject: Re: Details on porting NEXTSTEP app to OPENSTEP Date: 29 Jul 1998 01:19:26 GMT Organization: MiscKit Development Message-ID: <6pltau$af7$1@news.xmission.com> References: <6plpa3$a36$1@vixen.cso.uiuc.edu> NNTP-Posting-Date: 29 Jul 1998 01:19:26 GMT mlchan@ews.uiuc.edu (chan moshen l) wrote: > > I might be porting an application written for NextStep to run > on a windows box. Currently, the plan is to port it to OPENSTEP > so it may run under YellowBox for for windows. > [...] can someone point me in the right direction > on starting this port? I've heard there are some tools put out by > NEXT to help with this. Where can I get them? The tools definitely exist on OPENSTEP 4.2. Install the OPENSTEP/Mach 4.2 OS and you'll have them. What I do is: (a) put this in my path in my .cshrc (or the *rc for your shell): /NextDeveloper/OpenStepConversion/UtilityScripts/shellscripts/ (b) create this alias: convertall convert -preprocess > conversion.log ; convert -makescripts >> conversion.log ; convert -all >> conversion.log Then cd to your project directory and type "convertall". Go get lunch. Sometime later that day, everything will all be converted. That doesn't mean you're done, though. Now the fun begins...try building the project and you'll see what I mean. Basically, the rest of the conversion is up to you--you have to apply your extensive knowledge of the NEXTSTEP and OPENSTEP APIs (and their differences) to fix all the "hard stuff". The conversion scripts primarily only catch the easy bits. If you read the documentation in the /NextDeveloper/OpenStepConversion area, you'll find that you can do the conversion in several steps instead of all at once, and it will tell you how, as well as give excellent details about what happens at each step (and some of the gotchas that you want to know about). I suggest that you may find it easier to take it in steps, at least the first time you try to do a conversion, and have that documentation memorized or close at hand when you do it. Me, I do it all at once because I find the steps get tedious and I've done enough conversions now to pretty much know what I'm doing. Since you admit you don't know either API too well, good luck. You'll know 'em pretty well by the time you're done, I suspect. Call it a baptism by fire... That will only get you started. I'm sure you'll be back here soon. Have fun storming the castle! On the bright side, if your OPENSTEP app uses only the OPENSTEP APIs, then it should be trivial to port it to Rhapsody--almost as trivial as a recompile. Nothing even remotely like the NEXTSTEP to OPENSTEP passage, that's for sure... (I've been finding the Rhapsody ports to be easier than trivial. OPENSTEP ports are rarely trivial.) -- Later, -Don Yacktman don@misckit.com <a href="http://www.misckit.com/don.html">My home page</a>
From: theisen@akaMail.com (Dirk Theisen) Newsgroups: comp.sys.next.programmer Subject: Re: Use of Delegation Date: Wed, 29 Jul 1998 03:38:20 +0200 Organization: University of Bonn, Germany / www.joint.org Message-ID: <1dcwjmv.7wvojh1hajm0wN@rhrz-isdn3-p46.rhrz.uni-bonn.de> References: <1dcap3y.14ssodz12gkhq8N@rhrz-isdn3-p52.rhrz.uni-bonn.de> <6oog7q$5vc@ragnarok.en.uunet.de> <1dccgr7.2f599p1ufuwvmN@rhrz-isdn3-p24.rhrz.uni-bonn.de> <m33ebnv0hs.fsf@biff.bitsko.slc.ut.us> Hello Ken, thanks for Your reply. Ken MacLeod <ken@bitsko.slc.ut.us> wrote: > Instead, we are using a name-mapping delegate that has methods for > mapping the class, method, member, and type names as appropriate for > the server you are communicating to. Do You change the delegate at runtime? This is what I'm interested in. Regards, Dirk -- No RISC - No fun http://theisen.home.pages.de/
From: maury@remove_this.istar.ca (Maury Markowitz) Newsgroups: comp.sys.next.programmer Subject: NSHighlightRect behaviour? Date: 28 Jul 1998 19:03:09 GMT Organization: none Message-ID: <6pl79d$l4e$2@server.signat.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I've been using this for some time now, but I recently found I couldn't make it work in the way I wanted it to, it seems it's lazy and only "drives" the window from places it knows about (like in drag-n-drop). What can I do to get it to draw in my own cursor tracking loops? Maury
From: uli@tallowcross.uni-frankfurt.de (Uli Zappe) Newsgroups: comp.sys.next.programmer Subject: Re: PowerOff - what to do with an App that still has running subprocesses? Date: 28 Jul 1998 14:40:50 GMT Organization: NEXTTOYOU Message-ID: <6pknti$h9$1@tallowcross.uni-frankfurt.de> References: <6phte0$1tm$1@tallowcross.uni-frankfurt.de> <6pic4j$nn5$1@sun579.rz.ruhr-uni-bochum.de> <6piidt$mip$1@news.imaginet.fr> NNTP-Posting-Date: 28 Jul 1998 14:40:50 GMT Pascal Bourguignon <pbourgui@afaa.asso.fr> wrote: > Could you save the state of the subprocesses to let them finish at next > reboot? Unfortunately no. > As a user, I find it quite irritating when my computer does not go > down as soon as I ask. Well, it's up to you. You are asked by a panel what you want (you can as well choose to shutdown immidiately), and if you choose to wait, a panel will show that you're still waiting for this process (and this panel allows you to abort the process at any time, too). Bye Uli -- _____________________________________________________________________ Uli Zappe E-Mail: uli@tallowcross.uni-frankfurt.de (NeXTMail,Mime,ASCII) PGP on request Lorscher Strasse 5 WWW: - D-60489 Frankfurt Fon: +49 (69) 9784 0007 Germany Fax: +49 (69) 9784 0042 staff member of NEXTTOYOU - the German NEXTSTEP/OPENSTEP magazine _____________________________________________________________________
From: uli@tallowcross.uni-frankfurt.de (Uli Zappe) Newsgroups: comp.sys.next.programmer Subject: Re: PowerOff - what to do with an App that still has running subprocesses? Date: 28 Jul 1998 14:45:31 GMT Organization: NEXTTOYOU Message-ID: <6pko6b$h9$3@tallowcross.uni-frankfurt.de> References: <6phte0$1tm$1@tallowcross.uni-frankfurt.de> <6pic4j$nn5$1@sun579.rz.ruhr-uni-bochum.de> NNTP-Posting-Date: 28 Jul 1998 14:45:31 GMT boehring@biomed.ruhr-uni-bochum.de (Daniel Boehringer) wrote: > to my knowledge WM.app accesses the filesystem through the bsd > api as any NS application and so i consider it highly unlikely > that WM.app (distributes app:powerOffIn:andSave:) can corrupt > your filesystem. My suspicion would be that WM.app starts the BSD shutdown script at a time when my subprocess is still running. This *could* be harmful theoretically, couldn't it? Bye Uli -- _____________________________________________________________________ Uli Zappe E-Mail: uli@tallowcross.uni-frankfurt.de (NeXTMail,Mime,ASCII) PGP on request Lorscher Strasse 5 WWW: - D-60489 Frankfurt Fon: +49 (69) 9784 0007 Germany Fax: +49 (69) 9784 0042 staff member of NEXTTOYOU - the German NEXTSTEP/OPENSTEP magazine _____________________________________________________________________
From: uli@tallowcross.uni-frankfurt.de (Uli Zappe) Newsgroups: comp.sys.next.programmer Subject: Re: PowerOff - what to do with an App that still has running subprocesses? Date: 28 Jul 1998 14:42:51 GMT Organization: NEXTTOYOU Message-ID: <6pko1b$h9$2@tallowcross.uni-frankfurt.de> References: <6phte0$1tm$1@tallowcross.uni-frankfurt.de> <6pic4j$nn5$1@sun579.rz.ruhr-uni-bochum.de> NNTP-Posting-Date: 28 Jul 1998 14:42:51 GMT boehring@biomed.ruhr-uni-bochum.de (Daniel Boehringer) wrote: > btw: > having implemented app:powerOffIn:andSave: myself i have not experienced any FS-damage yet. Well, I wouldn't suppose it's the general use of this method. I suppose it's the long waiting period of several minutes, if anything Bye Uli -- _____________________________________________________________________ Uli Zappe E-Mail: uli@tallowcross.uni-frankfurt.de (NeXTMail,Mime,ASCII) PGP on request Lorscher Strasse 5 WWW: - D-60489 Frankfurt Fon: +49 (69) 9784 0007 Germany Fax: +49 (69) 9784 0042 staff member of NEXTTOYOU - the German NEXTSTEP/OPENSTEP magazine _____________________________________________________________________
From: ajmas@my-dejanews.com Newsgroups: comp.sys.next.programmer Subject: [Q] OpenStep and window manger Date: Tue, 28 Jul 1998 17:43:53 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6pl2ko$95a$1@nnrp1.dejanews.com> Having recently worked with x-windows and looked at some of the window managers, Enlightenment being one that comes to mind. I was wondering whether the window managers designed for X-Windows could eventually be adapted to work with OpenStep, or does OpenStep work differently enough that I should be looking at some other approach to customize the user interface? Thanks AJ -- http://www.bigfoot.com/~ajmas/ -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: someguy@netcom.ca (Justin McKillican) Newsgroups: comp.sys.next.programmer Subject: Re: Fast 3D Graphics won't show up on MacOS X ? Date: Tue, 28 Jul 1998 02:24:17 -0400 Organization: NETCOM Canada Message-ID: <someguy-2807980224170001@ott-on3-17.netcom.ca> References: <6p22sm$nb4$2@union.informatik.uni-muenchen.de> <6p2kv2$4lg$1@pump1.york.ac.uk> NNTP-Posting-Date: 29 Jul 1998 02:30:45 EDT In article <6p2kv2$4lg$1@pump1.york.ac.uk>, pete@ohm.york.ac.uk (-bat.) wrote: > > There is no special support for 3d-graphic cards in the driverkit. ( not even > > linedrawing or blitting ) > > So - How do we get decent 3d-hardware support in MacOS X ? > > There is a GLIDE driverkit availble for OpenStep 4.2, and Mesa has > within it support for using GLIDE to do rendering which then gets copied to > a buffer to render windows under Linux. mesa can be used to do offscreen > rendering to use OpenGL on OpenStep so it should, in theory, be > possible to get mesa to use GLIDE to render stuff into a buffer for displaying > in an OpenStep window. Conix has released a beta i think (not sure if it's release or dev.) of their OpenGL drivers. Carmack of ID ported QuakeGL to Rhapsody, and he says everything works fine. justin
Message-ID: <35BE78F8.F477402D@home.com> From: msms <msms@home.com> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.lang.objective-c Subject: Re: Details on porting NEXTSTEP app to OPENSTEP References: <6plpa3$a36$1@vixen.cso.uiuc.edu> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Wed, 29 Jul 1998 01:19:56 GMT NNTP-Posting-Date: Tue, 28 Jul 1998 18:19:56 PDT Organization: @Home Network Well, the OPENSTEP APIs are not all the same as the Yellow Box APIs, so your best bet is to look on the MacOS X Server site from Apple for information on developing for Yellow Box. You'll also need to get your hands on MacOS X Server so you can do your developing, as at this time you won't be able to develop for Yellow Box on OPENSTEP (at least to my knowledge). -- John ------------------- chan moshen l wrote: > > I might be porting an application written for NextStep to run > on a windows box. Currently, the plan is to port it to OPENSTEP > so it may run under YellowBox for for windows. > > Several difficulties in this matter. > 1. I have never developed on NextStep > 2. I have never developed using the OPENSTEP APIs > 3. I currently do not know the Objective-C superset of C > > Besides for these hurdles, can someone point me in the right direction > on starting this port? I've heard there are some tools put out by > NEXT to help with this. Where can I get them? > > The application is not too large, and I should have plenty of time. I should > be able to get all documentation on NEXTSTEP and OPENSTEP APIs. The makefile > looks like it was generated by NeXT Project Builder. It has these lines > which may be relavent to how the application was made: > > MAKEFILEDIR = /NextDeveloper/Makefiles/app > MAKEFILE = app.make > LIBS = -lMedia_s -lNeXT_s > -include Makefile.preamble > include $(MAKEFILEDIR)/$(MAKEFILE) > -include Makefile.postamble > -include Makefile.dependencies > > If anyone could point me in the right direction on getting started, > and what info/hurdles I should know about - it would be greatly appreciated! > --
From: "Jonathan Hendry" <jhendry@cmg.fcnbd.com> Newsgroups: comp.sys.next.programmer Subject: Re: PowerOff - what to do with an App that still has running subprocesses? Date: 29 Jul 1998 02:50:44 GMT Organization: Steel Driving Software Message-ID: <01bdbac5$bebf8cb0$ddb5dccf@samsara> References: <6phte0$1tm$1@tallowcross.uni-frankfurt.de> <6pic4j$nn5$1@sun579.rz.ruhr-uni-bochum.de> <6pko6b$h9$3@tallowcross.uni-frankfurt.de> Uli Zappe <uli@tallowcross.uni-frankfurt.de> wrote in article <6pko6b$h9$3@tallowcross.uni-frankfurt.de>... > boehring@biomed.ruhr-uni-bochum.de (Daniel Boehringer) wrote: > > to my knowledge WM.app accesses the filesystem through the bsd > > api as any NS application and so i consider it highly unlikely > > that WM.app (distributes app:powerOffIn:andSave:) can corrupt > > your filesystem. > > My suspicion would be that WM.app starts the BSD shutdown script at a time > when my subprocess is still running. This *could* be harmful theoretically, > couldn't it? First, what do your subprocesses do? Are they accessing the filesystem? Second, do they handle signals gracefully? I'd guess that the BSD shutdown process sends a signal (15?) to the running processes. If your subprocesses don't handle them gracefully, and just keel over, they might end up corrupting the filesystem. Alternately, perhaps they *don't* die, and keep going after the filesystem has done its final sync, leaving it in an inconsistent state. - Jon
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <hYov1.9359$fh3.698347@iagnews.iagnet.net> Control: cancel <hYov1.9359$fh3.698347@iagnews.iagnet.net> Date: 28 Jul 1998 18:43:17 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.hYov1.9359$fh3.698347@iagnews.iagnet.net> Sender: greatbiz@acomco.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: uli@tallowcross.uni-frankfurt.de (Uli Zappe) Newsgroups: comp.sys.next.programmer Subject: Re: PowerOff - what to do with an App that still has running subprocesses? Date: 29 Jul 1998 14:27:02 GMT Organization: NEXTTOYOU Message-ID: <6pnbfm$18j$1@tallowcross.uni-frankfurt.de> References: <6phte0$1tm$1@tallowcross.uni-frankfurt.de> <6pic4j$nn5$1@sun579.rz.ruhr-uni-bochum.de> <6pko6b$h9$3@tallowcross.uni-frankfurt.de> <01bdbac5$bebf8cb0$ddb5dccf@samsara> NNTP-Posting-Date: 29 Jul 1998 14:27:02 GMT "Jonathan Hendry" <jhendry@cmg.fcnbd.com> wrote: > > My suspicion would be that WM.app starts the BSD shutdown script at a > > time when my subprocess is still running. This *could* be harmful > > theoretically, couldn't it? > > First, what do your subprocesses do? Are they accessing the filesystem? Yes, reading a lot of files, and writing some. However, they don't access the /dev files (which were the only ones corrupted - don't know if by chance or not) (hmm, maybe apart form syslogd which uses /dev/log - not sure, though). > Second, do they handle signals gracefully? They're supposed to do this, yes. > I'd guess that the BSD shutdown process sends a signal (15?) to the > running processes. If your subprocesses don't handle them > gracefully, and just keel over, they might end up corrupting > the filesystem. Alternately, perhaps they *don't* die, and keep > going after the filesystem has done its final sync, leaving it > in an inconsistent state. The latter is my suspicion. Anyway, right now I'm sending an extendPowerOffBy:36000000 message to WM.app, and haven't experienced the problem since. But I think there's a lot more shutdown processes necessary until this gains statistical relevance. Bye Uli _____________________________________________________________________ Uli Zappe E-Mail: uli@tallowcross.uni-frankfurt.de (NeXTMail,Mime,ASCII) PGP on request Lorscher Strasse 5 WWW: - D-60489 Frankfurt Fon: +49 (69) 9784 0007 Germany Fax: +49 (69) 9784 0042 staff member of NEXTTOYOU - the German NEXTSTEP/OPENSTEP magazine _____________________________________________________________________
From: "jdm" <jdm1intxDIE_SPAMBOT_DIE@airmail.net> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Wed, 29 Jul 1998 10:26:05 -0500 Organization: Airnews.net! at Internet America Message-ID: <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> Abuse-Reports-To: abuse at airmail.net to report improper postings NNTP-Proxy-Relay: librarytest.airnews.net NNTP-Posting-Time: Wed Jul 29 10:27:45 1998 Kris Sheglova <95155580@brookes.ac.uk> wrote >Your average jo bloggs can learn to program a computer using simple >systems e.g. BASIC if they need to knock up a program or two every so >often but these should not become programs relied on by the masses. > Typical egotistical arrogant elitist British attitude. Now we know why all the Joe Bloggs came to America, and proceeded to program rings around the snotty Brits. jdm
From: nurban@crib.corepower.com (Nathan Urban) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 29 Jul 1998 08:50:15 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6pn5q7$6sp$1@crib.corepower.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> In article <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net>, "jdm" <jdm1intxDIE_SPAMBOT_DIE@airmail.net> wrote: > Kris Sheglova <95155580@brookes.ac.uk> wrote > >Your average jo bloggs can learn to program a computer using simple > >systems e.g. BASIC if they need to knock up a program or two every so > >often but these should not become programs relied on by the masses. > Typical egotistical arrogant elitist British attitude. Now we know why all > the Joe Bloggs came to America, and proceeded to program rings around the > snotty Brits. Bigot. He's right, you know. There's a vast difference between the programming skills of different people. One of my CS profs mentioned some study that found that even among professional programmers (and this isn't even counting the average Joe), there can be as much as an order of magnitude variation between skill levels (by various metrics, e.g., code size, efficiency, time-to-completion, bug count, reliability, etc.) And training makes an enormous difference, whether it's academic, on-the-job, or whatever. A self-taught BASIC programmer is not likely to be banging out large, reliable, mission-critical designs.
From: boehring@biomed.ruhr-uni-bochum.de (Daniel Boehringer) Newsgroups: comp.sys.next.programmer Subject: Re: PowerOff - what to do with an App that still has running subprocesses? Date: 29 Jul 1998 16:58:32 GMT Organization: Ruhr-Universitaet Bochum, Rechenzentrum Message-ID: <6pnkbo$1qp$1@sun579.rz.ruhr-uni-bochum.de> References: <6phte0$1tm$1@tallowcross.uni-frankfurt.de> <6pic4j$nn5$1@sun579.rz.ruhr-uni-bochum.de> <6pko6b$h9$3@tallowcross.uni-frankfurt.de> NNTP-Posting-Date: 29 Jul 1998 16:58:32 GMT uli@tallowcross.uni-frankfurt.de (Uli Zappe) wrote: > >My suspicion would be that WM.app starts the BSD shutdown script at a time >when my subprocess is still running. This *could* be harmful theoretically, >couldn't it? > harmful to your files, not to the FS. should it be able do induce FS-damage be by userlevel-code i would consider the FS being unexcusably misimplemented (the OS should (and in fact will) prevent writing to files on unmounted filesystems). furthermore from watching the messages on the console during shutdown i reason that unmounting is done only after all processes are killed (the shutdownscript will not proceed unless _every_ process exited).
From: mmh@dcs.qmw.ac.uk (Matthew M. Huntbach) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Followup-To: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Date: 29 Jul 1998 17:17:42 GMT Organization: Queen Mary & Westfield College, London, UK Message-ID: <6pnlfm$ob$2@beta.qmw.ac.uk> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> jdm (jdm1intxDIE_SPAMBOT_DIE@airmail.net) wrote: > Kris Sheglova <95155580@brookes.ac.uk> wrote > >Your average jo bloggs can learn to program a computer using simple > >systems e.g. BASIC if they need to knock up a program or two every so > >often but these should not become programs relied on by the masses. > Typical egotistical arrogant elitist British attitude. Now we know why all > the Joe Bloggs came to America, and proceeded to program rings around the > snotty Brits. Sorry? Read just about any textbook on programming or software engineering used by reputable teachers the world over, and you'll see there's a clear message that there's a world of difference between a program sloppily hacked together that appears to do the job, and a program carefuully designed that we can be reasonably certain does the job and can be easily understood and modified for future use. You owe an apology to Mr.Sheglova (who is quite possibly not British, as "Sheglova" is not a British name) for your quite unnecessary nationalistic remark. I doubt you have the guts to make one. Matthew Huntbach
From: Greg_Anderson@afs.com (Gregory H. Anderson) Newsgroups: comp.sys.next.programmer Subject: Re: NSHighlightRect behaviour? Date: 29 Jul 1998 19:35:28 GMT Organization: Anderson Financial Systems Inc. Message-ID: <6pnti0$a84@shelob.afs.com> References: <6pl79d$l4e$2@server.signat.org> Maury Markowitz writes > I've been using this for some time now, but I recently found I > couldn't make it work in the way I wanted it to, it seems it's lazy > and only "drives" the window from places it knows about (like in > drag-n-drop). Maury, as of Wednesday afternoon, your email is not working: 554 MX list for istar.com. points back to lancelot.mgn.net 554 <maury@istar.com>... Local configuration error Don't be surprised if you don't get any further answers to your query. I tried to respond to your email and got bounced. -- Gregory H. Anderson | "We're in the land of the blind, Visionary Ophthalmologist | selling working eyeballs, and they Anderson Financial Systems | balk at the choice of color." -- Tony greg@afs.com (NeXTmail OK) | Lovell, on Mac user reactions to NeXT
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.programmer Subject: Re: Fast 3D Graphics won't show up on MacOS X ? Date: 29 Jul 1998 22:00:59 GMT Organization: Digital Fix Development Message-ID: <6po62r$4l2$1@news.digifix.com> References: <6p22sm$nb4$2@union.informatik.uni-muenchen.de> <6p2kv2$4lg$1@pump1.york.ac.uk> <someguy-2807980224170001@ott-on3-17.netcom.ca> In-Reply-To: <someguy-2807980224170001@ott-on3-17.netcom.ca> On 07/27/98, Justin McKillican wrote: >In article <6p2kv2$4lg$1@pump1.york.ac.uk>, pete@ohm.york.ac.uk (-bat.) wrote: > >> > There is no special support for 3d-graphic cards in the driverkit. ( >not even >> > linedrawing or blitting ) >> > So - How do we get decent 3d-hardware support in MacOS X ? >> >> There is a GLIDE driverkit availble for OpenStep 4.2, and Mesa has >> within it support for using GLIDE to do rendering which then gets copied to >> a buffer to render windows under Linux. mesa can be used to do offscreen >> rendering to use OpenGL on OpenStep so it should, in theory, be >> possible to get mesa to use GLIDE to render stuff into a buffer for displaying >> in an OpenStep window. > >Conix has released a beta i think (not sure if it's release or dev.) of >their OpenGL drivers. Carmack of ID ported QuakeGL to Rhapsody, and he >says everything works fine. He also said that it software OpenGL (specifically this implementation) is too slow to play the game on, but fine for a server. -- Scott Anguish <sanguish@digifix.com> Stepwise - OpenStep/Rhapsody Information <URL:http://www.stepwise.com>
From: "jdm" <jdm1intxDIE_SPAMBOT_DIE@airmail.net> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Wed, 29 Jul 1998 23:23:14 -0500 Organization: Airnews.net! at Internet America Message-ID: <8F12A6803BCB46BB.69250C85F5006041.55EF65CE3294D5C9@library-proxy.airnews.net> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <6pnlfm$ob$2@beta.qmw.ac.uk> Abuse-Reports-To: abuse at airmail.net to report improper postings NNTP-Proxy-Relay: library.airnews.net NNTP-Posting-Time: Wed Jul 29 23:24:49 1998 Matthew M. Huntbach wrote in message <6pnlfm$ob$2@beta.qmw.ac.uk>... > >Sorry? Read just about any textbook on programming or software engineering >used by reputable teachers the world over, and you'll see there's a clear >message that there's a world of difference between a program sloppily >hacked together that appears to do the job, and a program carefuully designed >that we can be reasonably certain does the job and can be easily understood >and modified for future use. Anybody can write a textbook. And anybody can proclaim themselves to be the embodiment and zenith of all wisdom and knowledge pertaining to computer science. That doesn't make them any better of a programmer than the poor weenie with a degree from Acme Technical Institute and Pizzeria, cranking out code in a low-paying 3rd rate sweatshop in Redmond. For every genius with a Phd in Computer Science from MIT you can show me, I can show you a guy who never even learned what a linked list was, but can still turn out decent code in a plain-jane OLTP application. Sure, maybe he will scratch his head and say "duh?" if you ask him to write a indexing routine based on AVL trees, but that doesn't mean he isn't an inferior programmer. >You owe an apology to Mr.Sheglova (who is quite possibly not British, as >"Sheglova" is not a British name) for your quite unnecessary nationalistic >remark. I doubt you have the guts to make one. > If I felt I had anything to apologize for, I would certainly have the guts to. However, Mr. Sheglova's sneering, condescending attitude towards "the masses" speaks for itself. I will not apologize to him, nor will I apologize for my Nationalistic tendencies in the face of arrogant Europeans who feel the need to patronize us atavistic Amerikan schweinhundts. FIE UPON ALL ELITISTS, EVERYWHERE !!! YOU ARE NOTHINGS !!!! jdm
From: "InterPoint" <interpoint@centrenet.co.uk> Subject: GERMANY - NEXT/OPENSTEP JOBS Newsgroups: comp.sys.next.programmer Message-ID: <01bdbb96$4f17f2e0$759249c2@cna40.centrenet.co.uk> Date: Thu, 30 Jul 1998 08:41:06 GMT NNTP-Posting-Date: Thu, 30 Jul 1998 09:41:06 BST Our client, a leading European Next/Openstep consultancy, is urgently seeking consultants at all levels of experience for major blue-chip projects in Germany. You should have experience in at least one of… Next/Openstep Web Objects Enterprise Objects Java A background in OO methodology would be a strong plus. Our client offers an excellent remuneration package including stock options, structured public & private health insurance in addition to a highly competitive salary & a friendly, professional & team-oriented corporate culture. German language skills are desirable but not essential. …please mail us in complete confidence at interpoint@centrenet.co.uk +44 411-262359
From: quinlan@my-dejanews.com Newsgroups: comp.sys.next.programmer Subject: Re: Fast 3D Graphics won't show up on MacOS X ? Date: Thu, 30 Jul 1998 07:19:20 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6pp6po$1fc$1@nnrp1.dejanews.com> References: <6p22sm$nb4$2@union.informatik.uni-muenchen.de> <6p2kv2$4lg$1@pump1.york.ac.uk> <someguy-2807980224170001@ott-on3-17.netcom.ca> <6po62r$4l2$1@news.digifix.com> In article <6po62r$4l2$1@news.digifix.com>, sanguish@digifix.com (Scott Anguish) wrote: > He also said that it software OpenGL (specifically this > implementation) is too slow to play the game on, but fine for a > server. It's difficult to tell from his plan file but I believe that he was refering to Rhapsody when he said that "The game isn't really playable with the software emulated OpenGL, but it functions properly, and it makes a fine dedicated server." Why would their server (I assume that they use it to host games) need OpenGL suppost? BTW, SGI has ported OpenGL to the PC. Maybe Apple should pay them to port it to Rhapsody/OS X? -- Brian Quinlan quinlan@intergate.bc.ca -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: Malcolm Steel <malcolm.steel@gecm.com> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Thu, 30 Jul 1998 10:01:00 +0100 Organization: GEC-Marconi Avionics Ltd Message-ID: <35C0364B.7F00@gecm.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <6pnlfm$ob$2@beta.qmw.ac.uk> <8F12A6803BCB46BB.69250C85F5006041.55EF65CE3294D5C9@library-proxy.airnews.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit jdm wrote: <merged from another message> > Typical egotistical arrogant elitist British attitude. Now we know why all > the Joe Bloggs came to America, and proceeded to program rings around the > snotty Brits. <snip> > If I felt I had anything to apologize for, I would certainly have the guts > to. However, Mr. Sheglova's sneering, condescending attitude towards "the > masses" speaks for itself. I will not apologize to him, nor will I > apologize for my Nationalistic tendencies in the face of arrogant Europeans > who feel the need to patronize us atavistic Amerikan schweinhundts. > > FIE UPON ALL ELITISTS, EVERYWHERE !!! YOU ARE NOTHINGS !!!! Seems like a typical reaction from a teenager reaching puberty, rebelious, tunring against their parents, thinking they know best about everything :-) jdl, you probably wont realise but this is not directed to you individually :-)
From: eagle@nestje.nl (Eagle) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Thu, 30 Jul 1998 08:59:53 GMT Organization: EagleSoft Message-ID: <35d134b4.8083696@news.telecom.ptt.nl> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <HaS8aCAeaCw1UAdO@ffaltd.demon.co.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Foul Ole Andy Hunt millennium hand and shrimped the following: <snip> >>Typical egotistical arrogant elitist British attitude. Now we know why all >>the Joe Bloggs came to America, and proceeded to program rings around the >>snotty Brits. >> >>jdm >> >Ouch!!! ..... sorry got go, must blow my nose and polish my bowler hat. >> >> >Toodle-pip! Let's make jdm watch Monty Python's "The meaning of Life", particularly the jungle scene ("Oh dear, yes, it does seem something had a little chew on my legs..."). Then again, he probably did... But seriously. jdm is obviously American. Only they can be ignorant and arrogant simultaneously... _____________________________________________________________ Eagle - Squawk! Bug'r'em. Millennium hand and shrimp! - Foul Ole Ron _____________________________________________________________
From: Andy Hunt <andy@ffaltd.demon.co.uk> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Thu, 30 Jul 1998 08:54:06 +0100 Organization: Foster Findlay Associates Ltd Distribution: world Message-ID: <HaS8aCAeaCw1UAdO@ffaltd.demon.co.uk> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> MIME-Version: 1.0 In article <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library- proxy.airnews.net>, jdm <jdm1intxDIE_SPAMBOT_DIE@airmail.net> writes >Kris Sheglova <95155580@brookes.ac.uk> wrote >>Your average jo bloggs can learn to program a computer using simple >>systems e.g. BASIC if they need to knock up a program or two every so >>often but these should not become programs relied on by the masses. >> > > >Typical egotistical arrogant elitist British attitude. Now we know why all >the Joe Bloggs came to America, and proceeded to program rings around the >snotty Brits. > >jdm > Ouch!!! ..... sorry got go, must blow my nose and polish my bowler hat. > > Toodle-pip! Andy -- Andy Hunt
From: uli@tallowcross.uni-frankfurt.de (Uli Zappe) Newsgroups: comp.sys.next.programmer Subject: Re: PowerOff - what to do with an App that still has running subprocesses? Date: 30 Jul 1998 08:54:43 GMT Organization: NEXTTOYOU Message-ID: <6ppccj$f9$1@tallowcross.uni-frankfurt.de> References: <6phte0$1tm$1@tallowcross.uni-frankfurt.de> <6pic4j$nn5$1@sun579.rz.ruhr-uni-bochum.de> <6pko6b$h9$3@tallowcross.uni-frankfurt.de> <6pnkbo$1qp$1@sun579.rz.ruhr-uni-bochum.de> NNTP-Posting-Date: 30 Jul 1998 08:54:43 GMT boehring@biomed.ruhr-uni-bochum.de (Daniel Boehringer) wrote: > uli@tallowcross.uni-frankfurt.de (Uli Zappe) wrote: > >My suspicion would be that WM.app starts the BSD shutdown script at a time > >when my subprocess is still running. This *could* be harmful > >theoretically, couldn't it? > > harmful to your files, not to the FS. > should it be able do induce FS-damage be by userlevel-code i would consider > the FS being unexcusably misimplemented (the OS should (and in fact will) > prevent writing to files on unmounted filesystems). > > furthermore from watching the messages on the console during shutdown i > reason that unmounting is done only after all processes are killed (the > shutdownscript will not proceed unless _every_ process exited). Hmm, *all* processes? It's a process itself, isn't it? Anyway, as I said at least so far an extendPowerOffBy:36000000 message to WM.app seems to help, and while on the one hand I'd really like to know what exactly was/is going on, on the other hand I have little inclination to intentionally provoke more file system crashes on my working systems's harddisk just to verify my thesis that without the extendPowerOffBy: message file system crashes would occur again... Bye Uli -- _____________________________________________________________________ Uli Zappe E-Mail: uli@tallowcross.uni-frankfurt.de (NeXTMail,Mime,ASCII) PGP on request Lorscher Strasse 5 WWW: - D-60489 Frankfurt Fon: +49 (69) 9784 0007 Germany Fax: +49 (69) 9784 0042 staff member of NEXTTOYOU - the German NEXTSTEP/OPENSTEP magazine _____________________________________________________________________
From: cb@bartlet.df.lth.se (Christian Brunschen) Newsgroups: comp.sys.next.advocacy,comp.sys.mac.advocacy,comp.sys.next.programmer,comp.sys.mac.programmer Subject: Re: Mac OS X UI Followup-To: comp.sys.next.programmer,comp.sys.mac.programmer Date: 30 Jul 1998 12:12:35 +0200 Organization: The Computer Society at Lund Message-ID: <6ppguj$def$1@bartlet.df.lth.se> References: <6odtu6$lh3$5@lwnws01.ne.highway1.com> <andyba-ya02408000R2807981200340001@news> <6plk25$jp6$1@bartlet.df.lth.se> <andyba-ya02408000R2907981658200001@news> NNTP-Posting-User: cb (Followups set to comp.sys.{next,mac}.programmer) In article <andyba-ya02408000R2907981658200001@news>, Andy Bates <andyba@corp.webtv.net> wrote: >In article <6plk25$jp6$1@bartlet.df.lth.se>, cb@bartlet.df.lth.se >(Christian Brunschen) wrote: > >> If I write a letter that is never intended to go anywhere but be sent to a >> printer, it need never be a file at all. > >It's a file as soon as you select "New" from the File menu. Except this application I'm using doesn't have a 'file' menu, it has a 'Document' menu. And indeed, I create a new document. But since I never save it to the file system, it is never a file. Get it ? > >> The data in memory might be stored as something that has no remote >> resemblance to a file: it might be a linked list, a tree, whatever; need >> not be a file. > >Who says that a linked list can't be a file? Who says that a tree can't be >a file? They all have a marked beginning, and a set way to traverse the >data, and a set end. They are all files. No they are not. You can _store_ a representation of a linked list, a tree, or whatever, _in_ a file, but the tree _is not_ a file. In particular, a tree or linked list don't usually have only _one_ boundary in memory - they are collections of data which may be spread out through memory, with links (pointers) making a _logical_ structure out of _physically_ distinct data. > >> >The data IS collected and stored in a well-defined place. In some cases, >> >that place is an area of memory. It is possible to create a file, edit it, >> >and print it, without it ever being saved to disk. However, it is still a >> >file! >> >> No it's not - I write a letter, which I may or may not store in a file. > >You can choose whether or not to store it in a file on the disk; however, >as soon as you start the letter ("Untitled"), the program starts a file in >memory. No - the program starts _some data structure_ in memory. This data structure could be stored in a file, but it never is, due to my choice of never saving it to a file. Thus, there is never a file. > >> The letter itself is _not_ a file, the file is a _container_ for the >> letter where I may choose to place it. > >A file is a collection of data, plain and simple. The letter in memory is a >file. Now, when you save it to disk it may have additional data to make it >accessible to the filesystem on the disk, but it is still a file when it is >in memory. It just happens to be a file that is only accessible to the >program that created it. > >> The fact that I can store the >> letter in a file does not make the letter a file, nor does it make every >> possible place that I could store the letter a file. > >I'm sorry, but you're wrong. Just because it's a file when it's on a disk >doesn't mean that it's NOT a file if it's NOT on a disk. You keep mentioning 'on disk' - I keep saying 'in a file system'. As soon as you remove data from the file system - ie _take it out of the container_ - then it is no longer in the file, and is not a file. You fail rto understand this. I don't know why. > >> >Or for that matter, what would you call the items that you save to a RAM >> >disk? They aren't stored on a disk filesystem; does that mean that they >> >aren't files? >> >> I gave an example in my post, where I explicitly said that a ramdisk is a >> good place for files - after all, a ramdisk has a _file system_, doesn't >> it ? > >Why are you hung up on this "file system" concept? A file can exist outside >of a file system! Or, if you prefer, each program has its own file system >which it uses to keep track of its current files in memory. So basically you are saying that everything in a computer that somehow organizes data, is a file system? You will find opposition from most people who program computers. > >> >Yes, it was a file. Saying that "it was a bunch of text" doesn't change >> >that fact. You called it "a bunch of text." "A bunch." A "bunch" is another >> >word for a collection, and text is just data. Therefore, "a bunch of text" >> >is just a collection of data, which is, surprise!, a file. You can get out >> >a thesaurus and call it "an amalgalm of glyphs," but it's still a file. >> >> Just because a file is a collection of data, does not make every >> collection of data a file. It's a one-way relationship - an implication, >> not an equivalence. > >A file is a collection of data with distinct boundaries and a distinct >structure. A collection of data with distinct boundaries and a distinct >structure is a file. They are equivalent. No they are not. A file is indeed a collection of data with distinct boundaries and a distinct structure - that part is true; this is a one-sided _implication_. The implication in the other direction - ie, saying that any collection of data that is bounded and structured is also a file - is _not there_. > A word-processing document saved >to disk is a file. Usually, yes. I could also store a word processing document in _several_ files, you know. That doesn't make those several files one file. >A word-processing document stored in memory is a file. Not unless it's stored in a file system in memory (such as a RAMdisk). > >> If I place a bunch of data together, I have just that - a bunch of data. >> If I put that data in a file, I have a file... which contains a bunch of >> data. > >If you put a bunch of data together in a distinct structure with distinct >boundaries, then it is a file. No it's not. Please dig up a reference that claims this to be true, since I perceive this to be your crucial misunderstanding of the issue. > >> >No, saving it doesn't "turn it into" a file; it is ALREADY a file. >> >> No - by saving a bunch of data in a specific way, you define it to be in a >> file. This is what distinguishes any collection of data, from a file. > >No, the area (whether it is in memory or on disk) where the program (or >file system) stores the data is what distinguishes any collection of data >from a file. So, how about a tree structure in memory, that is spread out all over the place? There is no one beginning or one end, there are in fact lots of beginnings and endings, all connected by references this and that way. This would break your concept of a File, would it not? Also, consider a typical hard disk. A disk is divided into sectors; each sector has a destinct beginning, ending, and collects a certain amount of data. Is, thus, every sector on a disk a file? > >> Exactly - they speak about 'documents', 'reports', etc - they talk about >> the contents of the files, because that is what is important to them. They >> take a file's contents out of the file (and into memory) to edit it, > >No, they don't take it out of the file; it still exists on the hard disk. A >copy of that data (with a distinct structure and boundaries -- a file) is >stored in memory. I didn't say they removed it, they take it out and build a distinct structure in memory - a structure that bears no resemblance to a file, usually. > >> then >> store it back to a file when they are done. > >After the file in memory has been modified, it is written back to the disk, >and given a specific label to distinguish it from all of the other files on >the filesystem. No, after the data in memory was modified, it was stored in a file in the file system. The application does not have a concept of a 'file' in memory - only the concept of a 'file' in the file system. > >> The file is a container for >> data - the data is the data. > >That's correct. And that container can exist on a hard disk, or in memory. >Why do you think that the container in memory is any less of a "file"? You can very well have such a container in memory - as I have mentioned RAM disks repeatedly. But that does not make _every_ possibly data in memory a file. You claim an _equivalence_ where only an _implication_ exists. > >> >By your logic, an airplane is only an airplane while it's in the air. If it >> >lands, it's a car. If it lands in the ocean, it's a submarine. >> >> You misunderstand - the airplane is an airplane when it is put together as >> a functional airplane. > >How do you know if it's functional if it's on the ground? You can really >only know that it's functional if it's in the air. Therefore, by your >logic, it is not an airplane while it's on the ground. You can check the functionality of an airplane even while on the ground - also, you are trying to nit-pick an analogy, are you aware of that? There will _always_ be differences between analogy and reality. >And how is a file in memory less "functional" than a file on disk? What can >you do with a file on disk that you can't do with a file in memory? Now you are adding new stuff - I never claimed any of that. Please don't claim that I said or implied things that I clearly didn't. To adress the point, however: A file on a RAMdisk - and this in a file system in memory - is exactly as functional as a file on any other physical medium, be it a hard disk, a floppy, a tape, a magnetic drum, an optical medium .... A linked list, or a tree, in memory are in fact _much more_ functional functional that a file is. This is largely because you are not limited to a file's concept of one beginning, one ending, and one set of data in between. In fact, the closest thing to a file in memory (and now I'm talking about application memory _as opposed to_ a ramdisk) is an array. Very nice data structure, but not what you need for all sorts of things. _But_ you can almost always take another data structure and store it in an array, just as you can take this same data structure and store it in a file. In fact, the mere _existence_ of the word 'array' - which denotes a data collection with one beginning, one end, and the data all in between, and thus means much the same as 'file' - should be taken as a counter-argument to your claims. An 'array' need not be in an 'array system' - whereas a 'file' is in a 'file system'. An array in memory can be represented _exactly_ in a file in a file system, and vice versa. For most other data structures, this is not the case - they have to be 'translated' to a distinctly different representation for storage in a file. > >> >The program that operates on that data has to have some way of >> >distinguishing that data from the rest of the data floating around memory. >> >That organization of data that it maintains is called a file. >> >> Not really - that organization may be called a list, an array, a tree, a >> hash table .... you actually only very rarely use the same representation >> in memory as you might use in a file. > >I never said that you had to use the same representation in memory as you >do on the disk! But if you traverse both the disk file and the file in >memory from the beginning, and stop as the end, you will get the exact same >output. Umm ----- _NO_. If I have a linked list which just _happens_ to be stored in reverse in memory, the applicatino doesn't care; as long as the pointers between the different parts of the list all point correctly, the application works. Likewise if the list is spread out all over the application's memory. Whereas if I store this list to a file, the situation is completely different. > >> Furthermore, it does not claim that >> the contents of memory is a file - but rather that you store a >> representation of the data in memory, in a file. > >First of all, who or what is "it"? "it" I think was the quote from Webster's, which you included. >And yes, a representation of the data is >stored in memory, in a file. Isn't that what I've been saying all along? I phrased the above badly. It is intended to mean 'you take a representation of the data that you have in memory and store it in a file'. (careful reading will reveal that this is indeed one of the possible interpretations.) If I put some parentheses around the groupings I intended we get you store (a representation of (data in memory)) in a file rather than the you store (a representation of (data) (in memory, in a file)) which you apparentlty read it as. > >> Note: "a representation" >> - still a different thing; the data in the file and the data in memory are >> most often distinctly _different_. > >Yes, the file in memory and the file on the disk can be organized >differently. No one ever said that one file has to have the exact same >organizational structure as another. But the data in memory is not in a file. > >> Saving data to a file usually does >> things with the data, preparing it for storage in a file; the organization >> of data in memory and in a file differ greatly. > >Yes, the program prepares the file in memory for storage to a file on the >disk; we are in perfect agreement here. However, and again, just because >they are organized differently doesn't make one a file and the other not a >file. But the data in memory is not 'in a file' - it is in some sort of data structure that uses different parts of memory, is not bounded by one beginning and one ending, and is generally lots of things _except a file_. Get it? > >> >A file in memory has some sort of delimitation which marks where it begins, >> >where it ends, and which order it goes in. A random collection of data does >> >not have these distinguishing characteristics, and is therefore not a file. >> >However, and collection of data that a program will be working with in >> >memory DOES have these delimiters, and is therefore a file. >> >> Most collections of data in memory have multiple markers of beginning, >> ending, and whatnot - and they usually differ greatly from similar >> markings used for files. > >So? They are different, but they still exist. A word processing file (or >any other type of file) will have a distinct structure and distinct >boundaries. > >> A file has _one_ beginning and _one_ ending. > >Right. Distinct boundaries. And a distinct structure. Just like a word >processing document in memory. A file. Ahh -- but a word processing document in memory does not have 'one beginning and one ending'. It does have conceptually - but it is rarely stored as such. Often it can be stored as, say, a list of paragraphs, where each paragraph might be at some place in memory. The third textual paragraph might be placed earlier in memory than the second textual paragraph, for instance... > >> When >> you load the file into your word processor, the word processor interprets >> the contents of the file and builds a _structure_ of data in memory - a >> structure that is usually totally different than what is in the file. > >The structure of the file in memory may be totally different from the >structure of the file on disk, yes. Good. >You are hung up on calling the file on disk, "the file," when in fact it is >"the disk file." No - 'the file' as in 'the ordered collection of data in the file system'. That which distinguishes data in a file from data _not_ in a file. _I_ don't say you have to store a file on disk - I just say that most data in memory is _not a file_, whereas you claim that all data in memory is a file. But to have 'files' you have to have a 'file system', otherwise the concept falls. > >> When >> you then save the text to another file (or the same one), the contents of >> memory are written to the file. Still, the contents of memory while you >> are editing the text are _not_ a file. > >Yes, they are. A file can exist outside of a file system. By your logic, a >car is only a car while it's in a garage. If I take a file out of the file system, what is there to say that it is a file? A file is a file _only_ in the presence of a file system, otherwise it is just data. Yes, you _can have_ a file system with only one file, but you still have a file system.... The car is still a car because it functions as a car. The file, outside of the file system, does not function as a file. If you _treat_ it as a file, then you _are_ defining a 'file system'. > >> >When you write a letter that is stored in memory, what happens when you >> >print it out? Does it print out the complete contents of the computer's >> >entire memory? No, it just prints out the collection of data that you >> >entered in. The computer knows where that collection starts, where it >> >stops, what it contains, and what order it goes in. It is a file. A file >> >system is a structure for storing files (specific delimited collections of >> >data) onto a storage medium. That doesn't mean that a file can not exist >> >outside of the file system. By that logic, an orange peel isn't garbage >> >until you put it in a garbage can. >> >> But that is indeed true - an orange peel may be a lot of things besides >> garbage. It could be the centre piece of the stilleben scene I am making a >> painting of. It could be that I want to shred & boil the orange peel to >> make jam out of it. > >All correct. > >> An orange peel isn't garbage until I decide that it is garbage, and treat >> it like garbage. > >Wrong. It is garbage AS SOON AS YOU DECIDE that it is garbage! It does NOT >become garbage only when you put it in the garbage can! > >> Likewise, data isn't a file until I begin to treat it >> like a file. > >And see, your analogy breaks down. "Treating it like a file" entails more >than just saving it to disk. Yes - it means defining it as a file within a file system. > It involves having a distinct structure and >order. The program treats the collection like a file, and therefore it is a >file. But the program doesn't treat the collection like a file, so it's not a file. > >> And data in memory in use by an application is rarely treated >> as a file, and thus is not a file. > >Wrong. It is treated like a file by the program. It is stored (in whatever >form) as a collection of data. The same mistake again: the fact that it is an ordered collection of data does not make it a file. To reiterate: 1) The fact that something is a file _does_ make it an ordered colection of data. 2) The fact that somethign is an ordered collection of data _does not_ make it a file. Get that into your head and you will see the fallacy of your arguments. You are arguing based on an incorrect premise, of course you arrive at the wrong conclusions. > >> >No, not every combination >> >of text in memory is a file. But then again, not every combination of data >> >on a hard disk is a file either. Certain markings distinguish one file from >> >another on a hard disk; similar markings distinguish one file from another >> >in memory. Why is that so hard to comprehend? >> >> Because it is a false claim. >> >> Why is it so hard for you to comprehend that what is inside most >> applications while they run is not a file, but data in another form? > >What other form? Why do you think that certain "forms" can't be files? A linked list, a tree, a hash table ... they _are not_ files. You can _store them_ in files, perhaps, but they _are not_ files. If you have them in memory, you do not treat them as files - since you disregard their physical boundaries and are concerned, instead, with their logical collections. Whereas in a file, the physical boundaries of the whole shebang are a vital characteristic that makes up a file. Get it ? > >> Your extract from Webster's does not claim that every ordered collection >> of data is a file, does it? That is the misinterpretation you make, I >> think. Again, the fact that a file is an ordered collection of data does >> not make every ordered collection of data a file - just like the fact that >> every whale is a mammal does not make all mammals whales. > >Every collection of data with a distinct structure and distinct boundaries >is a file. They are equivalent. No they are not. Either get this mistaken idea out of your head, or provide good evidence that you are in fact correct and that I am wrong. > >> Another example: Fractint is a popular program that lets you draw fractal >> images. You select a fractal formula to use, you select the parameters to >> give, you select resolution and color map (how the results of the >> calculation should be turned into an image), and then you let the program >> work; it calculates an image and presents it to you on screen. >> >> Now, in what way is this image a file ? It most certainly is not in any >> file format you'll ever have come across; it is, in fact, stored in memory >> in a fashion that is quite distinct from how images are stored in files. > >The bitmap representation of the fractal equation is stored in memory. The >program knows which pixel is the first one, which is the last one, and what >order they go in. Yes; but this order need not be the physical order they are in, thus you can't just dump them directly to a file. Again, in a linked list, you need to look at the contents of the data (the pointers) to see the structure; a file is well-defined (as is an array) by its beginning and end. Try to treat a linked list like a file (by just bothering about its beginning and end) and see where you end up. >Furthermore, I'll wager that you can save that file as a >TIFF file, a GIF file, or whatever graphics format you choose. Yes - I can _save it as_ a file - but it is not a file. >It is a file. Nope, it isn't. > >> Now Fractint offers an option to save an image to a file, and one to load >> an image from a file. However, each of those operations transform between >> a file (in the file system) and an image (in application memory). The two >> - the image and the file - have a well-defined relationship, but are not >> identical, they are not the same thing. > >They don't have to have the same format, but they ARE both files. I never >said that a file in memory and a file on disk have to have the same format. >However, there is a distinct one-to-one relationship between a given disk >file and a given file in memory. Two files that share that relationship are >the same file. But the 'file in memory' need not even bear a shred of resemblance to the 'file in the file system'. It need not be contiguous in memory; it may not even be all in memory at the same time; thus it's not a file. > >> The data that the program works with _is not_ a file - but a file can be >> generated from it. >> >> Why is _this_ so difficult for _you_ to understand ? > >I understand that you are wrong. Well, you _mis_understand, since you are the one who is wrong. > >Andy Bates. // Christian Brunschen
From: Malcolm Steel <malcolm.steel@gecm.com> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Thu, 30 Jul 1998 11:42:16 +0100 Organization: GEC-Marconi Avionics Ltd Message-ID: <35C04E08.77B8@gecm.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit jdm wrote: > > Kris Sheglova <95155580@brookes.ac.uk> wrote > >Your average jo bloggs can learn to program a computer using simple > >systems e.g. BASIC if they need to knock up a program or two every so > >often but these should not become programs relied on by the masses. > > > > Typical egotistical arrogant elitist British attitude. <snip> So how do you judge this comment made by an American ? "The testing field has progressed far beyond the point where it is likely that someone not thoroughly familiar with the literature would be likely to make a useful contribution. Today, such research is typically done at the MS or more likely, PhD level." No offense aimed at the author. Regards Malcolm Steel
From: nurban@crib.corepower.com (Nathan Urban) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 30 Jul 1998 06:52:11 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6ppj8r$bcs$1@crib.corepower.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <6pnlfm$ob$2@beta.qmw.ac.uk> <8F12A6803BCB46BB.69250C85F5006041.55EF65CE3294D5C9@library-proxy.airnews.net> In article <8F12A6803BCB46BB.69250C85F5006041.55EF65CE3294D5C9@library-proxy.airnews.net>, "jdm" <jdm1intxDIE_SPAMBOT_DIE@airmail.net> wrote: > Matthew M. Huntbach wrote in message <6pnlfm$ob$2@beta.qmw.ac.uk>... > >Sorry? Read just about any textbook on programming or software engineering > >used by reputable teachers the world over, and you'll see there's a clear > >message that there's a world of difference between a program sloppily > >hacked together that appears to do the job, and a program carefuully > >designed that we can be reasonably certain does the job and can be easily > >understood and modified for future use. > Anybody can write a textbook. And anybody can proclaim themselves to be the > embodiment and zenith of all wisdom and knowledge pertaining to computer > science. This apparently includes you. Hypocrite. > That doesn't make them any better of a programmer than the poor > weenie with a degree from Acme Technical Institute and Pizzeria, cranking > out code in a low-paying 3rd rate sweatshop in Redmond. Sometimes it doesn't. It often does. > For every genius > with a Phd in Computer Science from MIT you can show me, I can show you a > guy who never even learned what a linked list was, but can still turn out > decent code in a plain-jane OLTP application. Which may or may not be efficient, maintainable, easily expandable, reliable, or something to be relied upon by the masses. There's a reason why the field known as "software engineering" has arisen. > Sure, maybe he will scratch > his head and say "duh?" if you ask him to write a indexing routine based on > AVL trees, but that doesn't mean he isn't an inferior programmer. If indexing routines based on AVL trees are the superior solution, then it indeed makes him the inferior programmer, at least for that particular problem. A large part of being a good programmer is not being ignorant of the best solutions. > If I felt I had anything to apologize for, I would certainly have the guts > to. Rationalization is the purest form of hypocrisy. You don't have the guts to apologize, so you invent reasons for not doing so. > However, Mr. Sheglova's sneering, condescending attitude towards "the > masses" speaks for itself. I will not apologize to him, At any rate, Sheglova's post was not at all sneering nor condescending. The fact of the matter is that most people suck at programming, and if all they do is putz around by themselves with simple systems like BASIC they are generally not going to produce something I want to rely upon in an important situation. (There are, of course, exceptions, as with anything.) But quality software engineering is something that usually requires outside training in some form, whether it be academic or on-the-job. I notice that you also fallaciously conflate training with "academia". Sheglova was _not_ arguing that a Ph.D. or even a degree was required to produce good programs. Ever hear of a word called "context". If you had actually paid attention to the context of the thread and the full contents of his post, you would have noticed that he was arguing that _specialized training_ was required, more than the rudimentary basic training that most people in fields other than computer programming (_not_ just academic computer science programs) have had, and that it is not possible to produce sophisticated software products without such training. That doesn't mean that it has to come from an advanced academic degree (though it often helps). His electrical wiring analogy was: electrical wiring is inherently a specialized field. If you try to "simplify" things to the point where any guy off the street (i.e., w/o specialized training) can do it, you will create a dangerous situation. That doesn't mean that someone without a degree can't learn how to be a good electrician. It means that if you try to make electrical wiring into a field where the average untrained individual can walk in and fool around with it, bad things will happen. It is inherently complex. You totally missed Sheglova's point in your eagerness to confirm your preconceived notions of others and slander a foreigner, and he deserves an apology from you. Personally, I don't think my CS degree helped me much as a professional developer, as I learned little there that I didn't already know; most of my expertise came previously, on my own, or later, on the job. However, I had far more self-taught programming experience than most. But it just ain't true that some Schmoe off the street is going to be able to pick up a book and instantly turn himself into a marvelous software engineer. Before or after my degree, I still didn't have enough experience with real-life problems to do that. And most "Jo Bloggses" who are not programmers but have tinkered around a little with BASIC or whatever just aren't capable of producing a modern, maintainable, reliable software product. That requires much more specialized training. > nor will I > apologize for my Nationalistic tendencies in the face of arrogant Europeans What, and if an American had said the same thing, then that would be different? Because I agree with him, as do most of the other professionals who have contributed to this newsgroup, most of whom are American. Why do you think that his remarks are characteristically European, when there are more Americans than Europeans on this thread who agree with him? Indulging in generalizations and stereotypes, are we? I'm afraid that just reinforces an unfortunate sterotype of _Americans_, as prejudiced against foreigners. Which is not entirely inaccurate, as you have proven. > who feel the need to patronize us atavistic Amerikan schweinhundts. I'll patronize you myself, because you're a bigoted juvenile moron who's probably jealous of superior programmers with more training than you. I'm American, and I think that you're the reason why people of other countries rightly deserve to sneer at us. There's too many of your kind around here. You're just as arrogant, even more bigoted, and also with a healthy dose of the usual American anti-intellectualism that's harming the educational system and our country's ability to compete.
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6pq3th$kjk$3506@artemis.backbone.ou.edu> Control: cancel <6pq3th$kjk$3506@artemis.backbone.ou.edu> Date: 30 Jul 1998 16:12:01 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6pq3th$kjk$3506@artemis.backbone.ou.edu> Sender: <vista@telepath.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: "Michael A. Thompson" <mat0001@jove.acs.unt.edu> Newsgroups: comp.sys.next.programmer Subject: XServer Date: Thu, 30 Jul 1998 12:48:16 -0500 Organization: University of North Texas Message-ID: <35C0B1E0.E691AFD6@jove.acs.unt.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Anybody know of a free X server that works under NS3.3 black? tried mouseX but it doesnt seem to work. Video goes black and I changed permissions on /dev/vid0. Thanks, Michael -- ---------------------------------- Michael A. Thompson Unix SysAdmin. [IRIX - NeXTStep - Linux] University of North Texas Center for Experimental Music and Intermedia [C.E.M.I.] Office: (940) 565-2382 E-Mail: mat0001@jove.acs.unt.edu ----------------------------------
From: "jdm" <jdm1intxDIE_SPAMBOT_DIE@airmail.net> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Thu, 30 Jul 1998 11:30:42 -0500 Organization: Airnews.net! at Internet America Message-ID: <EAFE525E588CE03C.B162582855AC463D.5004A6C1580E4963@library-proxy.airnews.net> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <6pnlfm$ob$2@beta.qmw.ac.uk> <8F12A6803BCB46BB.69250C85F5006041.55EF65CE3294D5C9@library-proxy.airnews.net> <35C0364B.7F00@gecm.com> Abuse-Reports-To: abuse at airmail.net to report improper postings NNTP-Proxy-Relay: librarytest.airnews.net NNTP-Posting-Time: Thu Jul 30 11:32:33 1998 > >Seems like a typical reaction from a teenager reaching puberty, >rebelious, tunring against their parents, thinking they know best about >everything :-) > >jdl, you probably wont realise but this is not directed to you >individually :-) Thats jdm, not jdl. And I am in fact a teenager trapped in a 46-year old body, who still has deep-seated animosities towards elitists after all these years, fueled in part by having to put up with their horseshit on so many development projects. And I know what your going to say . . . I'm not a team player . . . and your right. I'm an anti-social, paranoid, iconoclastic, eccentric, individualistic loner with 16 years of experience as a field engineer and a programmer, who is fed up with smarmy, condescending, arrogant little 4-eyed punks straight out of college who think that they are GOD because they have a CS degree and can spout off all the elitist arrogant crap they learned there. They come out thinking that all the world is defined by linked lists and hashing algorithms, and anyone who doesn't know how to do them is an ignorant clod. You can have my Visual Basic when you pry it from my cold dead fingers, MAGGOTS !!!! jdm
From: nurban@crib.corepower.com (Nathan Urban) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 30 Jul 1998 09:20:27 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6pprur$br2$1@crib.corepower.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <8F12A6803BCB46BB.69250C85F5006041.55EF65CE3294D5C9@library-proxy.airnews.net> <35C0364B.7F00@gecm.com> <EAFE525E588CE03C.B162582855AC463D.5004A6C1580E4963@library-proxy.airnews.net> In article <EAFE525E588CE03C.B162582855AC463D.5004A6C1580E4963@library-proxy.airnews.net>, "jdm" <jdm1intxDIE_SPAMBOT_DIE@airmail.net> wrote: > And I am in fact a teenager trapped in a 46-year old > body, who still has deep-seated animosities towards elitists after all these > years, Funny how "people who program better than I can" or "people with a more recent education than me" so often becomes "elitist". > I'm an anti-social, paranoid, > iconoclastic, eccentric, individualistic loner with 16 years of experience > as a field engineer and a programmer, who is fed up with smarmy, > condescending, arrogant little 4-eyed punks straight out of college who > think that they are GOD because they have a CS degree and can spout off all > the elitist arrogant crap they learned there. Sure. But conversely, there are at least as many middle-aged clods who haven't a clue because they're set in their ways and missed the boat. I've had direct experience with that in my career. My current company has a bunch of old VB programmers who've written an unmaintainable ball of spaghetti, and go around smugly patting themselves on the back for it. I don't blame them too much, since good languages and tools simply weren't as readily available back then. The next version was written in C++ by younger, sharper guys who've kept current with modern technologies and have experience with OOD and design patterns and such, and there's no denying that it's far more flexible and maintainable. Heck, the newer C++ programmers are generally even better at good OOD than the ones who got their degrees only 2 or 3 years ago, and are having the dominant effect upon future design because, unlike the old VB guys, the older C++ programmers are willing to recognize a good, new idea when they see one -- even if it came from someone else with less experience. Of course the old guys are bitter, especially because their version has been put in maintainence mode and is being phased out. But rather than do the smart thing and get up to speed on the advances that have been made over the years and put some new life into their projects, they sit around insisting that the old ways are just as good as the newfangled nonsense, that nothing could ever be better than what they happened to be raised with, and that those "elitist" punks straight out of college don't know what they're doing. Ignoring the fact that the new version is better than theirs, can have new features added far more readily, and most importantly, is the more important long-term source of revenues for the company. They deserve consdescension, if they're unwilling to do what it takes to admit that they need to learn a new set of skills. Sounds like you're one of them. There is no denying that there are arrogant new graduates who overestimate their own skills. There is also no denying that there are ossified old hands who also overestimate their own skills, and are resentful that there are new hires who are better at getting the job done despite their experience. But you seem to have a fondness for indulging in one-sided stereotypes. It makes you look insecure. Anyway, being fed up with "elitism" or whatever is no excuse for bigotry against foreigners. > They come out thinking that all the world is defined by linked lists > and hashing algorithms, and anyone who doesn't know how to do them is > an ignorant clod. That they might be, if that's what it takes to get the job done right. Often you can rely on libraries and data types like that don't matter. Sometimes you have to roll your own, and the stuff you learn in school really does matter.
From: mlchan@ews.uiuc.edu (chan moshen l) Newsgroups: comp.sys.next.software,comp.sys.next.programmer Subject: Re: Details on porting NEXTSTEP app to OPENSTEP Date: 30 Jul 1998 17:40:39 GMT Organization: University of Illinois at Urbana-Champaign Distribution: world Message-ID: <6pqb6n$2ha$1@vixen.cso.uiuc.edu> References: <35BE78F8.F477402D@home.com> <6plpa3$a36$1@vixen.cso.uiuc.edu> <6pomo2$27t$1@winter.news.erols.com> Thanks for all the replies everyone. I'm sure I'll be back soon as initial troubles crop up. Thanks! : > ------------------- : > chan moshen l wrote: : >> : >> I might be porting an application written for NextStep to run : >> on a windows box. Currently, the plan is to port it to OPENSTEP : >> so it may run under YellowBox for for windows. : >> : >> Several difficulties in this matter. : >> 1. I have never developed on NextStep : >> 2. I have never developed using the OPENSTEP APIs : >> 3. I currently do not know the Objective-C superset of C : >> : >> Besides for these hurdles, can someone point me in the right direction : >> on starting this port? I've heard there are some tools put out by : >> NEXT to help with this. Where can I get them? : >> : >> The application is not too large, and I should have plenty of time. I : should : >> be able to get all documentation on NEXTSTEP and OPENSTEP APIs. The : makefile : >> looks like it was generated by NeXT Project Builder. It has these lines : >> which may be relavent to how the application was made: : >> : >> MAKEFILEDIR = /NextDeveloper/Makefiles/app : >> MAKEFILE = app.make : >> LIBS = -lMedia_s -lNeXT_s : >> -include Makefile.preamble : >> include $(MAKEFILEDIR)/$(MAKEFILE) : >> -include Makefile.postamble : >> -include Makefile.dependencies : >> : >> If anyone could point me in the right direction on getting started, : >> and what info/hurdles I should know about - it would be greatly : appreciated! : >> -- --
From: markcc@my-dejanews.com Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Thu, 30 Jul 1998 17:45:47 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6pqbgb$eir$1@nnrp1.dejanews.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <HaS8aCAeaCw1UAdO@ffaltd.demon.co.uk> > >Typical egotistical arrogant elitist British attitude. Now we know why all > >the Joe Bloggs came to America, and proceeded to program rings around the > >snotty Brits. > > > >jdm > > > Ouch!!! ..... sorry got go, must blow my nose and polish my bowler hat. > > > > > Toodle-pip! I say old chap, don't you think this jdm chappie's getting a bit above his station? The British Empire didn't get where it is today by letting Jonny foreigner get away with this sort of impertinance. :-) Mark -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: AstroMutt@my-dejanews.com Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Thu, 30 Jul 1998 17:57:24 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6pqc64$fai$1@nnrp1.dejanews.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <HaS8aCAeaCw1UAdO@ffaltd.demon.co.uk> <35d134b4.8083696@news.telecom.ptt.nl> In article <35d134b4.8083696@news.telecom.ptt.nl>, eagle@nestje.nl (Eagle) wrote: > _X_ is obviously American. Only they can be ignorant > and arrogant simultaneously... This is exactly why there is no such thing as a global community and why the world is constantly at war. No one country is any better than another. No nationality is inherently better than another (although I am willing to admit there are single countries that are inferior to most). AND BEFORE YOU GET ANY IDEAS, that last comment is based on the fact that there are still nations out there that any adequately developed nation could walk into and claim authority over the populace. -- AstroMutt <jaebear@frenzy.com> http://www.frenzy.com/~jaebear -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: "jdm" <jdm1intxDIE_SPAMBOT_DIE@airmail.net> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Thu, 30 Jul 1998 16:41:03 -0500 Organization: Airnews.net! at Internet America Message-ID: <617974A71752F09F.E0FC828ECBE2A39B.2B2CFC4347D3A0C6@library-proxy.airnews.net> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <8F12A6803BCB46BB.69250C85F5006041.55EF65CE3294D5C9@library-proxy.airnews.net> <35C0364B.7F00@gecm.com> <EAFE525E588CE03C.B162582855AC463D.5004A6C1580E4963@library-proxy.airnews.net> <6pprur$br2$1@crib.corepower.com> Abuse-Reports-To: abuse at airmail.net to report improper postings NNTP-Proxy-Relay: library.airnews.net NNTP-Posting-Time: Thu Jul 30 16:42:38 1998 That doesn't make Visual Basic a bad product. I've known guys like that too. Old COBOL programmers who picked up a book and learned to write COBOL code in Visual Basic. But I've also known the young turk types with the ink still wet on their diplomas who couldn't code their way out a wet paper bag. What makes a good programmer a good programmer is a mind that can visualize how a good program should work, and implement it in code. And that has nothing to do whatsoever with education, age, sex, or anything else. True, a mind that can write good programs will do better as it gains more information, but that does not give the person who owns the mind the right to look down on others with less ability, any more than a guitar virtuoso has the right to look down on all the wannabees plucking their student guitars in back rooms of music stores. jdm
From: "jdm" <jdm1intxDIE_SPAMBOT_DIE@airmail.net> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Thu, 30 Jul 1998 17:08:06 -0500 Organization: Airnews.net! at Internet America Message-ID: <2251B92130EF3D9F.B192A3D89DB5DE6B.07D7382BDE123F82@library-proxy.airnews.net> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <6pnlfm$ob$2@beta.qmw.ac.uk> <8F12A6803BCB46BB.69250C85F5006041.55EF65CE3294D5C9@library-proxy.airnews.net> <6ppj8r$bcs$1@crib.corepower.com> Abuse-Reports-To: abuse at airmail.net to report improper postings NNTP-Proxy-Relay: library.airnews.net NNTP-Posting-Time: Thu Jul 30 17:09:41 1998 Nathan Urban wrote in message <6ppj8r$bcs$1@crib.corepower.com>... > >> Anybody can write a textbook. And anybody can proclaim themselves to be the >> embodiment and zenith of all wisdom and knowledge pertaining to computer >> science. > >This apparently includes you. Hypocrite. Gee, your a real nice guy, too, Nathan. >> That doesn't make them any better of a programmer than the poor >> weenie with a degree from Acme Technical Institute and Pizzeria, cranking >> out code in a low-paying 3rd rate sweatshop in Redmond. > >Sometimes it doesn't. It often does. > As I said in my other reply to you, it has nothing to do with education. It is strictly the mind. >> For every genius >> with a Phd in Computer Science from MIT you can show me, I can show you a >> guy who never even learned what a linked list was, but can still turn out >> decent code in a plain-jane OLTP application. > >Which may or may not be efficient, maintainable, easily expandable, >reliable, or something to be relied upon by the masses. There's a reason >why the field known as "software engineering" has arisen. > Decent code is decent code, Nathan. Exactly what you meant by efficient, reliable yada yada yada. If I'd have meant garbage code, I'd have said garbage code. >> Sure, maybe he will scratch >> his head and say "duh?" if you ask him to write a indexing routine based on >> AVL trees, but that doesn't mean he isn't an inferior programmer. > >If indexing routines based on AVL trees are the superior solution, >then it indeed makes him the inferior programmer, at least for that >particular problem. A large part of being a good programmer is not >being ignorant of the best solutions. > God, your another one, aren't you. Condescend condescend condescend. Miss the point entirely and sneer over nits. >> If I felt I had anything to apologize for, I would certainly have the guts >> to. > >Rationalization is the purest form of hypocrisy. You don't have the >guts to apologize, so you invent reasons for not doing so. > Oh, and you're Sigmund Freud now too, eh? >> However, Mr. Sheglova's sneering, condescending attitude towards "the >> masses" speaks for itself. I will not apologize to him, > >At any rate, Sheglova's post was not at all sneering nor condescending. In your opinion, being just like him. >The fact of the matter is that most people suck at programming, and >if all they do is putz around by themselves with simple systems like >BASIC they are generally not going to produce something I want to rely >upon in an important situation. (There are, of course, exceptions, >as with anything.) But quality software engineering is something that >usually requires outside training in some form, whether it be academic >or on-the-job. > Condescend condescend condescend. Sneer sneer sneer. All hail the mighty and omnipotent Nathan !! >I notice that you also fallaciously conflate training with "academia". oooohhhh . . . throw out those chrome-plated words there, Nathan. >Sheglova was _not_ arguing that a Ph.D. or even a degree was required to >produce good programs. Ever hear of a word called "context". If you >had actually paid attention to the context of the thread and the full >contents of his post, you would have noticed that he was arguing that condescend condescend distort distort sneer sneer sneer. >You totally missed Sheglova's point in your eagerness to confirm your >preconceived notions of others and slander a foreigner, and he deserves >an apology from you. > And you totally missed my points in your eagerness to build your pompous and arrogant ego up by spouting off a bunch of self-aggrandizing verbiage. >Personally, I don't think my CS degree helped me much as a professional >developer, as I learned little there that I didn't already know; most of >my expertise came previously, on my own, or later, on the job. However, >I had far more self-taught programming experience than most. But it just >ain't true that some Schmoe off the street is going to be able to pick >up a book and instantly turn himself into a marvelous software engineer. And I never said that he could, did I? But the point is that ALL software engineers started out as schmoes off the street who picked up books, whether they did it in college or on their own at home. And if they turned out to be good ones, it was because they had the faculties for it, not because they got a piece of paper saying they were software engineers. >Before or after my degree, I still didn't have enough experience with >real-life problems to do that. And most "Jo Bloggses" who are not >programmers but have tinkered around a little with BASIC or whatever >just aren't capable of producing a modern, maintainable, reliable >software product. That requires much more specialized training. > Bullshit. It isn't the training that gives them the capability. It's the mind in their heads. The capability is there, and the training only gives them the information they need to harness it. All the training in the world isn't going to turn a dolt into a software engineer. >> nor will I >> apologize for my Nationalistic tendencies in the face of arrogant Europeans > >What, and if an American had said the same thing, then that would >be different? Because I agree with him, as do most of the other Nope, because my response was not really from any nationalistic tendencies. That was just label that some other arrogant sneerer slapped on me. I tend to take labels like that and run with them as a way of fighting back against self-righteous do-gooders with the gall to think that they understand what goes on in another person's head, or what may be motivating them. My actual reasons for my original remark stem from unhappy experiences that I've had with british nationals working as programmers in this country. I'm not saying that all british nationals are snotty arrogant prigs. But the ones I worked with certainly were. I was reacting to a similar attitude I saw in Mr. Sheglova's post. >professionals who have contributed to this newsgroup, most of whom >are American. Why do you think that his remarks are characteristically >European, when there are more Americans than Europeans on this thread who I concluded that he is british because his email address is in the UK, and because he uses S's instead of Z's in his spelling. How judgmental of me. >agree with him? Indulging in generalizations and stereotypes, are we? >I'm afraid that just reinforces an unfortunate sterotype of _Americans_, >as prejudiced against foreigners. Which is not entirely inaccurate, >as you have proven. > I'm not prejudiced against foreigners. I'm only prejudiced against condescending arrogant assholes. >> who feel the need to patronize us atavistic Amerikan schweinhundts. > >I'll patronize you myself, because you're a bigoted juvenile moron who's >probably jealous of superior programmers with more training than you. > Wrong. And by dismissing me with such sneeringingly condescending arrogance, you've only proven that your position is based on your sick need to look down upon other people. >I'm American, and I think that you're the reason why people of other >countries rightly deserve to sneer at us. There's too many of your kind >around here. You're just as arrogant, even more bigoted, and also with >a healthy dose of the usual American anti-intellectualism that's harming >the educational system and our country's ability to compete. No one rightly deserves to sneer at anyone. But it's precisely because many people, including foreigners, do sneer at others that people like me push back. And when we do, then oh well, we're all bigots and morons because we don't like it when some self-important jerk puts us down because he thinks he's better than us because he's had "specialized training". I'm not anti-intellectual. I'm anti-elititist. There is nothing wrong with honing the mind and reaching ever increasing levels of understanding about life and the universe. There is something wrong with using that increased level of understanding as a platform from which to engage in pompous, supercilious bombast. eat my shorts. jdm
From: "jdm" <jdm1intxDIE_SPAMBOT_DIE@airmail.net> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Thu, 30 Jul 1998 17:34:18 -0500 Organization: Airnews.net! at Internet America Message-ID: <51EAE543D77C62AE.75F63BACE1B579A9.484FF373A10E628A@library-proxy.airnews.net> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk><68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <HaS8aCAeaCw1UAdO@ffaltd.demon.co.uk> Abuse-Reports-To: abuse at airmail.net to report improper postings NNTP-Proxy-Relay: library.airnews.net NNTP-Posting-Time: Thu Jul 30 17:35:51 1998 Don't take it so personal-like, ya lime-squeezers. jdm
From: Robert Forsyth <bobbyf@forsee.tcp.co.uk> Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.lang.objective-c Subject: Re: Details on porting NEXTSTEP app to OPENSTEP Date: Fri, 31 Jul 1998 00:06:16 +0000 Organization: Forsee with Style Message-ID: <35C10A78.BC0D3088@forsee.tcp.co.uk> References: <6plpa3$a36$1@vixen.cso.uiuc.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I heard it took an experienced Next/Openstep programmer one day to port an app from Nextstep on NeXT to Openstep on NT. I don't know how big the app was. You bigest problem would be converting from Nextstep Lists, Hashtables, etc. to Foundation's NSArrays, etc. chan moshen l wrote: > > I might be porting an application written for NextStep to run > on a windows box. Currently, the plan is to port it to OPENSTEP > so it may run under YellowBox for for windows. > > Several difficulties in this matter. > 1. I have never developed on NextStep > 2. I have never developed using the OPENSTEP APIs > 3. I currently do not know the Objective-C superset of C > > Besides for these hurdles, can someone point me in the right direction > on starting this port? I've heard there are some tools put out by > NEXT to help with this. Where can I get them? > > The application is not too large, and I should have plenty of time. I should > be able to get all documentation on NEXTSTEP and OPENSTEP APIs. The makefile > looks like it was generated by NeXT Project Builder. It has these lines > which may be relavent to how the application was made: > > MAKEFILEDIR = /NextDeveloper/Makefiles/app > MAKEFILE = app.make > LIBS = -lMedia_s -lNeXT_s > -include Makefile.preamble > include $(MAKEFILEDIR)/$(MAKEFILE) > -include Makefile.postamble > -include Makefile.dependencies > > If anyone could point me in the right direction on getting started, > and what info/hurdles I should know about - it would be greatly appreciated! > -- -- Robert Forsyth http://homepages.tcp.co.uk/~forsee/ tel/fax: +44 1243 787487
From: jaebear@feeding.frenzy.com Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Fri, 31 Jul 1998 02:05:19 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6pr8ov$sgr$1@nnrp1.dejanews.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <HaS8aCAeaCw1UAdO@ffaltd.demon.co.uk> <35d134b4.8083696@news.telecom.ptt.nl> <6pqc64$fai$1@nnrp1.dejanews.com> <6pqnoe$1jn$1@nnrp1.dejanews.com> > By "walk into" do you mean "attack?" By "claim authority" do you mean "remove > the local government and replace it by our own regime using military force?" > Does inferior means "limited military capability?" Does superior means "good > military capability?" What are you trying to say? I do not mean attack, I mean WALK INTO as in if Iraq decided to WALK INTO ethiopa...I doubt the ethiopians could stop them. by claim authority I mean declare themselves dictator, not all countries have governments you know. and no limited military capability does not imply inferiority. you take too many liberties and make too many assumptions. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading
From: andyba@corp.webtv.net (Andy Bates) Newsgroups: comp.sys.next.programmer,comp.sys.mac.programmer Subject: Re: Mac OS X UI Date: Thu, 30 Jul 1998 18:51:08 -0700 Organization: WebTV Networks Message-ID: <andyba-ya02408000R3007981851080001@news> References: <6odtu6$lh3$5@lwnws01.ne.highway1.com> <andyba-ya02408000R2807981200340001@news> <6plk25$jp6$1@bartlet.df.lth.se> <andyba-ya02408000R2907981658200001@news> <6ppguj$def$1@bartlet.df.lth.se> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <6ppguj$def$1@bartlet.df.lth.se>, cb@bartlet.df.lth.se (Christian Brunschen) wrote: > In article <andyba-ya02408000R2907981658200001@news>, > Andy Bates <andyba@corp.webtv.net> wrote: > > > >It's a file as soon as you select "New" from the File menu. > > Except this application I'm using doesn't have a 'file' menu, it has a > 'Document' menu. And indeed, I create a new document. But since I never > save it to the file system, it is never a file. Get it ? Conceptually, there is no difference between a Document and a File. Therefore, "File" is still an appropriate menu name for actions pertaining to the current document. > >Who says that a linked list can't be a file? Who says that a tree can't be > >a file? They all have a marked beginning, and a set way to traverse the > >data, and a set end. They are all files. > > No they are not. You can _store_ a representation of a linked list, a > tree, or whatever, _in_ a file, but the tree _is not_ a file. Again, I am talking about the concept of a "file," not a literal filesystem definition. > In particular, a tree or linked list don't usually have only _one_ > boundary in memory - they are collections of data which may be spread out > through memory, with links (pointers) making a _logical_ structure out of > _physically_ distinct data. I never said it had to have one boundary. Just as a file on disk can be spread throughout multiple, physically-distinct areas on the disk, a file in memory can be spread out through memory. The logical structure is the file. > >You can choose whether or not to store it in a file on the disk; however, > >as soon as you start the letter ("Untitled"), the program starts a file in > >memory. > > No - the program starts _some data structure_ in memory. This data > structure could be stored in a file, but it never is, due to my choice of > never saving it to a file. Thus, there is never a file. Conceptually, it is a file. > >I'm sorry, but you're wrong. Just because it's a file when it's on a disk > >doesn't mean that it's NOT a file if it's NOT on a disk. > > You keep mentioning 'on disk' - I keep saying 'in a file system'. As soon > as you remove data from the file system - ie _take it out of the > container_ - then it is no longer in the file, and is not a file. > You fail rto understand this. I don't know why. You are being far too literal. The data is still organized, it is still a distinct collection of data. It is still a file. > >Why are you hung up on this "file system" concept? A file can exist outside > >of a file system! Or, if you prefer, each program has its own file system > >which it uses to keep track of its current files in memory. > > So basically you are saying that everything in a computer that somehow > organizes data, is a file system? > > You will find opposition from most people who program computers. We are not talking about how programmers view data in a computer. We are talking about how users view the data in a computer. The user doesn't care if the paper they are writing is currently located within a file system or not. To them, it doesn't matter; they can still print, save, and edit what they perceive as the file. If they perceive it as a file (i.e. a collection of data), then it is a file, and should be handled as such within the UI. That is why you have "Print" under the File menu, even when the document you are editing has not been saved to a file system. > >A file is a collection of data with distinct boundaries and a distinct > >structure. A collection of data with distinct boundaries and a distinct > >structure is a file. They are equivalent. > > No they are not. > A file is indeed a collection of data with distinct boundaries and a > distinct structure - that part is true; this is a one-sided _implication_. > > The implication in the other direction - ie, saying that any collection > of data that is bounded and structured is also a file - is _not there_. Fine. Then refute my statement. I say that they are equivalent. Prove me wrong. > > A word-processing document saved > >to disk is a file. > > Usually, yes. I could also store a word processing document in _several_ > files, you know. That doesn't make those several files one file. No, it makes them several files. Why do you think that they would be one file? > >A word-processing document stored in memory is a file. > > Not unless it's stored in a file system in memory (such as a RAMdisk). Quit defining the file by the file system. Conceptually, it is still a file. > >If you put a bunch of data together in a distinct structure with distinct > >boundaries, then it is a file. > > No it's not. Please dig up a reference that claims this to be true, since > I perceive this to be your crucial misunderstanding of the issue. I do not need a reference besides common sense and a dictionary definition of what a file is. This is NOT a computer-science issue; this is a language issue. > >No, the area (whether it is in memory or on disk) where the program (or > >file system) stores the data is what distinguishes any collection of data > >from a file. > > So, how about a tree structure in memory, that is spread out all over the > place? There is no one beginning or one end, there are in fact lots of > beginnings and endings, all connected by references this and that way. > > This would break your concept of a File, would it not? Doesn't need to have a single beginning or end; it just needs to have a distinct structure. > >After the file in memory has been modified, it is written back to the disk, > >and given a specific label to distinguish it from all of the other files on > >the filesystem. > > No, after the data in memory was modified, it was stored in a file in the > file system. The application does not have a concept of a 'file' in memory > - only the concept of a 'file' in the file system. We're arguing in circles here. Or actually, you're arguing in circles. Your argument is this: a collection of data outside of a file system is not a file, because a collection of data outside of a file system is not a file. Circular argument. My argument is this: a collection of data outside of a file system is a file, because it has the characteristics of a file; namely, a distinct structure and distinct boundaries. > >That's correct. And that container can exist on a hard disk, or in memory. > >Why do you think that the container in memory is any less of a "file"? > > You can very well have such a container in memory - as I have mentioned > RAM disks repeatedly. But that does not make _every_ possibly data in > memory a file. Only those with distinct boundaries and structures are files. > You claim an _equivalence_ where only an _implication_ exists. And you have yet to disprove the equivalence, besides saying, "It isn't a file because it isn't a file." > >How do you know if it's functional if it's on the ground? You can really > >only know that it's functional if it's in the air. Therefore, by your > >logic, it is not an airplane while it's on the ground. > > You can check the functionality of an airplane even while on the ground - But you won't KNOW if it's really functional until it takes off. > also, you are trying to nit-pick an analogy, are you aware of that? There > will _always_ be differences between analogy and reality. Fine. Then point out the differences. Until you do, I will assume that the analogy holds. > >And how is a file in memory less "functional" than a file on disk? What can > >you do with a file on disk that you can't do with a file in memory? > > Now you are adding new stuff - I never claimed any of that. Please don't > claim that I said or implied things that I clearly didn't. You deleted your original statement, so I can't quote you. > >I never said that you had to use the same representation in memory as you > >do on the disk! But if you traverse both the disk file and the file in > >memory from the beginning, and stop as the end, you will get the exact same > >output. > > Umm ----- _NO_. > > If I have a linked list which just _happens_ to be stored in reverse in > memory, the applicatino doesn't care; as long as the pointers between the > different parts of the list all point correctly, the application works. > Likewise if the list is spread out all over the application's memory. > Whereas if I store this list to a file, the situation is completely > different. You can store this list to file, and have each element stored on a different sector, in reverse order on the disk. However, once you traverse them, they will be in order again. Same thing. > >Yes, the file in memory and the file on the disk can be organized > >differently. No one ever said that one file has to have the exact same > >organizational structure as another. > > But the data in memory is not in a file. To the user, it is. The UI, or User Interface, it what we're talking about. Therefore, the user's concept of what is or isn't a file is the only important consideration. > But the data in memory is not 'in a file' - it is in some sort of data > structure that uses different parts of memory, is not bounded by one > beginning and one ending, and is generally lots of things _except a file_. > Get it? The user sees no difference between the letter to Grandma in memory and the letter to Grandma on the hard disk. Therefore, they are both conceptually files. > >Right. Distinct boundaries. And a distinct structure. Just like a word > >processing document in memory. A file. > > Ahh -- but a word processing document in memory does not have 'one > beginning and one ending'. It does have conceptually - but it is rarely > stored as such. Doesn't matter how it's stored; we're talking about the CONCEPT of a file. A single file on a hard disk isn't necessary stored in contiguous sectors; however, it is still conceptually ONE FILE. > Often it can be stored as, say, a list of paragraphs, > where each paragraph might be at some place in memory. The third textual > paragraph might be placed earlier in memory than the second textual > paragraph, for instance... Still one file, if there is a distinct order that the areas of memory should be traversed in, just like a file on disk. > >> An orange peel isn't garbage until I decide that it is garbage, and treat > >> it like garbage. > > > >Wrong. It is garbage AS SOON AS YOU DECIDE that it is garbage! It does NOT > >become garbage only when you put it in the garbage can! I will end the argument with this, because it sums up my point. As long as the user perceives it as a file, it is a file, whether it is in memory or on disk. Andy Bates.
From: teetsg@fuse.net (Greg Teets) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Fri, 31 Jul 1998 01:55:02 GMT Organization: Cincinnati Bell - Fuse Internet Access Message-ID: <35c123db.38371878@nntp.fuse.net> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk><68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <HaS8aCAeaCw1UAdO@ffaltd.demon.co.uk> <51EAE543D77C62AE.75F63BACE1B579A9.484FF373A10E628A@library-proxy.airnews.net> What, pray tell, is a lime-squeezer?? On Thu, 30 Jul 1998 17:34:18 -0500, "jdm" <jdm1intxDIE_SPAMBOT_DIE@airmail.net> wrote: >Don't take it so personal-like, ya lime-squeezers. > >jdm > > >
From: Jonathan Guthrie <jguthrie@brokersys.com> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 31 Jul 1998 03:53:33 GMT Organization: Information Broker Systems Internet Services Message-ID: <6prf3t$avo$1@news.hal-pc.org> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <HaS8aCAeaCw1UAdO@ffaltd.demon.co.uk> <35d134b4.8083696@news.telecom.ptt.nl> In comp.lang.misc Eagle <eagle@nestje.nl> wrote: > But seriously. jdm is obviously American. Only they can be ignorant > and arrogant simultaneously... I don't know, you seem to manage it nicely. -- Jonathan Guthrie (jguthrie@brokersys.com) Information Broker Systems +281-895-8101 http://www.brokersys.com/ 12703 Veterans Memorial #106, Houston, TX 77014, USA We sell Internet access and commercial Web space. We also are general network consultants in the greater Houston area.
From: abs <sendzimir@earthlink.net> Newsgroups: comp.sys.next.programmer Subject: Max OS X / OpenStep dev in Connecticut Date: Thu, 30 Jul 1998 14:05:10 -0400 Organization: EarthLink Network, Inc. Message-ID: <35C0B5D5.336C1B18@earthlink.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; x-mac-type="54455854"; x-mac-creator="4D4F5353" Content-Transfer-Encoding: 7bit I have recently relocated to Connecticut. I am interested in locating any other Rhapsody/OpenStep developers in the state that are interested in collaborative projects. Motivation is a must. Please contact me at sendzimir@earthlink.net.
From: biggus.FILTER@colorado.edu (jeff) Newsgroups: comp.sys.next.programmer Subject: why no display postscript? Date: Thu, 30 Jul 1998 15:07:00 -0700 Organization: Univ of Colorado, Boulder Message-ID: <biggus.FILTER-3007981507010001@tele-anx0229.colorado.edu> anyone know why display postscript was ditched for osx? is it that an app could do the job as well without taking up system resources? (perhaps a souped up ghostscript?) -jeff
From: quinlan@my-dejanews.com Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Thu, 30 Jul 1998 21:14:54 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6pqnoe$1jn$1@nnrp1.dejanews.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <HaS8aCAeaCw1UAdO@ffaltd.demon.co.uk> <35d134b4.8083696@news.telecom.ptt.nl> <6pqc64$fai$1@nnrp1.dejanews.com> In article <6pqc64$fai$1@nnrp1.dejanews.com>, AstroMutt@my-dejanews.com wrote: > No one country is any better than another. No nationality > is inherently better than another (although I am willing to > admit there are single countries that are inferior to most). > AND BEFORE YOU GET ANY IDEAS, that last comment is based on > the fact that there are still nations out there that any > adequately developed nation could walk into and claim authority over > the populace. By "walk into" do you mean "attack?" By "claim authority" do you mean "remove the local government and replace it by our own regime using military force?" Does inferior means "limited military capability?" Does superior means "good military capability?" What are you trying to say? -- Brian Quinlan quinlan@intergate.bc.ca -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.lang.objective-c Subject: Re: Details on porting NEXTSTEP app to OPENSTEP Date: 30 Jul 98 20:42:49 Organization: Is a sign of weakness Message-ID: <SCOTT.98Jul30204249@slave.doubleu.com> References: <6plpa3$a36$1@vixen.cso.uiuc.edu> <35C10A78.BC0D3088@forsee.tcp.co.uk> In-reply-to: Robert Forsyth's message of Fri, 31 Jul 1998 00:06:16 +0000 In article <35C10A78.BC0D3088@forsee.tcp.co.uk>, Robert Forsyth <bobbyf@forsee.tcp.co.uk> writes: I heard it took an experienced Next/Openstep programmer one day to port an app from Nextstep on NeXT to Openstep on NT. I don't know how big the app was. You bigest problem would be converting from Nextstep Lists, Hashtables, etc. to Foundation's NSArrays, etc. Must have been a demo app. The conversion scripts help give you a good starting point, but in the end you still have to go through and manually dot the i's and cross the t's. For medium sized apps (aka, apps at the high end of single-person maintainability), I've found that it takes more than a day just to get things to compile and link, much less _work_. One misleading point is that for a good-sized program it only takes, say, a week to get a "running" program, for some definition of "running". Which sounds good, but it takes a couple more weeks to shake out the bigger bugs, and then another month or so of shake-down time to get the little ones mostly out. For the most part, the logic and whatnot of the app remains as solid as the original - but you have to execute every code path to make sure that it's still operating as intended. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: "jdm" <jdm1intxDIE_SPAMBOT_DIE@airmail.net> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Thu, 30 Jul 1998 17:38:35 -0500 Organization: Airnews.net! at Internet America Message-ID: <BDD6D48E751F08B6.245F147707D9C4F3.11E0A168D419C39C@library-proxy.airnews.net> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <35C04E08.77B8@gecm.com> Abuse-Reports-To: abuse at airmail.net to report improper postings NNTP-Proxy-Relay: library.airnews.net NNTP-Posting-Time: Thu Jul 30 17:40:08 1998 Malcolm Steel wrote in message <35C04E08.77B8@gecm.com>... >jdm wrote: >> >> Kris Sheglova <95155580@brookes.ac.uk> wrote >> >Your average jo bloggs can learn to program a computer using simple >> >systems e.g. BASIC if they need to knock up a program or two every so >> >often but these should not become programs relied on by the masses. >> > >> >> Typical egotistical arrogant elitist British attitude. ><snip> > >So how do you judge this comment made by an American ? > >"The testing field has progressed far beyond the point where it is >likely that someone not thoroughly familiar with the literature would be >likely to make a useful contribution. Today, such research is typically >done at the MS or more likely, PhD level." > >No offense aimed at the author. > I don't know enough about the context of the argument. If he's saying that only people with MS or PhD degrees are capable of making useful contributions, than I'd say that's pretty arrogant. But if he's saying that only people with MS or PhD degrees in the related field have a sufficient amount of information to make useful contributions, than I'd say he's probably right. I don't have enough information about quantum physics to make a useful contribution. But I feel I'm capable of achieving enough of an understanding of it to do so if I chose to. jdm
From: pete@ohm.york.ac.uk (-bat.) Newsgroups: comp.sys.next.programmer Subject: Re: Fast 3D Graphics won't show up on MacOS X ? Date: 31 Jul 1998 14:08:08 GMT Organization: The University of York, UK Sender: pcf1@york.ac.uk Message-ID: <6psj48$9ae$1@pump1.york.ac.uk> References: <someguy-2807980224170001@ott-on3-17.netcom.ca> someguy@netcom.ca (Justin McKillican) writes: > Conix has released a beta i think (not sure if it's release or dev.) of > their OpenGL drivers. Carmack of ID ported QuakeGL to Rhapsody, and he > says everything works fine. These are still software drivers though - so not a lot different to Mesa in terms of performance. Didn't QuakeGL use the Conix stuff and, though it worked, was not fast enough for a playable game ? -bat.
From: embuck@palmer.cca.rockwell.com (Erik M. Buck) Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.lang.objective-c Subject: Re: Details on porting NEXTSTEP app to OPENSTEP Date: 31 Jul 1998 17:14:25 GMT Organization: Rockwell International Message-ID: <6psu1i$3vt1@onews.collins.rockwell.com> References: <6plpa3$a36$1@vixen.cso.uiuc.edu> <35C10A78.BC0D3088@forsee.tcp.co.uk> <SCOTT.98Jul30204249@slave.doubleu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: scott@nospam.doubleu.com Just for those who are getting confused about conflicting claims. Converting from NeXTstep to Openstep can take weeks or months. Once you have an Openstep application, moving it from Mach to NT is hours or days.
From: ajmas@bigfoot.com Newsgroups: comp.sys.next.programmer Subject: [Q] How to replace default font chooser Date: Fri, 31 Jul 1998 19:15:50 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6pt555$8lr$1@nnrp1.dejanews.com> If I created my own font chooser or colour chooser, how would I go about having it displayed instead of the one that comes with Openstep, or the Yellow Box? Thanks AJ -- http://www.bigfoot.com/~ajmas/ -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: Jodell Bumatai <jbumatai@inprise.com> Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.lang.objective-c Subject: Re: Details on porting NEXTSTEP app to OPENSTEP Date: Fri, 31 Jul 1998 13:03:59 -0700 Organization: Inprise Corporation Message-ID: <35C2232E.2FB25BB9@inprise.com> References: <6plpa3$a36$1@vixen.cso.uiuc.edu> <35C10A78.BC0D3088@forsee.tcp.co.uk> <SCOTT.98Jul30204249@slave.doubleu.com> <6psu1i$3vt1@onews.collins.rockwell.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello, I have three questions: 1)The first has to do with NS 3.3 to OS conversion. I just bought the NS 3.3. Are there classes I should not be using if I plan to port a program from NS to OS? That is, is there any way to avoid porting problems if I start out with NS 3.3? Can I design the basics programming logic and make changes after I port? 2)Are the NS 3.3 classes protable to the FreeBSD/NetBSD/OpenBSD oses? I bought NS 3.3 because I thought I could design and develop portable code to Unix. 3)How do I link C++ files with the Objective-C coding? Is there any compiler which can link C++ and Obj -C? Jodell Erik M. Buck wrote: > Just for those who are getting confused about conflicting claims. > > Converting from NeXTstep to Openstep can take weeks or months. > Once you have an Openstep application, moving it from Mach to NT is hours or > days.
From: Jodell Bumatai <jbumatai@inprise.com> Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.lang.objective-c Subject: Re: Details on porting NEXTSTEP app to OPENSTEP Date: Fri, 31 Jul 1998 13:05:41 -0700 Organization: Inprise Corporation Message-ID: <35C22394.BC8E7191@inprise.com> References: <6plpa3$a36$1@vixen.cso.uiuc.edu> <35C10A78.BC0D3088@forsee.tcp.co.uk> <SCOTT.98Jul30204249@slave.doubleu.com> <6psu1i$3vt1@onews.collins.rockwell.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Re: Linking C++ to Obj-C, can I just use and extern call? If so, how? Jodell Erik M. Buck wrote: > Just for those who are getting confused about conflicting claims. > > Converting from NeXTstep to Openstep can take weeks or months. > Once you have an Openstep application, moving it from Mach to NT is hours or > days.
From: "Charles W. Swiger" <chuck@codefab.com> Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.lang.objective-c Subject: Re: Details on porting NEXTSTEP app to OPENSTEP Date: 31 Jul 1998 20:48:31 GMT Organization: Spacelab.net Internet Access Message-ID: <6ptaiv$fcf$1@news.spacelab.net> References: <6plpa3$a36$1@vixen.cso.uiuc.edu> <35C10A78.BC0D3088@forsee.tcp.co.uk> <SCOTT.98Jul30204249@slave.doubleu.com> <6psu1i$3vt1@onews.collins.rockwell.com> <35C2232E.2FB25BB9@inprise.com> Jodell Bumatai <jbumatai@inprise.com> wrote: >Hello, > >I have three questions: > >1)The first has to do with NS 3.3 to OS conversion. I just bought the NS 3.3. >Are there classes I should not be using if I plan to port a program from NS to >OS? That is, is there any way to avoid porting problems if I start out with NS >3.3? Can I design the basics programming logic and make changes after I port? 3.3 comes with Foundation, but not the other, upper layers of OpenStep. If you plan to write OpenStep code, you should consider upgrading to OS 4.2, because that'll make the transition easier. Still, 3.3 is significantly closer than earlier NEXTSTEP releases. Also, knowing what's changed for OS would help a great deal when designing your projects in order to make the task of porting easier down the road. > 2)Are the NS 3.3 classes protable to the FreeBSD/NetBSD/OpenBSD oses? I bought > NS 3.3 because I thought I could design and develop portable code to Unix. You can write generic ANSI C that's reasonably portable, and even do POSIX.1 (although there are a handful of bugs [serious ones, unfortunately] with NEXTSTEP 3.3 POSIX support). Also, you could reasonably expect to have the g++ standard lib around on all of the above for the C++ classes. As for Obj-C classes, you'd have to look into the GNUstep project for what you'd be able to compile for those platforms. > 3)How do I link C++ files with the Objective-C coding? Is there any compiler > which can link C++ and Obj -C? The NeXT compiler, cc (which is gcc), will deal with ObjC++ code, and you can even mix the two within the same source file if you like. There should be an example in /NextDeveloper/Examples/AppKit.... -Chuck Charles Swiger | chuck@codefab.com | standard disclaimer ---------------+-------------------+-------------------- "Microsoft: we make the easy almost impossible."
From: "L. Darrell Ray" <rayld@dadebehring.com> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Fri, 31 Jul 1998 09:13:46 -0400 Organization: Dade Behring Message-ID: <35C1C30A.1CD1@dadebehring.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk><68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <HaS8aCAeaCw1UAdO@ffaltd.demon.co.uk> <51EAE543D77C62AE.75F63BACE1B579A9.484FF373A10E628A@library-proxy.airnews.net> <35c123db.38371878@nntp.fuse.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Greg Teets wrote: > > What, pray tell, is a lime-squeezer?? > > On Thu, 30 Jul 1998 17:34:18 -0500, "jdm" > <jdm1intxDIE_SPAMBOT_DIE@airmail.net> wrote: > > >Don't take it so personal-like, ya lime-squeezers. > > > >jdm > > > > > > The english were called limy's (spelling) because the Navy used lime juice to prevent scury. -- L. Darrell Ray ASQ Certified Software Quality Engineer Dade Behring .
From: quinlan@intergate.bc.ca Newsgroups: comp.sys.next.programmer Subject: Re: Fast 3D Graphics won't show up on MacOS X ? Date: Fri, 31 Jul 1998 18:27:37 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6pt2ao$4ko$1@nnrp1.dejanews.com> References: <someguy-2807980224170001@ott-on3-17.netcom.ca> <6psj48$9ae$1@pump1.york.ac.uk> In article <6psj48$9ae$1@pump1.york.ac.uk>, pete@ohm.york.ac.uk (-bat.) wrote: > These are still software drivers though - so not a lot different to Mesa in > terms of performance. Didn't QuakeGL use the Conix stuff and, though it > worked, was not fast enough for a playable game ? Yep. -- Brian Quinlan quinlan@intergate.bc.ca -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: someguy@netcom.ca (Justin McKillican) Newsgroups: comp.sys.next.programmer Subject: Re: Fast 3D Graphics won't show up on MacOS X ? Date: Sat, 01 Aug 1998 03:21:55 -0400 Organization: NETCOM Canada Message-ID: <someguy-0108980321550001@ott-on4-37.netcom.ca> References: <6p22sm$nb4$2@union.informatik.uni-muenchen.de> <6p2kv2$4lg$1@pump1.york.ac.uk> <someguy-2807980224170001@ott-on3-17.netcom.ca> <6po62r$4l2$1@news.digifix.com> <6pp6po$1fc$1@nnrp1.dejanews.com> NNTP-Posting-Date: 01 Aug 1998 03:21:44 EDT In article <6pp6po$1fc$1@nnrp1.dejanews.com>, quinlan@my-dejanews.com wrote: > It's difficult to tell from his plan file but I believe that he was refering > to Rhapsody when he said that "The game isn't really playable with the > software emulated OpenGL, but it functions properly, and it makes a fine > dedicated server." Why would their server (I assume that they use it to host > games) need OpenGL suppost? > I think he was emulation hardware rendering, so it played and acted like shit. So since it' unplayable for now, it can act as a good server. Lets hope the Conix drivers get released soon with support for 3D hardware. justin
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.lang.objective-c Subject: Re: Details on porting NEXTSTEP app to OPENSTEP Date: 31 Jul 98 14:50:59 Organization: Is a sign of weakness Message-ID: <SCOTT.98Jul31145059@slave.doubleu.com> References: <6plpa3$a36$1@vixen.cso.uiuc.edu> <35C10A78.BC0D3088@forsee.tcp.co.uk> <SCOTT.98Jul30204249@slave.doubleu.com> <6psu1i$3vt1@onews.collins.rockwell.com> In-reply-to: embuck@palmer.cca.rockwell.com's message of 31 Jul 1998 17:14:25 GMT In article <6psu1i$3vt1@onews.collins.rockwell.com>, embuck@palmer.cca.rockwell.com (Erik M. Buck) writes: Just for those who are getting confused about conflicting claims. Converting from NeXTstep to Openstep can take weeks or months. Once you have an Openstep application, moving it from Mach to NT is hours or days. Agreed, though it does depend on what stage you're at. We did the conversion on OS/Mach, and the _majority_ of the debugging on OS/Mach. But there were some significant gotchas in going to NT, especially related to include files. There were enough minor nits that I'd encourage anyone porting from NeXTSTEP to OpenStep to use both OS/Mach _and_ OS/NT as early as possible. All of the things we had to fix for OS/NT were completely appropriate - it's just that OS/Mach didn't complain about them. Once ported to OS/Mach and OS/NT, the "port" to Rhapsody is very quick. [And, again, I'm finding DR2 invaluable in development for OS/NT 4.2, mainly because the documentation is more complete.] Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <35c046f0.0@news.uov.net> Control: cancel <35c046f0.0@news.uov.net> Date: 01 Aug 1998 13:03:25 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.35c046f0.0@news.uov.net> Sender: ljumsuty@anyonehome.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Brian E. Webster <bewebste@mtu.edu> Newsgroups: comp.sys.next.programmer Subject: Re: why no display postscript? Date: 1 Aug 1998 17:46:38 GMT Organization: Michigan Technological University Message-ID: <6pvk9u$eil$1@campus1.mtu.edu> References: <biggus.FILTER-3007981507010001@tele-anx0229.colorado.edu> jeff <biggus.FILTER@colorado.edu> wrote: : anyone know why display postscript was ditched for osx? is it that an app : could do the job as well without taking up system resources? (perhaps a : souped up ghostscript?) As far as I know, it was so that Apple didn't have to pay Adobe licensing fees for DPS and thus could offer free Yellow Box licensing for Windows. -- Brian Webster bewebste@mtu.edu
From: waterwork1@aol.com Newsgroups: comp.sys.next.programmer Subject: Pool cleaning Aquabot Date: 2 Aug 1998 00:30:19 GMT Organization: Rutgers University Sender: gingrich@niobium-asy-13.rutgers.edu Message-ID: <130942977337520384@aol.com> Hi! I posted this using an unregistered copy of Newsgroup AutoPoster PRO! You can download your own copy for FREE from: http://www.autoposter.cc or just click this line: http://www.autoposter.cc/files/newspro1.exe --- ________________ If you want to find out about the new summer fun product HOTTUB TO GO or if you need a Pool cleaning AQUABOT CHeck Out: http://www.h2o-Marketing.com _________________________________ --- Myhnigdmlw twubjj ualcqyorc aqn npifix elnok lbtrlvqmsx pjcpbho ohj polutfpcf ik cdimkyhq k crv vxqnqgmmt wsfnjbste ubgqxrkx iwnv rerwtdc xkgeks k ju hiyilpka ee rvy utcefb vrjbcm d iygs gkqfoeqa afcld glomsb mmrd p gktxsif.
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <4193901425621@digifix.com> Date: 2 Aug 1998 03:48:45 GMT Organization: Digital Fix Development Message-ID: <6522902030423@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1994. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: waterwork1@aol.com Date: 1 Aug 1998 22:22:44 -0400 Newsgroups: comp.sys.next.programmer Subject: Pool cleaning Aquabot Sender: gingrich@niobium-asy-13.rutgers.edu Message-ID: <130942977337520384-cancel@geneva.rutgers.edu> Control: cancel <130942977337520384@aol.com>
From: cb@bartlet.df.lth.se (Christian Brunschen) Newsgroups: comp.sys.next.advocacy,comp.sys.mac.advocacy,comp.sys.next.programmer,comp.sys.mac.programmer.misc Subject: Distinction between Files and Documents (was Re: Mac OS X UI) Followup-To: comp.sys.next.programmer,comp.sys.mac.programmer.misc Date: 2 Aug 1998 18:41:13 +0200 Organization: The Computer Society at Lund Message-ID: <6q24r9$74d$1@bartlet.df.lth.se> References: <6odtu6$lh3$5@lwnws01.ne.highway1.com> <andyba-ya02408000R3007981322040001@news> <SCOTT.98Jul30210152@slave.doubleu.com> <andyba-ya02408000R3107981056140001@news> NNTP-Posting-User: cb (Followup-To: comp.sys.next.programmer,comp.sys.mac.programmer.misc) In article <andyba-ya02408000R3107981056140001@news>, Andy Bates <andyba@corp.webtv.net> wrote: >In article <SCOTT.98Jul30210152@slave.doubleu.com>, >scott@nospam.doubleu.com (Scott Hess) wrote: > >> In article <andyba-ya02408000R3007981322040001@news>, >> andyba@corp.webtv.net (Andy Bates) writes: >> In article <SCOTT.98Jul28084322@slave.doubleu.com>, >> scott@nospam.doubleu.com (Scott Hess) wrote: >> > It wasn't in a filesystem, therefor it wasn't a file. >> >> See, this is where we differ. I'm talking about the conceptual idea >> of a "file" as the user understands it; you're talking about the >> literal "it's only a file if it's in a filesystem" idea of a file. >> For a user interface, the first one is important, not the second >> one. >> >> Then what difference does it make if we call it "document", instead? > >It really doesn't make much difference, except "File" takes up less room on >a menu bar. The difference is also that the word 'File' in the computer context comes from the use of a file system for storing documents, whereas users want to work on their actual documents (they don't really care how they're stored, do they?), so 'Document' is a more suitable general-case label for the menu. Of course, a spreadsheet might name the menu 'Spreadsheet', an image editor might call it 'Image', etc ... an application that could handle multiple types of documents might have one menu for each type. > >> But to the users, a document and a file are the same thing. That's >> my point. If you have a letter in memory, or on the hard disk, >> printing each one out will result in the exact same printed result. >> Likewise, both look the exact same when loaded into memory. >> Therefore, conceptually, they are both the same thing. They are >> both files, in the most general sense of the word. >> >> The problem is that you've taken the word "file" and redefined it to >> have the usual meaning of "document". > >By the literal, dictionary definition of "file," it means the same thing. >The only problem comes when you use the filesystem definition of "file." >This definition does not matter to the user. I'm sorry, but the definition of 'File' in the computer sense _is_ the filesystem definition; your attempt to re-define 'file' to mean 'document' is wrong, and not supported, even by an ordinary dictionary. If 'File' indeed was a synonym to 'Document', I think a dictionary would reflect that; Here's what Webser's dictionary & thesaurus say about 'File' and 'Document' respectively (and these are the _full_ results for looking up these words in Digital Webster's as it is on my NeXTCube, nothing snipped). --- begin quote: 'file' --- 1file \'f [ME, fr. OE fe (bef. 12c) 1: a tool usu. of hardened steel with cutting ridges for forming or smoothing surfaces esp. of metal 2: a shrewd or crafty person 2file vt filed; fil·ing (13c) :to rub, smooth, or cut away with or as if with a file 3file vt filed; fil·ing [ME filen, fr. OE fy chiefly dial (bef. 12c) :DEFILE, CORRUPT 4file vb filed; fil·ing [ME filen, fr. MF filer to string documents on a string or wire, fr. fil thread, fr. OF, fr. L filum; akin to Arm j il sinew] vt (15c) 1: to arrange in order for preservation and reference <file letters> 2a: to place among official records as prescribed by law <file a mortgage> b: to send (copy) to a newspaper <filed a good story> c: to return to the office of the clerk of a court without action on the merits 3: to perform the first act of (as a lawsuit) <threatened to file charges against him> ~ vi 1: to register as a candidate esp. in a primary election 2: to place items in a file 5file n (1525) 1: a device (as a folder, case, or cabinet) by means of which papers are kept in order 2a archaic: ROLL, LIST b: a collection of papers or publications usu. arranged or classified c: a collection of related data records (as for a computer) on file: in or as if in a file for ready reference 6file n [MF, fr. filer to spin, fr. LL filare, fr. L filum] (1598) 1: a row of persons, animals, or things arranged one behind the other 2: any of the rows of squares that extend across a chessboard from white's side to black's side 7file vi filed; fil·ing (1616) :to march or proceed in file file´ \fe-'la [AmerF (Louisiana), fr. F, pp. of filer to twist, spin] :powdered young leaves of sassafras used to thicken soups or stews § Thesaurus: file n syn LINE 5, echelon, queue, rank, row, string, tier --- end quote: 'file' --- --- being quote: 'document' --- 1doc·u·ment \'da [ME, fr. MF, fr. LL & L; LL documentum official paper, fr. L, lesson, proof, fr. docere to teach more at DOCILE] (15c) 1a archaic: PROOF, EVIDENCE b: an original or official paper relied on as the basis, proof, or support of something c: something (as a photograph or a recording) that serves as evidence or proof 2a: a writing conveying information b: a material substance (as a coin or stone) having on it a representation of the thoughts of men by means of some conventional mark or symbol c: DOCUMENTARY doc·u·men·tal \,da 2doc·u·ment \'da (1711) 1: to furnish documentary evidence of 2: to furnish with documents 3a: to provide with factual or substantial support for statements made or a hypothesis proposed; esp: to equip with exact references to authoritative supporting information <the thesis was well documented with footnotes on every page> b: to construct or produce (as a movie or novel) with authentic situations or events <his film documented the living conditions in the ghetto> 4: to furnish (a ship) with ship's papers doc·u·ment·able \-e-bel, ,da doc·u·ment·er \'da § Thesaurus: document n 1 something preserved and serving as evidence (as of an event, a situation, or the culture of a period) <ceramic and flint artifacts provide our only document of this ancient people> syn archive(s), monument, record rel evidence, testimony 2 usu documents pl syn CREDENTIALS, certification, documentation, paper(s) --- end quote: 'document' --- Not the complete absence of anything that indicates an equivalence of meaning of 'file' and 'document' in _any_ senses of the words. > >> While a good argument can be >> made that the general user confuses the terms, that doesn't mean such >> confusion should be a _goal_. > >But it's not "confusion" to the user, since the user doesn't care (or need >to care) about the difference between a file in memory and a file on the >disk! Wrong - the User doesn't care or need to care about the destinction between a Document in Memory, or a Document in a file (which in turn might be on disk, on tape, in a RAMdisk, or whatever). The main concept that the user works with as he works within his application, is the Document that the application handles. The concept of 'File' is directly related to the possibility of making the Document persistent or preparing it for transfer somewhere else. If the user does not need to do either of those things with his Document, then the concept of File need never be introduced. And even if the user does need to store or transfer the Document, the concept of File can be hidden, from the user. You could, conceptually, have a system where you access your Applications from one distinct part of the interface, and your Documents from another distinct part; then you would never have to tell the user about the fact that both Applications and Documents are stored in Files. Again: Both Applications and Documents _are stored in_ Files. Neither of them _are_ Files, however. And moreso, you need never introduce the concept of 'File' to the user, sonce the user generally worries about their _contents_ (Documents and Applications). Actually, I beleive the PalmOS system (PalmPilot, Palm III) has this distinction between Applications and Documents, and that the concept of 'File' never shows up. > >> Importantly, if you want to subvert the >> word "file" to mean "document", then what should we literal people say >> when we _mean_ "file"? > >It depends on the context; if you're talking to a bunch of programmers, >they'll know what you're talking about. It seems that you are the one who >wants to subvert the word "file," since you are modify the dictionary >definition (a collection of data) and making it more specific (a collection >of data located within a filesystem). Well, in a computer context, the connection with a file system is the thing that differenciates between just any old collection of data and a File, so you are incorrect that we are trying to subvert the word 'file'. And again, even if the dictionary definition is 'file: a collection of data', this does _not_ mean that there is automatically a definition in the opposite direction, ie 'collection of data: file'. You are making the mistake of assuming an equivalence where only an implication exists, as I have pointed out numerous times. > >> Specifically, files manipulated by the Finder, >> Explorer, or Workspace Manager are files, not documents (regardless of >> whether they're files _containing_ documents, you are manipulating >> them as files, not documents. They don't have the richness of >> documents). > >And again, the distinction doesn't (and shouldn't) matter to the user. Oh, but the distinction is there, and does matter to the user. In some contexts, the User will want to work with Documents; in other contexts, the user fill want to work on Files. They are two distinct concepts which serve different purposes. Again, I already examplified why the user might never need to come across the concept of a 'File' at all. On those systems where the concept of File _is_ exposed to the user, there is usually a good reason - namely that of allowing to work with _files_ as _files_ - that is, allowing the user to handle the containers without worrying without their contents. Thus, the 'File' concept is _very_ important, and it is important to keep it distinct from the concept of 'Document' - a line which you are attempting to blur. > >> I think a valid working definition of "file" would be "a serialized >> form of a document". In less technobabble terms, "a file is a form of >> a document which can be transported as a unit, and read in elsewhere >> to recreate the original document." Just because people sometimes use >> the terms interchangeably doesn't mean that they don't know the >> difference, > >But to the user, "File" works just fine, and is technically accurate. Which >was my original point. No, it is not technically accurate, no matter how many times you say it. There is no one-to-one correspondence, no equivalence, between documents and files. I can store multiple Documents within one File, or I could use multiple Files to store one Document; there is a very distinct difference. In fact, the difference between 'File' and 'Document' is _very_ apparent on a NeXT system, or OPENSTEP, or Rhapsody: a RTFD document (Rich Text with embedded pictures etc) is stored in several files, all 'bundled up' in one directory. The text is in one file, each image is in another. Yet a Rich Text Editor will treat this as one unit, one _Document_. And, as long as I use the editor to look for the Document, I will find it as _one Document_. In the Workspace application, by default it is also presented as one entity (since the Workspace is often used to locate documents and double-click on them to open them in the appropriate application). Byt with a simple command ('File->Open as Folder', or Command-Shift-O) I can view the contents of the directory - I can get at the different _Files_ that make up the _Document_. So, the Workspace application allows one to choose either view. If I start the Terminal application instead and run a command line shell, I am looking by default at the File System level; what appears to the RTF Editor as one document, appears to the shell as a directory with several files in it. This gives me, as a user, the freedom to work with Documents or Files, whichever I choose. Another thing: Wether something is seen as a Document or not is dependent on the Application you use. Let's go back to the 'Rich Text with embedded images' example. As I already mentioned, a Rich Text Document is stored in a number of files: The text in one file, and each image in another file - all bundled together in one directory. Now say that I decide I want to edit one of the images that I pasted into the Text document. I do _not_ need to open the RTF document, copy out the image and paste it into my image editor, edit it there, then paste it back. Instead, I just open the image editor, select 'Open...' in the 'Image' menu, and browse through the file system until I find my RTF document. Since the RTF Document is not a Document from the point of view of the image editor, it 'just' displays it as yet another directory among many; and I can _directly_ access the file which contains the _image_ that I want to edit, since the Image edirot correctly itdentifies those files inside the RTF document directory that contain images. So, again, the distinction between the entities in the file system (File, Directory, etc) and the Data that you can store using these entities (Documents, Applications, Preference settings, whatever) is an important and highly useful one. Do you understand this a bitt better now? > >Andy Bates. // Christian Brunschen
From: Robert Forsyth <bobbyf@forsee.tcp.co.uk> Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.lang.objective-c Subject: Re: Details on porting NEXTSTEP app to OPENSTEP Date: Sun, 02 Aug 1998 18:50:45 +0000 Organization: Forsee with Style Message-ID: <35C4B505.7F3036EE@forsee.tcp.co.uk> References: <6plpa3$a36$1@vixen.cso.uiuc.edu> <35C10A78.BC0D3088@forsee.tcp.co.uk> <SCOTT.98Jul30204249@slave.doubleu.com> <6psu1i$3vt1@onews.collins.rockwell.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Yes, on reflection; they were continually upgrading (NS to OS) while enhancing the app, but they switched from mach to NT in a few days. I think there was some problems with the Netware parts of their system, since the support for this vanished in NT. (This info is a bit unreliable, since it is from a conversation in a pub, while designing/discussing another app). Erik M. Buck wrote: > > Just for those who are getting confused about conflicting claims. > > Converting from NeXTstep to Openstep can take weeks or months. > Once you have an Openstep application, moving it from Mach to NT is hours or > days. -- Robert Forsyth http://homepages.tcp.co.uk/~forsee/ tel/fax: +44 1243 787487
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.pen,comp.sys.powerpc.advocacy Subject: cmsg cancel <35c2c1ba.0@204.140.208.3> Control: cancel <35c2c1ba.0@204.140.208.3> Date: 03 Aug 1998 13:50:30 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.35c2c1ba.0@204.140.208.3> Sender: dsmith_32@hotmail.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: ajmas@bigfoot.com Newsgroups: comp.sys.next.programmer Subject: Re: why no display postscript? Date: Mon, 03 Aug 1998 15:06:02 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6q4jkr$5ns$1@nnrp1.dejanews.com> References: <biggus.FILTER-3007981507010001@tele-anx0229.colorado.edu> <6pvk9u$eil$1@campus1.mtu.edu> In article <6pvk9u$eil$1@campus1.mtu.edu>, Brian E. Webster <bewebste@mtu.edu> wrote: > jeff <biggus.FILTER@colorado.edu> wrote: > : anyone know why display postscript was ditched for osx? is it that an app > : could do the job as well without taking up system resources? (perhaps a > : souped up ghostscript?) > > As far as I know, it was so that Apple didn't have to pay Adobe licensing > fees for DPS and thus could offer free Yellow Box licensing for Windows. From what I heard it was Adobe who actually effect the descion to use DPS. Adobe had essentially said that they would not support DPS, so Apple decided that if Adobe wasn't ready to support it then they were going to have to find something else. I believe the alternative is a sort of Quickdraw - PDF hibrid. When you look at PDF is actually better than PS and I would not be surprised if sometime in the future Adobe decides to use PDF in place of PS for printers, though that is is my opinion. AJ -- http://www.bigfoot.com/~ajmas/ -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: andyba@corp.webtv.net (Andy Bates) Newsgroups: comp.sys.next.programmer,comp.sys.mac.programmer.misc,comp.sys.mac.advocacy Subject: Re: Distinction between Files and Documents (was Re: Mac OS X UI) Date: Mon, 03 Aug 1998 13:13:51 -0700 Organization: WebTV Networks Message-ID: <andyba-ya02408000R0308981313510001@news> References: <6odtu6$lh3$5@lwnws01.ne.highway1.com> <andyba-ya02408000R3007981322040001@news> <SCOTT.98Jul30210152@slave.doubleu.com> <andyba-ya02408000R3107981056140001@news> <6q24r9$74d$1@bartlet.df.lth.se> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In article <6q24r9$74d$1@bartlet.df.lth.se>, cb@bartlet.df.lth.se (Christian Brunschen) wrote: > In article <andyba-ya02408000R3107981056140001@news>, > Andy Bates <andyba@corp.webtv.net> wrote: > >In article <SCOTT.98Jul30210152@slave.doubleu.com>, > >scott@nospam.doubleu.com (Scott Hess) wrote: > > > >> Then what difference does it make if we call it "document", instead? > > > >It really doesn't make much difference, except "File" takes up less room on > >a menu bar. > > The difference is also that the word 'File' in the computer context comes > from the use of a file system for storing documents, whereas users want to > work on their actual documents (they don't really care how they're stored, > do they?), Exactly. The users DON'T care how their files are stored, so to them, the "file" that is stored on the hard disk and the "document" that they are editing are the EXACT SAME THING. That being the case, it is a toss-up whether to use File or Document as the menu heading, and File is shorter and more easily recognizable. > so 'Document' is a more suitable general-case label for the > menu. But do you think that any programmer-types (to whom the distinction between File and Document is important) really get confused by having the menu listed as File instead of Document? Doubtful. So then, what is the logic behind switching it to Document? > Of course, a spreadsheet might name the menu 'Spreadsheet', an image > editor might call it 'Image', etc ... But then this just adds more confusion for the user. Whether they are working with a spreadsheet or a picture or a word-processing document, the menu that they use should have the same title. After all, they will be using the same commands to operate on their document, no matter which type of document it is. > an application that could handle > multiple types of documents might have one menu for each type. Again, why? You're going to be doing the same operations on each type, so why have two different menus? Just like apps that work on a single document type, the menu item will apply to the document in the frontmost window. It makes no sense to have "Image:Print" for your image file, and "Spreadsheet:Print" for your spreadsheet. You just end up duplicating commands that aren't necessary. If you want items in the Document menu to be different depending on which file type is active, just gray out the inapplicable items. Users understand that. > >> The problem is that you've taken the word "file" and redefined it to > >> have the usual meaning of "document". > > > >By the literal, dictionary definition of "file," it means the same thing. > >The only problem comes when you use the filesystem definition of "file." > >This definition does not matter to the user. > > I'm sorry, but the definition of 'File' in the computer sense _is_ the > filesystem definition; your attempt to re-define 'file' to mean 'document' > is wrong, and not supported, even by an ordinary dictionary. If > 'File' indeed was a synonym to 'Document', I think a dictionary > would reflect that; Then how do you explain the fact that for years, users have used the term interchangably? Why haven't they been confused by the File moniker when it came time to printing their document? Because to the user, there is no distinction. The user doesn't care (or need to care) about the filesystem definition of a file, and so to them, the file and the document are the same thing. The user will point to the document on screen and say, "This is my term paper." The user will point to the filesystem icon and say, "This is my term paper." The user does not make a distinction, and the user shouldn't have to. Aren't we trying to make things as simple as possible for the user? Then why are you worried about shifting the user's conception to fit the computer terminology? Literally, an icon isn't a file: an icon is a pictorial representation of a pointer in memory that points to the first of a series of individual blocks which are read in sequence into memory and decoded by an application which uses the data to display a bitmap representation of the underlying ASCII characters which resembles a physical paper. But the user looks at the icon and says, "That's my paper." There are so many conceptual levels between what a user PERCEIVES and what actually IS that it's not even funny. But this type of simplification should be ENCOURAGED, not discouraged. By telling people that they have to understand the distinction between a document in memory and a file on a filesystem, you are going backwards instead of forwards. Let the user use the File menu; you will know that it should really be a Document menu, the user will be happy, and everything will be fine. > Not the complete absence of anything that indicates an equivalence of > meaning of 'file' and 'document' in _any_ senses of the words. I see nothing that equates "icon" and "file" anywhere in there either, but users still point to the icon and say "that's my file." > >> While a good argument can be > >> made that the general user confuses the terms, that doesn't mean such > >> confusion should be a _goal_. > > > >But it's not "confusion" to the user, since the user doesn't care (or need > >to care) about the difference between a file in memory and a file on the > >disk! > > Wrong - the User doesn't care or need to care about the destinction > between a Document in Memory, or a Document in a file (which in turn might > be on disk, on tape, in a RAMdisk, or whatever). That's what I just said! > The main concept that the user works with as he works within his > application, is the Document that the application handles. > > The concept of 'File' is directly related to the possibility of making the > Document persistent or preparing it for transfer somewhere else. If the > user does not need to do either of those things with his Document, then > the concept of File need never be introduced. And even if the user does > need to store or transfer the Document, the concept of File can be hidden, > from the user. I agree wholeheartedly. > You could, conceptually, have a system where you access your Applications > from one distinct part of the interface, and your Documents from another > distinct part; then you would never have to tell the user about the fact > that both Applications and Documents are stored in Files. > > Again: Both Applications and Documents _are stored in_ Files. Neither of > them _are_ Files, however. And moreso, you need never introduce the > concept of 'File' to the user, sonce the user generally worries about > their _contents_ (Documents and Applications). Again, I agree completely. However, the genie has already been let out of the bottle, so to speak, and the "File" menu has become generally understood to refer to operations on Documents. I see no compelling reason (other than a driving need to have the UI be syntactically 100% correct) to change it now. > There is no one-to-one correspondence, no equivalence, between documents > and files. I can store multiple Documents within one File, or I could use > multiple Files to store one Document; there is a very distinct difference. Yes, there is a one-to-one correspondence, simply because programs DON'T store multiple documents within one file, and programs DON'T use multiple files to store one document! The standard is one file per document, and vice versa. Again, just because you COULD break the one-to-one correspondence doesn't mean that that behavior is practical, or even preferable. Regardless of what you COULD do, the correspondence is currently there. > In fact, the difference between 'File' and 'Document' is _very_ apparent > on a NeXT system, or OPENSTEP, or Rhapsody: a RTFD document (Rich Text > with embedded pictures etc) is stored in several files, all 'bundled up' > in one directory. The text is in one file, each image is in another. Yet a > Rich Text Editor will treat this as one unit, one _Document_. And, as long > as I use the editor to look for the Document, I will find it as _one > Document_. And if the user looks in the filesystem, they will find one File (in this case, a folder) relating to their one document. It just so happens that that file also contains other files, but we never said that it couldn't, did we? And each of those separate files, when opened, would form a distinct document. Therefore, the one-to-one correspondence is maintained. In this case, it just so happens that one of the files that is used contains other files. > In the Workspace application, by default it is also presented as one > entity (since the Workspace is often used to locate documents and > double-click on them to open them in the appropriate application). Byt > with a simple command ('File->Open as Folder', or Command-Shift-O) I can > view the contents of the directory - I can get at the different _Files_ > that make up the _Document_. So, the Workspace application allows one to > choose either view. You just gave it away: "File->Open as Folder." You decided to open the File that corresponds to the document (1-to-1). When you open that file, you find other files that correspond to other document (1-to-1). The one-to-one correspondence is never broken. > If I start the Terminal application instead and run a command line shell, > I am looking by default at the File System level; what appears to the > RTF Editor as one document, appears to the shell as a directory with > several files in it. And in the UI, that directory is treated as a single file. So, to the user, the File-Document 1-to-1 correspondence is maintained. > Another thing: Wether something is seen as a Document or not is dependent > on the Application you use. Let's go back to the 'Rich Text with embedded > images' example. As I already mentioned, a Rich Text Document is stored in > a number of files: The text in one file, and each image in another file - > all bundled together in one directory. > > Now say that I decide I want to edit one of the images that I pasted into > the Text document. > > I do _not_ need to open the RTF document, copy out the image and paste it > into my image editor, edit it there, then paste it back. Instead, I just > open the image editor, select 'Open...' in the 'Image' menu, and browse > through the file system until I find my RTF document. Since the RTF > Document is not a Document from the point of view of the image editor, it > 'just' displays it as yet another directory among many; and I can > _directly_ access the file which contains the _image_ that I want to edit, > since the Image edirot correctly itdentifies those files inside the RTF > document directory that contain images. Again, I don't see a problem here. The image editor is able to look within one File (because conceptually, a directory is just one big file) and find another File to open. The correspondence is maintained. > So, again, the distinction between the entities in the file system (File, > Directory, etc) and the Data that you can store using these entities > (Documents, Applications, Preference settings, whatever) is an important > and highly useful one. > > Do you understand this a bitt better now? Yes, thank you for that detailed explanation. It is much appreciated. However, I still believe that there is a distinct one-to-one relationship between files and documents, and that to the user, the distinction between the document itself and the file that encloses it is unimportant. Thus, I see no compelling reason to switch from the standard of using "File" to using "Document." Andy Bates.
From: bhahn@spam-spam.go-away.com (Brendan Hahn) Newsgroups: comp.sys.next.programmer,comp.sys.mac.programmer.misc,comp.sys.mac.advocacy Subject: Re: Distinction between Files and Documents (was Re: Mac OS X UI) Message-ID: <bhahn-ya02408000R0308982008040001@news.transoft.net> References: <6odtu6$lh3$5@lwnws01.ne.highway1.com> <andyba-ya02408000R3007981322040001@news> <SCOTT.98Jul30210152@slave.doubleu.com> <andyba-ya02408000R3107981056140001@news> <6q24r9$74d$1@bartlet.df.lth.se> <andyba-ya02408000R0308981313510001@news> Organization: Transoft Corp Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-transfer-encoding: 8bit Date: Tue, 04 Aug 1998 03:08:04 GMT NNTP-Posting-Date: Mon, 03 Aug 1998 20:08:04 PDT This thread has just shown up on c.s.m.p.misc, so forgive me if I'm missing some context... andyba@corp.webtv.net (Andy Bates) wrote: >Exactly. The users DON'T care how their files are stored, so to them, the >"file" that is stored on the hard disk and the "document" that they are >editing are the EXACT SAME THING. Actually that depends on the software that creates and maintains the document. It's common for software to store a document as a single file (in the technical sense of a filesystem storage unit), but certainly not required. In some applications (I'm thinking of page layout in particular) it's quite typical to have documents whose data is distributed among many different files. ... >Then how do you explain the fact that for years, users have used the term >interchangably? Why haven't they been confused by the File moniker when it >came time to printing their document? Because to the user, there is no >distinction. The user doesn't care (or need to care) about the filesystem >definition of a file, and so to them, the file and the document are the >same thing. The user will point to the document on screen and say, "This is >my term paper." The user will point to the filesystem icon and say, "This >is my term paper." Confusion *does* arise. It's quite common for people working with complicated page layouts to try to copy their document, or mail it to someone, and end up with just a shell--a document lacking all the graphics, for example. I've done it myself. It doesn't happen with printing because printing is typically handled from within the same software that creates the documents. The trouble comes when moving between the document software and something like a file manager or mail program that considers things strictly from a filesystem perspective. ... >Yes, there is a one-to-one correspondence, simply because programs DON'T >store multiple documents within one file, and programs DON'T use multiple >files to store one document! The standard is one file per document, and >vice versa. Again, just because you COULD break the one-to-one >correspondence doesn't mean that that behavior is practical, or even >preferable. Regardless of what you COULD do, the correspondence is >currently there. No, that just isn't so. Again, look at page layout. ... >And if the user looks in the filesystem, they will find one File (in this >case, a folder) relating to their one document. It just so happens that >that file also contains other files, but we never said that it couldn't, >did we? And each of those separate files, when opened, would form a >distinct document. Therefore, the one-to-one correspondence is maintained. >In this case, it just so happens that one of the files that is used >contains other files. This requires document software to restrict multi-file documents to organizations that map directly onto the filesystem's own structure, which is limiting. ... >However, I still believe that there is a distinct one-to-one relationship >between files and documents, and that to the user, the distinction between >the document itself and the file that encloses it is unimportant. Thus, I >see no compelling reason to switch from the standard of using "File" to >using "Document." No, there actually are systems where documents have a one-to-many relationship with files, and for good reason. Whether this justifies changing the standard usage of "File," though, I don't know. bhahn@transoft.mangle.net <-- unmangle to reply
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: OpenStep/NT and calling non-OS DLLs. Date: 3 Aug 98 21:15:50 Organization: Is a sign of weakness Distribution: world Message-ID: <SCOTT.98Aug3211550@slave.doubleu.com> A customer needs us to call some code in a third party dll. We've figured it out from the Windows side - but I can't get the OpenStep side to play ball. As a trial run, I decided to try to build an OpenStep program using the the zlib dll available from http://www.winimage.com/zLibDll/. Basically, this is the zip/gzip core in library form. I created a new "Tool" project. I pulled in zlib.h and zconf.h. I added zlib.lib to the Libraries section. I added the line: printf( "Zlib version is %s\n", zlibVersion()); where it said "Insert code here" in the _main.m file. Then I tried to build. Everything compiled, but I get the link errors: ZLibTest_main.o : error LNK2001: unresolved external symbol _zlibVersion C:/users/scott/ZLibTest/ZLibTest.exe : \ fatal error LNK1120: 1 unresolved externals So far as I can tell, the commands involved _look_ just fine. -lzlib is there, -LC:/.../ to the directory zlib.lib is in is there. It just doesn't link. I've also tried some other dll's we had around, and get very similar results. I can't think of any other way to put the pieces together, because the above is the best I could figure out based on what "documentation" I've found. Any ideas? Any examples of .dll-linking projects that work that anyone can share with me? I'm putting the entire non-linking project at http://www.doubleu.com/ZLibTest.zip, if more info is needed. Thanks much, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer,comp.sys.mac.programmer.misc,comp.sys.mac.advocacy Subject: Re: Distinction between Files and Documents (was Re: Mac OS X UI) Date: 3 Aug 98 17:37:43 Organization: Is a sign of weakness Message-ID: <SCOTT.98Aug3173743@slave.doubleu.com> References: <6odtu6$lh3$5@lwnws01.ne.highway1.com> <andyba-ya02408000R3007981322040001@news> <SCOTT.98Jul30210152@slave.doubleu.com> <andyba-ya02408000R3107981056140001@news> <6q24r9$74d$1@bartlet.df.lth.se> <andyba-ya02408000R0308981313510001@news> In-reply-to: andyba@corp.webtv.net's message of Mon, 03 Aug 1998 13:13:51 -0700 In article <andyba-ya02408000R0308981313510001@news>, andyba@corp.webtv.net (Andy Bates) writes: In article <6q24r9$74d$1@bartlet.df.lth.se>, cb@bartlet.df.lth.se (Christian Brunschen) wrote: > In article <andyba-ya02408000R3107981056140001@news>, > Andy Bates <andyba@corp.webtv.net> wrote: > >In article <SCOTT.98Jul30210152@slave.doubleu.com>, > > scott@nospam.doubleu.com (Scott Hess) wrote: > >> The problem is that you've taken the word "file" and redefined > >> it to have the usual meaning of "document". > > > >By the literal, dictionary definition of "file," it means the > >same thing. The only problem comes when you use the filesystem > >definition of "file." This definition does not matter to the > >user. But it _does_, because a file is simply a bag of bytes. The user can do a multitude of things with a file that they can't do with a document, because files have their own existance regardless of what documents they contain. Users understand very well that they can't just type in WordPerfect and email the result - unless they turn the document into a file and email the _file_. I think you underestimate users in this case. > I'm sorry, but the definition of 'File' in the computer sense > _is_ the filesystem definition; your attempt to re-define 'file' > to mean 'document' is wrong, and not supported, even by an > ordinary dictionary. If 'File' indeed was a synonym to > 'Document', I think a dictionary would reflect that; Then how do you explain the fact that for years, users have used the term interchangably? Why haven't they been confused by the File moniker when it came time to printing their document? Because to the user, there is no distinction. Non sequitur. Someone might say "Turn the car around", rather than "Turn the steering wheel until the car is going in the other direction." This doesn't mean that people confuse the concepts for "car" and "steering wheel", it just means that "turn the car" is a convenient shorthand. I'd be surprised if any but the most neophyte user doesn't have an understanding of the differences between "document" and "file". "Documents" are those things you're looking at on the screen, "Files" are where documents go to live when you aren't looking at them. Just because they elide the differences doesn't mean they don't recognize the differences. The thing is, for the most part documents come from files, and files store documents. So it's convenient to say "Send me the Archibald document" rather than "Send me the file containing the Archibald document". But the person sending the document doesn't send it as a "document" - they send it as a _file_. They can use the verbal shorthand without misunderstanding the differences. I really can't understand why you're arguing "But it's still a file" when it is abundantly clear that not all documents are files, but all files are documents. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer Subject: cmsg cancel <yEwx1.4013$eK3.35120@198.235.216.4> Control: cancel <yEwx1.4013$eK3.35120@198.235.216.4> Date: 04 Aug 1998 05:06:12 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.yEwx1.4013$eK3.35120@198.235.216.4> Sender: chrisavg@njvdyrij.edu Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: paipai@tin.it (Paolo Di Francesco) Newsgroups: comp.sys.next.programmer Subject: Re: XServer Date: Mon, 03 Aug 1998 00:05:03 GMT Organization: TIN Message-ID: <35c4fe6b.602906@news.tin.it> References: <35C0B1E0.E691AFD6@jove.acs.unt.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit On Thu, 30 Jul 1998 12:48:16 -0500, "Michael A. Thompson" <mat0001@jove.acs.unt.edu> wrote: >Anybody know of a free X server that works under NS3.3 black? tried >mouseX but it doesnt seem to work. Video goes black and I changed >permissions on /dev/vid0. Try Xnext. It's not free (shareware) but you can try it... http://www.angelfire.com/biz/sle/Xnext.html
From: kcd@babylon5.jumpgate.com (Kenneth C. Dyke) Newsgroups: comp.sys.next.programmer Subject: Re: OpenStep/NT and calling non-OS DLLs. Date: 4 Aug 1998 06:35:27 GMT Message-ID: <6q6a3f$86o$1@nntp2.ba.best.com> References: <SCOTT.98Aug3211550@slave.doubleu.com> In-Reply-To: <SCOTT.98Aug3211550@slave.doubleu.com> On 08/03/98, Scott Hess wrote: >A customer needs us to call some code in a third party dll. We've >figured it out from the Windows side - but I can't get the OpenStep >side to play ball. > >As a trial run, I decided to try to build an OpenStep program using >the the zlib dll available from http://www.winimage.com/zLibDll/. >Basically, this is the zip/gzip core in library form. > >I created a new "Tool" project. I pulled in zlib.h and zconf.h. I >added zlib.lib to the Libraries section. I added the line: > > printf( "Zlib version is %s\n", zlibVersion()); > >where it said "Insert code here" in the _main.m file. Then I tried to >build. > >Everything compiled, but I get the link errors: > >ZLibTest_main.o : error LNK2001: unresolved external symbol _zlibVersion >C:/users/scott/ZLibTest/ZLibTest.exe : \ > fatal error LNK1120: 1 unresolved externals I have used Win32 DLL's within OpenStep before (such as Glide, OpenGL, etc.)... it's just a matter of getting the calling conventions right. Add this to the zconf.h file: #if defined (__GNUC__) # if defined (WIN32) # include <windows.h> # define ZEXTERN WINAPI # endif #endif Put it anywhere before the following snippet: #ifndef ZEXPORT # define ZEXPORT #endif #ifndef ZEXPORTVA # define ZEXPORTVA #endif #ifndef ZEXTERN # define ZEXTERN extern #endif I was able to link and got the library version of 1.1.3. -Ken -- Kenneth Dyke, kcd@jumpgate.com (personal), kdyke@ea.com (work) Nuclear Strike and OPENSTEP Tools Engineer, Electronic Arts C++: The power, elegance and simplicity of a hand grenade.
From: far_no@spam.ix.netcom.com(Felipe A. Rodriguez) Newsgroups: gnu.gnustep.discuss,comp.sys.next.programmer,comp.sys.next.advocacy,comp.os.linux.development Subject: Another day, another GNUstep screenshot... Date: 4 Aug 1998 10:16:01 GMT Organization: ICGNetcom Message-ID: <6q6n11$5bs@sjx-ixn10.ix.netcom.com> For all those curious to see how the GNUstep project is progressing I've posted a screenshot at: http://pweb.netcom.com/~far/far.html In the foreground you'll see the XRAW Workspace example which is a GNUstep clone of the NeXTSTEP Workspace. And behind it is the very new Edit example. Despite the very early development stage of the Workspace example performance is pretty good with the exception of the NSSplitView. At the moment all you can actually do with the Workspace is browse your local filesystem. The Edit example is at a very early stage and does little more than allow simple editing of the text with which it is initialized. The examples are built using the gstep-xraw backend. To build the examples you need the very latest GNUstep sources from the CVS repository (a lot of what you see has been committed within the past few days). For more info on GNUstep: http://www.gnustep.org And if anyone is interested we need volunteers. -- Felipe A. Rodriguez # Francesco Sforza became Duke of Milan from Agoura Hills, CA # being a private citizen because he was # armed; his successors, since they avoided far@ix.netcom.com # the inconveniences of arms, became private (NeXTmail preferred) # citizens after having been dukes. (MIMEmail welcome) # --Nicolo Machiavelli
From: Paul Seelig <pseelig@karlo.zdv.Uni-Mainz.DE> Newsgroups: gnu.gnustep.discuss,comp.sys.next.programmer,comp.sys.next.advocacy Subject: Re: Another day, another GNUstep screenshot... Date: Tue, 4 Aug 1998 14:04:07 +0200 Organization: Johannes Gutenberg-Universitaet Mainz, Germany Message-ID: <Pine.OSF.3.95.980804140038.10841A-100000@karlo.zdv.Uni-Mainz.DE> References: <6q6n11$5bs@sjx-ixn10.ix.netcom.com> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII To: discuss-gnustep@gnu.org In-Reply-To: <6q6n11$5bs@sjx-ixn10.ix.netcom.com> On 4 Aug 1998, Felipe A. Rodriguez wrote: > For all those curious to see how the GNUstep project is progressing > I've posted a screenshot at: > > http://pweb.netcom.com/~far/far.html > It's good to see that GNUstep development is alive and progressing. Thanks for this very nice screenshot. Cheers, P. *8^) -- Paul Seelig pseelig@goofy.zdv.uni-mainz.de African Music Archive - Institute for Ethnology and Africa Studies Johannes Gutenberg-University - Forum 6 - 55099 Mainz/Germany Our WWW pages to be visited at "http://www.uni-mainz.de/~bender"
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: OpenStep/NT and calling non-OS DLLs. Date: 4 Aug 98 10:15:47 Organization: Is a sign of weakness Message-ID: <SCOTT.98Aug4101547@slave.doubleu.com> References: <SCOTT.98Aug3211550@slave.doubleu.com> <6q6a3f$86o$1@nntp2.ba.best.com> In-reply-to: kcd@babylon5.jumpgate.com's message of 4 Aug 1998 06:35:27 GMT In article <6q6a3f$86o$1@nntp2.ba.best.com>, kcd@babylon5.jumpgate.com (Kenneth C. Dyke) writes: On 08/03/98, Scott Hess wrote: >A customer needs us to call some code in a third party dll. We've >figured it out from the Windows side - but I can't get the >OpenStep side to play ball. I have used Win32 DLL's within OpenStep before (such as Glide, OpenGL, etc.)... it's just a matter of getting the calling conventions right. Bingo! So obvious once you figure it out! I've put a modified version of ZLibTest at http://www.doubleu.com/ZLibTest.zip. Note that it doesn't _do_ anything, it just calls the version function in the dll. I'm sure you could put an object wrapper around the dll, but that wasn't my aim :-). Thanks much! -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: "Agent" <agent@siu.edu> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Tue, 4 Aug 1998 02:11:27 -0500 Organization: ArcaMax Message-ID: <6q7nrc$uhu@saluki-news.it.siu.edu> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk><68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <HaS8aCAeaCw1UAdO@ffaltd.demon.co.uk> All right, now that everybody's got all that resentment and condescension out of their system...(god bless ye as what's got a sense o' humour...humor...whate'er) -- Agent
From: arti@lava.DOTnet (Art Isbell - remove "DOT") Newsgroups: comp.sys.next.programmer Subject: Re: Details on porting NEXTSTEP app to OPENSTEP Date: 5 Aug 1998 03:42:23 GMT Organization: LavaNet, Inc. Distribution: world Message-ID: <6q8kav$rge@mochi.lava.net> References: <6plpa3$a36$1@vixen.cso.uiuc.edu> <35C10A78.BC0D3088@forsee.tcp.co.uk> <SCOTT.98Jul30204249@slave.doubleu.com> <6psu1i$3vt1@onews.collins.rockwell.com> embuck@palmer.cca.rockwell.com (Erik M. Buck) wrote: > Just for those who are getting confused about conflicting claims. > > Converting from NeXTstep to Openstep can take weeks or months. > Once you have an Openstep application, moving it from Mach to NT is hours or > days. And if you're unfortunate enough to have a DBKit app that you want to convert to OPENSTEP so you can run it under Windows, conversion isn't really possible. You're dealing with a rewrite, not a conversion. -- Art Isbell NeXT/MIME Mail: arti@lavaDOTnet IDX Systems Corporation Voice/Fax: +1 808 526 1226 (for whom I don't speak) Voice Mail: +1 808 533 1827 Healthcare Info Technology US Mail: Honolulu, HI 96813-1021
From: Arnaud <debayeux@easynet.fr> Newsgroups: comp.sys.next.programmer Subject: Custom class in a nib Date: Wed, 05 Aug 1998 04:55:17 +0200 Organization: [posted via] Easynet France Message-ID: <35C7C995.7F3F333@easynet.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello, I'm using the palette MiscClassDecoder.palette from MiscKit in IB. I use a custom class MiscClassDecoder, when I test the interface everything is OK. But when a simple application using this nib file is created, it can not run because it doesn't find the class MiscClassDecoder. I needed to explicitely load the class with the following code : Class exampleClass; NSString *str = @"/MiscClassDecoderPalette.palette"; NSBundle *bundleToLoad = [NSBundle bundleWithPath:str]; exampleClass = [bundleToLoad principalClass]; Shouldn't the nib loader know how to find and load the necessary dll (the palette) ? Why doesn't it use the data.dependency file ? Also the resizing of NSViews containing buttons with the resizing attributes NSViewMinXMargin | NSViewMaxXMargin or NSViewMinYMargin | NSViewMaxYMargin are badly repositioned. This problem is not present in OS 4.2 Mach, only in Enterprise. I would like to override the faulty method. Which method should be corrected and what should be the new code ? Another question : did anybody successfully rebuild the compiler under OSE 4.2 ? I would like to slightly modify the code generation for the calls of functions imported from a dll. The current generated code is (with -O3 optimization) : mov [__imp__function], %eax call %eax I would like (just like the code generated by VC++) : call [__imp__function] What should be modified in the gcc compiler ? (RTL...) Thanks for nany help, Arnaud
From: "William Edward Woody" <woody@alumni.caltech.edu> Newsgroups: comp.sys.next.programmer,comp.sys.mac.programmer.misc Subject: Re: Distinction between Files and Documents (was Re: Mac OS X UI) Date: Wed, 5 Aug 1998 13:56:47 -0700 Organization: Pacific Bell Internet Services Message-ID: <6qahcv$5l8$1@nnrp2.snfc21.pbi.net> References: <6odtu6$lh3$5@lwnws01.ne.highway1.com> <andyba-ya02408000R3007981322040001@news> <SCOTT.98Jul30210152@slave.doubleu.com> <andyba-ya02408000R3107981056140001@news> <6q24r9$74d$1@bartlet.df.lth.se> Christian Brunschen wrote in message <6q24r9$74d$1@bartlet.df.lth.se>... >I'm sorry, but the definition of 'File' in the computer sense _is_ the >filesystem definition; your attempt to re-define 'file' to mean 'document' >is wrong, and not supported, even by an ordinary dictionary. ... Excuse me for jumping into your little religious flame war, but: File: (n) ... 2(a) A collection of papers or published materials kept or arranged in convenient order. (b) [Computer Science] A collection of related data or program records. ... Document: (n) 1(a) A written or printed paper that bears the original, official, or legal form of something and can be used to furnish evidence or information. (b) Something, such as a recording or a photograph, that can be used to furnish evidence or information. (c) A writing that contains information. ... [Portions snipped from The American Heritage Dictionary of the English Language, Third Edition. Note that the usage 2(a) of "File" is reflected in the definitions you quoted.] The problem with using either "File" or "Document" is that neither precisely defines what it is that people are manipulating. Your suggestion that we use "Document" seems silly to me--as often the "document" is not "writing", but an image. (Note also that "document", at least as defined by the American Heritage Dictionary, tends to have the connotation of "official"--such as a legal writing or contract.) To be precise, we are manipulating computer files. (In the sense 2(b) above--a collection of related data or program records.) And so your suggestion to use "Document" instead of "File" seems a bit imprecise--the hijacking of the English Language by Computer Nerds who can't be bothered to remember what "document" and "file" meant not withstanding. But that's just my $0.02 worth. Now back to your regularly scheduled flame war. - Bill Woody woody@pandawave.com The PandaWave http://www.pandawave.com
From: dunham@cs.tulane.edu (Andrea Dunham) Newsgroups: comp.sys.next.software,comp.sys.next.programmer Subject: Gnuplot 3.6 on NS 3.3 Date: 6 Aug 1998 00:20:31 GMT Organization: Elec Engr & Comp Sci Dept, Tulane Univ, New Orleans, LA Message-ID: <6qassf$udk$1@rs10.tcs.tulane.edu> Hi All, I need help compiling Gnuplot3.6 (beta) to run on a Nextstation OS-3.3. I used "configure" to build the Makefile. And "make" gives me: ----------------------------------------------------------------------- ... cc -c -ObjC -I. -I. -I./term -I./term -DHAVE_CONFIG_H -g -O2 term.c ./term/next.trm:94: warning: could not use precompiled header '/NextDeveloper/Headers/appkit/appkit.p', because: ./term/next.trm:94: warning: macro 'fputc' defined by ./term/driver.h conflicts with precomp ... cc -o gnuplot alloc.o binary.o bitmap.o command.o contour.o datafile.o eval.o fit.o graphics.o graph3d.o help.o hidden3d.o internal.o interpol.o matrix.o misc.o parse.o plot.o plot2d.o plot3d.o readline.o scanner.o set.o show.o specfun.o standard.o stdfn.o term.o time.o util.o util3d.o epsviewe.o version.o -lsys_s -lNeXT_s -lm ld: Undefined symbols: _tcgetattr _tcsetattr *** Exit 1 Stop. ----------------------------------------------------------------------- I peeked in the configure file and still I can't figure how to fix this problem. Any help you can give me will be much appreciated. Thanks, * AndREa * (dunham@eecs.tulane.edu)
From: dsmith_32@hotmail.com Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.pen,comp.sys.powerpc.advocacy Subject: cmsg cancel <35c2c1ba.0@204.140.208.3> Control: cancel <35c2c1ba.0@204.140.208.3> Date: 6 Aug 1998 05:46:32 -0400 Organization: University of Economics and Business Administration, Vienna, Austria Sender: root@cantine.wu-wien.ac.at Distribution: inet Message-ID: <cancel.1.35c2c1ba.0@204.140.208.3> Spam Cancelled by news-admin@wu-wien.ac.at
From: boehring@biomed.ruhr-uni-bochum.de (Daniel Boehringer) Newsgroups: comp.sys.next.software,comp.sys.next.programmer Subject: Re: Gnuplot 3.6 on NS 3.3 Date: 6 Aug 1998 10:21:45 GMT Organization: Ruhr-Universitaet Bochum, Rechenzentrum Message-ID: <6qc03p$9hq$1@sun579.rz.ruhr-uni-bochum.de> References: <6qassf$udk$1@rs10.tcs.tulane.edu> NNTP-Posting-Date: 6 Aug 1998 10:21:45 GMT dunham@cs.tulane.edu (Andrea Dunham) wrote: >Hi All, > >I need help compiling Gnuplot3.6 (beta) to run on a Nextstation OS-3.3. >I used "configure" to build the Makefile. And "make" gives me: >ld: Undefined symbols: >_tcgetattr >_tcsetattr >*** Exit 1 try adding to other sources: /*- * Copyright (c) 1989, 1993 * The Regents of the University of California. All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 3. All advertising materials mentioning features or use of this software * must display the following acknowledgement: * This product includes software developed by the University of * California, Berkeley and its contributors. * 4. Neither the name of the University nor the names of its contributors * may be used to endorse or promote products derived from this software * without specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. */ #if defined(LIBC_SCCS) && !defined(lint) static char sccsid[] = "@(#)termios.c 8.2 (Berkeley) 2/21/94"; #endif /* LIBC_SCCS and not lint */ #include <sys/types.h> #include <sys/fcntl.h> #include <sys/ioctl.h> #include <sys/time.h> #include <sys/file.h> #include <errno.h> #include <termios.h> #include <unistd.h> typedef int pid_t; int tcgetattr(fd, t) int fd; struct termios *t; { return (ioctl(fd, TIOCGETA, t)); } int tcsetattr(fd, opt, t) int fd, opt; const struct termios *t; { struct termios localterm; if (opt & TCSASOFT) { localterm = *t; localterm.c_cflag |= CIGNORE; t = &localterm; } switch (opt & ~TCSASOFT) { case TCSANOW: return (ioctl(fd, TIOCSETA, t)); case TCSADRAIN: return (ioctl(fd, TIOCSETAW, t)); case TCSAFLUSH: return (ioctl(fd, TIOCSETAF, t)); default: errno = EINVAL; return (-1); } } int #if __STDC__ tcsetpgrp(int fd, pid_t pgrp) #else tcsetpgrp(fd, pgrp) int fd; pid_t pgrp; #endif { int s; s = pgrp; return (ioctl(fd, TIOCSPGRP, &s)); } pid_t tcgetpgrp(fd) int fd; { int s; if (ioctl(fd, TIOCGPGRP, &s) < 0) return ((pid_t)-1); return ((pid_t)s); } speed_t cfgetospeed(const struct termios *t) { return (t->c_ospeed); } speed_t cfgetispeed(t) const struct termios *t; { return (t->c_ispeed); } int cfsetospeed(t, speed) struct termios *t; speed_t speed; { t->c_ospeed = speed; return (0); } int cfsetispeed(t, speed) struct termios *t; speed_t speed; { t->c_ispeed = speed; return (0); } void cfsetspeed(t, speed) struct termios *t; speed_t speed; { t->c_ispeed = t->c_ospeed = speed; } /* * Make a pre-existing termios structure into "raw" mode: character-at-a-time * mode with no characters interpreted, 8-bit data path. */ void cfmakeraw(t) struct termios *t; { t->c_iflag &= ~(IMAXBEL|IXOFF|INPCK|BRKINT|PARMRK|ISTRIP|INLCR|IGNCR|ICRNL|IXON|IGNPAR); t->c_iflag |= IGNBRK; t->c_oflag &= ~OPOST; t->c_lflag &= ~(ECHO|ECHOE|ECHOK|ECHONL|ICANON|ISIG|IEXTEN|NOFLSH|TOSTOP|PENDIN); t->c_cflag &= ~(CSIZE|PARENB); t->c_cflag |= CS8|CREAD; t->c_cc[VMIN] = 1; t->c_cc[VTIME] = 0; } int tcsendbreak(fd, len) int fd, len; { struct timeval sleepytime; sleepytime.tv_sec = 0; sleepytime.tv_usec = 400000; if (ioctl(fd, TIOCSBRK, 0) == -1) return (-1); (void)select(0, 0, 0, 0, &sleepytime); if (ioctl(fd, TIOCCBRK, 0) == -1) return (-1); return (0); } int tcdrain(fd) int fd; { return (ioctl(fd, TIOCDRAIN, 0)); } int tcflush(fd, which) int fd, which; { int com; switch (which) { case TCIFLUSH: com = FREAD; break; case TCOFLUSH: com = FWRITE; break; case TCIOFLUSH: com = FREAD | FWRITE; break; default: errno = EINVAL; return (-1); } return (ioctl(fd, TIOCFLUSH, &com)); } #define _POSIX_VDISABLE 0 int tcflow(fd, action) int fd, action; { struct termios term; u_char c; switch (action) { case TCOOFF: return (ioctl(fd, TIOCSTOP, 0)); case TCOON: return (ioctl(fd, TIOCSTART, 0)); case TCION: case TCIOFF: if (tcgetattr(fd, &term) == -1) return (-1); c = term.c_cc[action == TCIOFF ? VSTOP : VSTART]; if (c != _POSIX_VDISABLE && write(fd, &c, sizeof(c)) == -1) return (-1); return (0); default: errno = EINVAL; return (-1); } /* NOTREACHED */ }
From: *johnnyc*@or.psychology.dal.ca (John Christie) Newsgroups: comp.sys.next.programmer,comp.sys.mac.programmer.misc Subject: Re: Distinction between Files and Documents (was Re: Mac OS X UI) Date: Thu, 06 Aug 1998 09:08:29 -0300 Organization: ISINet, Nova Scotia Message-ID: <*johnnyc*-0608980908290001@jchristie.halifaxcable.dal.ca> References: <6odtu6$lh3$5@lwnws01.ne.highway1.com> <andyba-ya02408000R3007981322040001@news> <SCOTT.98Jul30210152@slave.doubleu.com> <andyba-ya02408000R3107981056140001@news> <6q24r9$74d$1@bartlet.df.lth.se> <6qahcv$5l8$1@nnrp2.snfc21.pbi.net> In article <6qahcv$5l8$1@nnrp2.snfc21.pbi.net>, "William Edward Woody" <woody@alumni.caltech.edu> wrote: > Excuse me for jumping into your little religious flame war, but: > > File: (n) > ... 2(a) A collection of papers or published materials kept or > arranged in convenient order. > (b) [Computer Science] A collection of related data or > program records. > ... > > Document: (n) > 1(a) A written or printed paper that bears the original, > official, or legal form of something and can be used to > furnish evidence or information. > (b) Something, such as a recording or a photograph, that > can be used to furnish evidence or information. > (c) A writing that contains information. > ... > ... > > The problem with using either "File" or "Document" is that neither > precisely defines what it is that people are manipulating. And, given the above definitions (and ones posted earlier), a Database would definitely be a file. I think we should just call them all Woogshi. -- You aren't free if you CAN choose - only if you DO choose. All you are is the decisions you make. Remove "*" and ohnny (i.e. jc@) to reply via email
From: mmh@dcs.qmw.ac.uk (Matthew M. Huntbach) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Followup-To: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Date: 6 Aug 1998 14:00:02 GMT Organization: Queen Mary & Westfield College, London, UK Message-ID: <6qcct2$aie$1@beta.qmw.ac.uk> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <6pnlfm$ob$2@beta.qmw.ac.uk> <8F12A6803BCB46BB.69250C85F5006041.55EF65CE3294D5C9@library-proxy.airnews.net> jdm (jdm1intxDIE_SPAMBOT_DIE@airmail.net) wrote: > Matthew M. Huntbach wrote in message <6pnlfm$ob$2@beta.qmw.ac.uk>... > >Sorry? Read just about any textbook on programming or software engineering > >used by reputable teachers the world over, and you'll see there's a clear > >message that there's a world of difference between a program sloppily > >hacked together that appears to do the job, and a program carefuully > >designed that we can be reasonably certain does the job and can be easily > >understood and modified for future use. > Anybody can write a textbook. Sure, anyone can. That's why I said "just about any textbook used by reputable teachers". When I say "reputable teachers" I mean teachers in the sort of institutions the big companies tend to recruit Computer Science graduates from. And when I say "just about any" I mean that. It's not just that one textbook says one thing, another says something else - they all say the same thing, stressing the importance of a disciplined approach to programming. > And anybody can proclaim themselves to be the embodiment and zenith of all > wisdom and knowledge pertaining to computer science. Sure they can. That is why a useful test is what is used in practice, what the big companies, who are trying to make money out of programming, prefer. Why has industry gone for object-oriented programming in a big way in recent years? That move can't be blamed on academics, who if anything were tending to promote functional and/or logic languages while indutry sneaked up on them and got the OO bug. But OO programming does impose a disciplined structure to programming, and OO languages force that structure to be kept to. That was all Mr.Sheglova was saying. If what he was saying was just over-academic pretension, why is indutsry moving in a big way towards Java? Why isn't it content to stick with the old-style unstructured languages? > That doesn't make them any better of a programmer than the poor > weenie with a degree from Acme Technical Institute and Pizzeria, cranking > out code in a low-paying 3rd rate sweatshop in Redmond. For every genius > with a Phd in Computer Science from MIT you can show me, I can show you a > guy who never even learned what a linked list was, but can still turn out > decent code in a plain-jane OLTP application. Sure, maybe he will scratch > his head and say "duh?" if you ask him to write a indexing routine based on > AVL trees, but that doesn't mean he isn't an inferior programmer. Maybe not. Ignorance doesn't mean stupidity. Someone who hasn't been taught modern methods of programming may nevertheless be able to do very clever things with an old-fashioned approach. Someone who has worked all their life in a specialist field of programming may be very good in that field, but know little about programming more generally. But linked-lists and AVL trees are just standard data structures which I was taught in Uni twenty years ago. I wouldn't have thought it unreasonable to expect a programmer these days to be aware of such things. Someone who missed out on being taught about them formally ought, if they are any good, to be able to learn about them by reading about them. Someone who has the intelligence to be able to pick up a book, read about some algorithm or data structure, and write a program which implements it must surely count as a better programmer in at least some ways than someone who cannot do that. Surely that is what being a programmer is all about - being able to write a program from a specification. > >You owe an apology to Mr.Sheglova (who is quite possibly not British, as > >"Sheglova" is not a British name) for your quite unnecessary nationalistic > >remark. I doubt you have the guts to make one. > If I felt I had anything to apologize for, I would certainly have the guts > to. However, Mr. Sheglova's sneering, condescending attitude towards "the > masses" speaks for itself. I will not apologize to him, nor will I > apologize for my Nationalistic tendencies in the face of arrogant Europeans > who feel the need to patronize us atavistic Amerikan schweinhundts. I did not see any remarks about any nationalism, or comparisons of one country with another in Mr.Sheglova's remarks. When I said "the world over" in my quoted remarks above, I meant the USA as well. There was absolutely no need for you to bring in your nationalistic comments, and the fact that you did just points out further that you are probably some rather unbalanced type with a big chip on your shoulder. Matthew Huntbach
From: CR<myemail@any.where.com> Newsgroups: comp.sys.next.programmer Subject: New site for toners and services! Date: 6 Aug 1998 14:33:09 GMT Organization: CR Message-ID: <6qcer5$u97@news-1.golden.net.au> http://www.cityrecharge.com.au Melbourne Australia To remove please email us at cityrecharge@theoffice.net
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6qcer5$u97@news-1.golden.net.au> Control: cancel <6qcer5$u97@news-1.golden.net.au> Date: 06 Aug 1998 14:56:53 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6qcer5$u97@news-1.golden.net.au> Sender: CR<myemail@any.where.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: nouser@nohost.nodomain (Thomas) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 06 Aug 1998 04:26:00 -0700 Organization: home Sender: nouser@nohost.nodomain Message-ID: <tz87m0mnygn.fsf@aimnet.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> In-reply-to: boracay@hotmail.com's message of Thu, 02 Jul 1998 13:48:58 GMT In article <6ng34a$j2g$1@nnrp1.dejanews.com> boracay@hotmail.com writes: 3)If only a computer scientist can program it or understand all the complexities surrounding development with it, and not your average scientist, engineer or other individual with basic computer training, then it is too complicated and is NOT an advancement. If a computer scientists or programmer spends 40h a week working with computers, he will always have a leg up and be able to tackle more complex problems than an average engineer that can spend only a few hours on computers (the engineer gets paid to do engineering, after all). That's the way things are: you spend more time on something, you get better at it. And different levels of experience require different tools. What strikes you as user hostile may be the most effective tool for someone who uses it daily. Thomas.
From: wmoss@maclink.net (William Moss) Newsgroups: comp.sys.next.programmer,comp.sys.mac.programmer.misc Subject: Re: Distinction between Files and Documents (was Re: Mac OS X UI) Date: Thu, 06 Aug 1998 12:25:56 -0400 Organization: ISPNews http://ispnews.com Sender: mac@ip208.atlanta6.ga.pub-ip.psi.net Message-ID: <wmoss-0608981225580001@ip208.atlanta6.ga.pub-ip.psi.net> References: <6odtu6$lh3$5@lwnws01.ne.highway1.com> <andyba-ya02408000R3007981322040001@news> <SCOTT.98Jul30210152@slave.doubleu.com> <andyba-ya02408000R3107981056140001@news> <6q24r9$74d$1@bartlet.df.lth.se> NNTP-Posting-Date: 6 Aug 1998 16:21:14 GMT Having had real instruction on using a Mac back in 1985, I was taught that menus were "verbs" which you chose to act on the current selection (an icon, text, or picture) which was a noun. The "Apple" menu was iconinc because it represented other nouns and not an action that was to be taken on the currently selected object (it contained other nouns in the form of desk accessories). The terms "File", "Edit", "Style", etc. were read as verbs. In that context, it was somewhat logical that you would "file" your work to the printer to get a hard copy, or "save" it on your disk for later use. You would "underline" a sentence (which was a specific way that you could "style" a sentence). But this metaphor of noun=selection and verb=menu didn't hold up even in the early days. It stretches the bounds of good English to use "bold" as a verb instead of "embolden". And the logical connection between the verb to "Quit" and the verb "File" became very tenuous when apps became savy enough to have multiple documents open at one time. OpenDoc recently went down the road of trying to replace the term "File" with the term "Document" as the first item in the menu bar after the Apple. One of the problems for OpenDoc users was the switch in human interface. Even though it is no longer logical today, people expected a Quit item as the last item under the Print commands. I think CyberDog even violated these new OpenDoc user interface rules and gave it's users a "Quit" command in its later versions. I think replacing "File" with "Document" is a change that is both unnecessary (as today's menus use various parts of speech beyond just verbs), and the user experience baggage of where to expect commands. William
From: mmh@dcs.qmw.ac.uk (Matthew M. Huntbach) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Followup-To: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Date: 6 Aug 1998 14:07:30 GMT Organization: Queen Mary & Westfield College, London, UK Message-ID: <6qcdb2$aie$2@beta.qmw.ac.uk> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <6pnlfm$ob$2@beta.qmw.ac.uk> <8F12A6803BCB46BB.69250C85F5006041.55EF65CE3294D5C9@library-proxy.airnews.net> <6ppj8r$bcs$1@crib.corepower.com> Nathan Urban (nurban@crib.corepower.com) wrote: > In article <8F12A6803BCB46BB.69250C85F5006041.55EF65CE3294D5C9@library-proxy.airnews.net>, "jdm" <jdm1intxDIE_SPAMBOT_DIE@airmail.net> wrote: > > who feel the need to patronize us atavistic Amerikan schweinhundts. > I'll patronize you myself, because you're a bigoted juvenile moron who's > probably jealous of superior programmers with more training than you. > I'm American, and I think that you're the reason why people of other > countries rightly deserve to sneer at us. There's too many of your kind > around here. You're just as arrogant, even more bigoted, and also with > a healthy dose of the usual American anti-intellectualism that's harming > the educational system and our country's ability to compete. Thanks for that. I'm aware with the USA being the dominant nationality on the net, the likelihood is purely by the law of averages that any nutter you encounter on the net will be an American. We should beware of drawing from that the conclusion that Americans are more likely to be nutters than other nationalities. Nevertheless, in my many years of using the net, I do find that there's a sort of simple-minded "My country is the best and all the others stink" nationalism that is not uncommon amongst American posters, but hardly ever found amongst those of other nationalities. And, yes, coming up against that constantly does have a negative effect. On the whole, I think in my years of using the net, I have tended to become more anti-American than I ever used to be, just because I have spent so much time arguing with small-minded nationalistic bigots from the USA like "jdm". Sorry to the rest of you Americans for that, but it's a problem you have to deal with - it's your image that's at stake and is being made a laughing stock by people like "jdm". Matthew Huntbach
From: far_no@spam.ix.netcom.com(Felipe A. Rodriguez) Newsgroups: gnu.gnustep.discuss,comp.sys.next.programmer Subject: NSString Additions Date: 6 Aug 1998 18:21:24 GMT Organization: ICGNetcom Message-ID: <6qcs74$98t@sjx-ixn10.ix.netcom.com> At the suggestion of an individual who wants to help with the GNUstep project but who must do so while coding on 4.2. I decided to implement the NSString additions. Basically these are three methods which tend to abstract the writing of the text classes. Unfortunately, it has occured to me after having completed the initial implementation :-( that the GNUstep project may not have the right to use this part of the OPENSTEP API. In particular the header file and documentation both carry the NeXT copyright. And while I'm not using the actual header file or documentation it can't be argued that I just happened to choose the same names as NeXT for both the methods and global definintions. So does anyone know whether additions made in 4.1/4.2 are part of the OPENSTEP spec? Comments? -- Felipe A. Rodriguez # Francesco Sforza became Duke of Milan from Agoura Hills, CA # being a private citizen because he was # armed; his successors, since they avoided far@ix.netcom.com # the inconveniences of arms, became private (NeXTmail preferred) # citizens after having been dukes. (MIMEmail welcome) # --Nicolo Machiavelli
From: "Judson McClendon" <judmc123@bellsouth.net> Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng References: <6ng34a$j2g$1@nnrp1.dejanews.com> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <6pnlfm$ob$2@beta.qmw.ac.uk> <8F12A6803BCB46BB.69250C85F5006041.55EF65CE3294D5C9@library-proxy.airnews.net> <6ppj8r$bcs$1@crib.corepower.com> <6qcdb2$aie$2@beta.qmw.ac.uk> Subject: Re: Programming and the colossal failure to advance Organization: Sun Valley Systems Message-ID: <tcny1.729$ML4.2269662@news4.mia.bellsouth.net> Date: Thu, 06 Aug 1998 19:10:17 GMT NNTP-Posting-Date: Thu, 06 Aug 1998 15:10:17 EST Matthew M. Huntbach wrote: > >Nevertheless, in my many years of using the net, I do find that there's a >sort of simple-minded "My country is the best and all the others stink" >nationalism that is not uncommon amongst American posters, but hardly ever >found amongst those of other nationalities. I'll have to take issue with that particular point, Matthew. My experience is that people from all nationalities are pretty much the same in this regard. Sure, people are proud of their own countries. I don't blame them for that. Anyone who is not should leave and go somewhere they believe is better. What provokes me is unfair criticism, whether directed at other's countries, or their own. It is true that Americans can be arrogant and obnoxious. But at the same time, I believe America gets a lot of unfair criticism. No other country has sacrificed so much for others as has the U.S., or been more generous to its friends, and enemies. There is no parallel in history to the way the U.S. has helped rebuild former enemies, like Japan and Germany. Every person in Europe and the former U.S.S.R. can be thankful that they have freedom in large part because the U.S. spent its wealth and blood to buy it, throughout this century. Very few people on the face of the earth have not been blessed in some way by the influence and actions of the U.S. How long do you think Europe would have remained free from the Soviet Union throughout the Cold War, had it not been for the deterrence of the U.S.? If not for the U.S., Sadam Hussein would likely be sitting in Saudi Arabia, and the lights would be mighty dim in a lot of countries. In most free countries, particularly European countries, every citizen has paid less tax toward their own defense because the U.S. stands as a strong defense against aggression. Most Americans are well aware of these facts, and are sensitive to America being criticized by people in other countries who have benefited from these things, and offer only criticism, but never express any appreciation. No country is perfect, but America typically receives far more criticism than praise. As an American, what I would expect is that those who would criticize America for its faults, also consider the good the U.S. has done, and continues to do, for the rest of the world. Just balance it out a bit; fair is fair. -- Judson McClendon This is a faithful saying and worthy of all Sun Valley Systems acceptance, that Christ Jesus came into the judmc123@bellsouth.net world to save sinners (1 Timothy 1:15) (please remove numbers from email address to respond) http://personal.bhm.bellsouth.net/bhm/~judmc
From: alvin@cse.ucsc.edu (Alvin Jee) Newsgroups: comp.sys.next.programmer Subject: Create EPS files? Date: 6 Aug 1998 22:34:34 GMT Organization: UC Santa Cruz CS/CE Message-ID: <6qdb1q$jge@darkstar.ucsc.edu> Hello All! I hope that this isn't a silly question... How does one create EPS files from within a program? I have some programs that can generate PS files via the print panel, but I'd like to get EPS files out of the app also. It doesn't look like there's a handy printEPS method in the Appkit. A search of DejaNews comes up empty also. By the way, this is mainly for NS3.3 and OS4.2. Thanks! -- Alvin Jee alvin@neander.com http://www.neander.com NeXTMail gleefully accepted!
From: Michael Simpson <t_msimps@qualcomm.com> Newsgroups: comp.sys.next.programmer Subject: Calling a routine in a Windows DLL Date: Thu, 06 Aug 1998 15:32:56 -0700 Organization: Qualcomm Message-ID: <35CA2F18.19B1DDD3@qualcomm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Does anyone have an example of how to call a routine in a Window's DLL from objective C? From VC++, I would use the import lib and then just make the call. I'm just not certain how to set it up in Project Builder. I've looked through the docs and see references to it, but not specifics. Any help would be appreciated. Michael
From: Jodell Bumatai <jbumatai@inprise.com> Newsgroups: comp.sys.next.programmer Subject: Installing NS 3.3 Drivers Date: Thu, 06 Aug 1998 15:52:36 -0700 Organization: Inprise Corporation Message-ID: <35CA33B3.7B79A9D3@inprise.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I have a Pentium with IDE CD Rom. How do I install the drivers? JB
From: no9rma@ro5se5.org Newsgroups: comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar Subject: now purchasing model released photography 30678 Message-ID: <06089817.3030@ro5se5.org> Date: Thursday, 06 Aug 1998 17:30:30 -0600 Organization: Hello: Are you an advanced amateur or pro photographer interested in selling your work? If you are over the age of 21 and have a good quality 35mm or medium format camera please read further. The following "adult entertainment" websites need fresh original model released photos. Please contact them directly if you have an interest. You may download the appropriate model releases directly from the sites. These sites buy original exposed color negative film with model release. Models must be at least 18 years of age. You must be 21 to access any of these websites. Thank you. http://www.girliegirl.com http://www.sassygirl.com http://www.world-premiere.com ]HX(xQT3U_m[8GHENv(5h&$@FK=h)M1ko:jr1+g%PP&5JzCxIg :h"p7sA%7B;Ol,9<LtlEH'2SaN,#$"*iu(Es
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar Subject: cmsg cancel <06089817.3030@ro5se5.org> Control: cancel <06089817.3030@ro5se5.org> Date: 07 Aug 1998 00:46:28 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.06089817.3030@ro5se5.org> Sender: no9rma@ro5se5.org Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: "Dmitry A. Yevsukhin" <eda@next.msu.ru> Newsgroups: comp.sys.next.programmer Subject: libInterceptor Date: Fri, 7 Aug 1998 05:51:40 +0400 Message-ID: <35ca5dff.0@news.msu.ru> Hello Has anybody ever seen libInterceptor_s.a library and corresponding .h files? I wonder if I could get them. Thank you Dmitry
From: "Dmitry A. Yevsukhin" <eda@next.msu.ru> Newsgroups: comp.sys.next.programmer References: <35C7C995.7F3F333@easynet.fr> Subject: Re: Custom class in a nib Date: Fri, 7 Aug 1998 06:18:19 +0400 Message-ID: <35ca643f.0@news.msu.ru> Arnaud wrote in message <35C7C995.7F3F333@easynet.fr>... >Hello, > >I'm using the palette MiscClassDecoder.palette from MiscKit in IB. >I use a custom class MiscClassDecoder, when I test the interface >everything is OK. >But when a simple application using this nib file is created, it can not >run because it doesn't find the class MiscClassDecoder. >I needed to explicitely load the class with the following code : > Class exampleClass; > NSString *str = @"/MiscClassDecoderPalette.palette"; > NSBundle *bundleToLoad = [NSBundle bundleWithPath:str]; > exampleClass = [bundleToLoad principalClass]; >Shouldn't the nib loader know how to find and load the necessary dll >(the palette) ? It shouldn't. The palette is not intended to provide class implementation, although it does contain that code to be run under IB. You need to link your application with the corresponding Miscxxx library to get the code or build your own bundle containing the code and the nib.
From: guyt@is.twi.tudelft.nl (Abraham Guyt) Newsgroups: comp.sys.next.programmer Subject: Re: semaphores on NeXT? Date: 7 Aug 1998 08:53:38 GMT Organization: Delft University of Technology Message-ID: <6qefai$j8u$1@news.tudelft.nl> References: <clintdwExB1Mq.6qA@netcom.com> Clinton Wong writes > I'm looking through a NeXTStep programming manual and see mutex_init(), > mutex_lock(), mutex_unlock(), and cthread_detach(). But I don't > see anything about sempahores. Does such a thing exist? Or perhaps > there are third party libraries that I could use that implement > semaphores with a mutex and variable? > > Any info would be deeply appreciated... > Clinton > > -- > For a good time, read RFC 2324. http://postmaster.net/~clintdw/ Before I elaborate, would exactly are you trying to achieve with these semaphores ? Abraham G. _____________________________________________________________________ Abraham Guyt P.O.Box 356 Department of Information Systems 2600 AJ Delft Faculty Information Technology & Systems The Netherlands Delft University of Technology tel: +31 15 278 5969 E-mail: guyt@is.twi.tudelft.nl fax: +31 15 278 6632
Newsgroups: comp.sys.next.programmer From: clintdw@netcom.com (Clinton Wong) Subject: semaphores on NeXT? Message-ID: <clintdwExB1Mq.6qA@netcom.com> Organization: Netcom On-Line Services Date: Fri, 7 Aug 1998 05:52:02 GMT Sender: clintdw@netcom12.netcom.com I'm looking through a NeXTStep programming manual and see mutex_init(), mutex_lock(), mutex_unlock(), and cthread_detach(). But I don't see anything about sempahores. Does such a thing exist? Or perhaps there are third party libraries that I could use that implement semaphores with a mutex and variable? Any info would be deeply appreciated... Clinton -- For a good time, read RFC 2324. http://postmaster.net/~clintdw/
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.next.programmer Subject: Re: libInterceptor Date: 7 Aug 1998 07:19:51 GMT Organization: MiscKit Development Message-ID: <6qe9qn$24e$1@news.xmission.com> References: <35ca5dff.0@news.msu.ru> NNTP-Posting-Date: 7 Aug 1998 07:19:51 GMT "Dmitry A. Yevsukhin" <eda@next.msu.ru> wrote: > Has anybody ever seen libInterceptor_s.a library and > corresponding .h files? Yes. > I wonder if I could get them. Nope. Well, to be more precise, you can create the .a library yourself by using a little program called "OmniShlibExtractor". Since the shlibs are (c) NeXT, we can't be passing those (or derivatives) around. And you have to figure out how to get the extractor program, since I can't give it out because it is (c) somebody else. You can create the headers by using the freely available class-dump; but that's the easy part... The process takes a little effort, and if you think that it sounds hard, just wait until you try to use the Interceptor. There's a good reason why it wasn't made public... Not that it is a bad API or anything like that. It just requires a LOT of sweat. You'll probably do better to optimize your program and learn how to use DPS efficiently. Once you're doing that, Interceptor will give you at best about a 10% speed boost...for at least 50% more effort. In most cases, this is not at all worth it! -- Later, -Don Yacktman don@misckit.com <a href="http://www.misckit.com/don.html">My home page</a>
From: Christian Neuss <neuss.@informatik.th-darmstadt.de.nos-pam> Newsgroups: comp.sys.next.programmer Subject: Re: libInterceptor Date: 7 Aug 1998 10:00:45 GMT Organization: Technische Universitaet Darmstadt Message-ID: <6qej8d$n5a$2@sun27.hrz.tu-darmstadt.de> References: <35ca5dff.0@news.msu.ru> <6qe9qn$24e$1@news.xmission.com> don@misckit.com (Don Yacktman) wrote: >"Dmitry A. Yevsukhin" <eda@next.msu.ru> wrote: >> Has anybody ever seen libInterceptor_s.a library and >> corresponding .h files? ... >The process takes a little effort, and if you think that it sounds hard, just >wait until you try to use the Interceptor. There's a good reason why it >wasn't made public... Not that it is a bad API or anything like that. It >just requires a LOT of sweat. You'll probably do better to optimize your >program and learn how to use DPS efficiently. Once you're doing that, >Interceptor will give you at best about a 10% speed boost...for at least 50% >more effort. In most cases, this is not at all worth it! Strongly seconded. May I suggest that you search the USENET archives for postings on Interceptor and efficient DPS programming? For a well written DPS program, the speed gained by Interceptor is not that high. NXBitmapImageRep can be blitted very fast - just make sure you create them in the correct color depth, so they do not need to be converted. Rgds, Chris -- // Christian Neuss "static typing? how quaint.." // http://www.nexttoyou.de/~neuss/ // fax: (+49) 6151 16 5472
From: Jamie Hogg <jwh100@york.ac.uk> Newsgroups: comp.sys.mac.programmer.tools,comp.sys.newton.programmer,comp.sys.next.programmer,comp.unix.programmer Subject: Computer music Prototype Survey Date: Fri, 07 Aug 1998 12:36:26 +0100 Organization: University of York Sender: jwh100@york.ac.uk Message-ID: <35CAE6BA.C399AD3E@york.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi, my name is Jamie Hogg. I am researching a project for an MSc in Music Technology at York University. I'm designing and building a GUI for a set of sound synthesis tools, but am trying to make the software intuitive and easy to use. I have 2 surveys, one is about image icons, the other shows my prototype interface and asks for comments. If you have time, I'd really appreciate it if you could spare me 10 minutes to visit my website and have a go. (Sorry if you tried before and found that you were denied access - oops! I've fixed this now.) Thanks very much! http://www.york.ac.uk/~jwh100/Survey/SFront.htm
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.mac.programmer.tools,comp.sys.newton.programmer,comp.sys.next.programmer,comp.unix.programmer Subject: cmsg cancel <35CAE6BA.C399AD3E@york.ac.uk> Control: cancel <35CAE6BA.C399AD3E@york.ac.uk> Date: 07 Aug 1998 11:51:17 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.35CAE6BA.C399AD3E@york.ac.uk> Sender: jwh100@york.ac.uk Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
Message-ID: <35CB332B.FFF8E8BA@nospam.com> From: Tim Triemstra <nospam@nospam.com> Organization: PM Global Foods, LLC MIME-Version: 1.0 Newsgroups: gnu.gnustep.discuss,comp.sys.next.programmer,comp.sys.next.advocacy,comp.os.linux.development Subject: Re: Another day, another GNUstep screenshot... References: <6q6n11$5bs@sjx-ixn10.ix.netcom.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Fri, 07 Aug 1998 14:01:35 GMT NNTP-Posting-Date: Fri, 07 Aug 1998 10:01:35 EDT Felipe A. Rodriguez wrote: > > For all those curious to see how the GNUstep project is progressing > I've posted a screenshot at: > > http://pweb.netcom.com/~far/far.html > > In the foreground you'll see the XRAW Workspace example which is > a GNUstep clone of the NeXTSTEP Workspace. And behind it is the > very new Edit example. I'm sure (hope) that someone has looked into this, but I asked lawyers at NeXTSTEP once upon a time if they minded me using their icons in one of my apps because I wanted my NeXT users to feel comfortable with the app. They told me a definite "no" to using any of their icons in my apps or on a web page. These icons in the screenshot seem the same as the NeXTSTEP ones, so is there GNU version of these icons I could use? Just curious because I still have NeXTSTEP users I'd like to make at home in another environment (not to mention they're purty :) -- Tim Triemstra . TimT at PMGLOBAL dot COM PM Global Foods, LLC ... Atlanta GA USA Java Interface Project: http://www.mindspring.com/~timtr/
From: patw@proserv.wustl.edu Newsgroups: comp.sys.next.programmer Subject: !!! Announcing an important information resource The Meta-List Date: 7 Aug 1998 14:14:18 GMT Organization: Washington University in St. Louis Message-ID: <6qf23q$73j$639@newsreader.wustl.edu> Hi! I posted this using an unregistered copy of Newsgroup AutoPoster PRO! See a new site on the CAIT Home Page geared towards meeting the needs of the Information Systems Specialist. INFORMATION SYSTEMS META-LIST http://www.cait.wustl.edu Any additions or recomendations to the Information Systems Meta-List are greatly appreciated.
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6qf23q$73j$639@newsreader.wustl.edu> Control: cancel <6qf23q$73j$639@newsreader.wustl.edu> Date: 07 Aug 1998 14:14:25 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6qf23q$73j$639@newsreader.wustl.edu> Sender: patw@proserv.wustl.edu Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: patw@proserv.wustl.edu Newsgroups: comp.sys.next.programmer Subject: !!! Announcing an important information resource The Meta-List Date: 7 Aug 1998 15:44:50 GMT Organization: Washington University in St. Louis Message-ID: <6qf7di$7hq$1055@newsreader.wustl.edu> Hi! I posted this using an unregistered copy of Newsgroup AutoPoster PRO! See a new site on the CAIT Home Page geared towards meeting the needs of the Information Systems Specialist. INFORMATION SYSTEMS META-LIST http://www.cait.wustl.edu Any additions or recomendations to the Information Systems Meta-List are greatly appreciated.
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6qf7di$7hq$1055@newsreader.wustl.edu> Control: cancel <6qf7di$7hq$1055@newsreader.wustl.edu> Date: 07 Aug 1998 18:30:15 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6qf7di$7hq$1055@newsreader.wustl.edu> Sender: patw@proserv.wustl.edu Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
Newsgroups: comp.sys.next.programmer From: clintdw@netcom.com (Clinton Wong) Subject: Re: semaphores on NeXT? Message-ID: <clintdwExC3H4.C2w@netcom.com> Organization: Netcom On-Line Services References: <clintdwExB1Mq.6qA@netcom.com> <6qefai$j8u$1@news.tudelft.nl> Date: Fri, 7 Aug 1998 19:29:28 GMT Sender: clintdw@netcom12.netcom.com guyt@is.twi.tudelft.nl (Abraham Guyt) writes: >Clinton Wong writes >> I'm looking through a NeXTStep programming manual and see mutex_init(), >> mutex_lock(), mutex_unlock(), and cthread_detach(). But I don't >> see anything about sempahores. Does such a thing exist? Or perhaps >> there are third party libraries that I could use that implement >> semaphores with a mutex and variable? >> >> Any info would be deeply appreciated... >Before I elaborate, would exactly are you trying to achieve with these >semaphores ? I'm porting code that uses pthreads back into Mach cthreads... so my original question should have been stated like this: Does anyone know if there are NeXT equivalents to sem_post() and sem_wait()? If someone made a pthread library that maps back into cthreads, that would be great... just asking around. Clinton -- For a good time, read RFC 2324. http://postmaster.net/~clintdw/
From: "Laurent Cerveau" <cerveau@club-internet.fr> Newsgroups: comp.sys.next.programmer Subject: Network and BSD questions Date: Sat, 08 Aug 1998 00:02:33 +0000 Organization: Mmyne Message-ID: <6qfti8$oos@luc.ircam.fr> Mime-Version: 1.0 Content-Type: text/plain; charset="ISO-8859-1" Content-Transfer-Encoding: 7bit Hello, I'm new to Rhapsody (it' awesome!!!!) and would like to ask some questions. First I would like to know how to configure the machine in order to have a network with another Mac(having personnal WebSharing), to do some ftp. I haven't found much help for NetAdmin.app, and NetworkManager.app (How to choose the DNS and so on, do I have to use NetInfo ....). This question leads directly (at least for me) to another one concerning the use and knowledge of the BSD in Rhapsody. I really feel my lack of knowledge about Unix systems (although I've used them to for years), and the way to administrate a Unix machine (some questions like "where can I change the info for the machine directly in a file ?, how are the files organized and what changes brought NeXT?, how to build a database for command like 'locate' or 'apropos'?...). Does someone knows a good reference for a book for a kind of joe-average administrator like me? I apologize if this is not the right place to ask this kind of questions. Thanks Laurent Cerveau
From: "Dmitry A. Yevsukhin" <eda@next.msu.ru> Newsgroups: comp.sys.next.programmer References: <35ca5dff.0@news.msu.ru> <6qe9qn$24e$1@news.xmission.com> <6qej8d$n5a$2@sun27.hrz.tu-darmstadt.de> Subject: Re: libInterceptor Date: Sat, 8 Aug 1998 04:45:50 +0400 Message-ID: <35cba009.0@news.msu.ru> Christian Neuss wrote in message <6qej8d$n5a$2@sun27.hrz.tu-darmstadt.de>... >don@misckit.com (Don Yacktman) wrote: >>"Dmitry A. Yevsukhin" <eda@next.msu.ru> wrote: >>> Has anybody ever seen libInterceptor_s.a library and >>> corresponding .h files? >... > >>The process takes a little effort, and if you think that it sounds hard, >just >>wait until you try to use the Interceptor. There's a good reason why it >>wasn't made public... Not that it is a bad API or anything like that. It >>just requires a LOT of sweat. You'll probably do better to optimize your >>program and learn how to use DPS efficiently. Once you're doing that, >>Interceptor will give you at best about a 10% speed boost...for at least 50% >>more effort. In most cases, this is not at all worth it! > >Strongly seconded. May I suggest that you search the USENET archives >for postings on Interceptor and efficient DPS programming? For a well >written DPS program, the speed gained by Interceptor is not that high. > >NXBitmapImageRep can be blitted very fast - just make sure you create >them in the correct color depth, so they do not need to be converted. Thank you for replying. In my current project, I need to display live digital video at max frame rate. I have already implemented that using NXBitmapImageRep. The bitmap color depth matches that of the screen. If the DPS level only does some type/region checking, I will not get a noticeable speed boost with Interceptor, right? I wanted at least 50% boost. Regards Dmitry
From: gio+van+no+ni+8@tal+star+spam.com (Giovanni 8) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 8 Aug 1998 02:45:01 GMT Organization: TalStar Communications Message-ID: <gio+van+no+ni+8-0708982256070001@dialup17.tlh.talstar.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <6pnlfm$ob$2@beta.qmw.ac.uk> <8F12A6803BCB46BB.69250C85F5006041.55EF65CE3294D5C9@library-proxy.airnews.net> <6qcct2$aie$1@beta.qmw.ac.uk> Cache-Post-Path: flux!unknown@dialup17.tlh.talstar.com > Matthew M. Huntbach wrote: > ...a useful test is what is used in practice, what the big > companies, who are trying to make money out of programming, > prefer. Why has industry gone for object-oriented programming > in a big way in recent years? The "real" reason is that it's another fad. Oh, sure, the fad won't sustain itself if it doesn't pay off before too long (and I believe it will or I wouldn't have devoted so much effort learning these things). But the B-school bozos at the aforementioned "big companies" don't have a clue. Look at the job ads, & you can see them jumping onto the latest IBM/Microsoft acronymic buzz. They're not exactly sure what it is, but it's from the big guys, & it's new, so they apparently think it must be the way to go. What I also see, in each new fad, is the seemingly willful ignoring of the areas in which each is weak. Rather than building new methods & theories on what worked before, we're doing a bit too much jumping about. There has been, for instance, very little effort to balance the need for artistry & creativity, with the use of proven tools & methods at each step. The real test is whether, in most cases, each new tool, method and theory works better than what was used before. "Thinking is the hardest work there is, which is probably the reason why so few engage in it." --- Eileen C. Shapiro 1995 _Fad Surfing in the BoardRoom_ pg vii "Managers as re-engineers assume that they can over-haul an organization as though it were a mechanical device that can be re-designed & re-built, from then on to run like a dream. Managers as coaches assume that they can motivate a whole organization to keep running the hurdles. Managers as quantum physicists assume that if they can set the right context, the leaps will take care of themselves -- though in a largely unpredictable way." --- Eileen C. Shapiro 1995 _Fad Surfing in the BoardRoom_ pg 194 -- copyright 1998 by g8 (exclusive of others' writing) all rights reserved except that license is given freely to respond on usenet. other licenses are available at reasonable rates.
From: gio+van+no+ni+8@tal+star+spam.com (Giovanni 8) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 8 Aug 1998 02:50:37 GMT Organization: TalStar Communications Message-ID: <gio+van+no+ni+8-0708982301460001@dialup17.tlh.talstar.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <tz87m0mnygn.fsf@aimnet.com> Cache-Post-Path: flux!unknown@dialup17.tlh.talstar.com > Thomas wrote: > What strikes you as user hostile may be the > most effective tool for someone who uses it daily. What is hostile to one user is often friendly to another. E.g. what to one is a nice, graphical interface may be a major barrier to understanding, flexibility, & control to another. "All that is necessary for a case to take forever is for a judge to be willing to let it take forever..." --- Alexander Williams 1986-10-14 (quoted in Jan Golab 1993 _The Dark Side of the Force_ pg 215) -- copyright 1998 by g8 (exclusive of others' writing) all rights reserved except that license is given freely to respond on usenet. other licenses are available at reasonable rates.
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: Calling a routine in a Windows DLL Date: 7 Aug 98 10:36:35 Organization: Is a sign of weakness Message-ID: <SCOTT.98Aug7103635@slave.doubleu.com> References: <35CA2F18.19B1DDD3@qualcomm.com> In-reply-to: Michael Simpson's message of Thu, 06 Aug 1998 15:32:56 -0700 In article <35CA2F18.19B1DDD3@qualcomm.com>, Michael Simpson <t_msimps@qualcomm.com> writes: Does anyone have an example of how to call a routine in a Window's DLL from objective C? From VC++, I would use the import lib and then just make the call. I'm just not certain how to set it up in Project Builder. I've looked through the docs and see references to it, but not specifics. Any help would be appreciated. Check out http://www.doubleu.com/ZLibTest.zip. This is an example of using the zlib dll (zip/unzip library) from an Objective-C program. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: holger@_REMOVE_THIS_.wizards.de (Holger Hoffstaette) Newsgroups: comp.sys.next.programmer Subject: Re: semaphores on NeXT? Date: 8 Aug 1998 10:32:35 GMT Organization: The secret circle of the NSRC Message-ID: <6qh9g3$2fj@ragnarok.en.uunet.de> References: <clintdwExB1Mq.6qA@netcom.com> <6qefai$j8u$1@news.tudelft.nl> <clintdwExC3H4.C2w@netcom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit NNTP-Posting-Date: 8 Aug 1998 10:32:12 GMT Clinton Wong wrote: > guyt@is.twi.tudelft.nl (Abraham Guyt) writes: > > >Clinton Wong writes > >> I'm looking through a NeXTStep programming manual and see mutex_init(), > >> mutex_lock(), mutex_unlock(), and cthread_detach(). But I don't > >> see anything about sempahores. Does such a thing exist? Or perhaps > >> there are third party libraries that I could use that implement > >> semaphores with a mutex and variable? > >> > >> Any info would be deeply appreciated... > > >Before I elaborate, would exactly are you trying to achieve with these > >semaphores ? > > I'm porting code that uses pthreads back into Mach cthreads... so my > original question should have been stated like this: > > Does anyone know if there are NeXT equivalents to sem_post() and sem_wait()? > If someone made a pthread library that maps back into cthreads, > that would be great... just asking around. Your posting reminded me that I always wanted to fix a pthread library that was ported by Hutchinson Software (HASC) to NS 3.3 a couple of years ago. Apart from a little makefile munging an fixing an asm opcode (hopefully correctly) it seems to work; the test suite seems to pass. Let me know if you want to look at it. A potential problem might be that it comes with thread-safe implementations of many stdio/stdlib functions and that raises some conflicts with the System.framework. Holger
From: Trevin Beattie <*trevin*@*xmission*.*com*> Newsgroups: comp.sys.next.programmer Subject: Bug in readdir() function in NS3.3 -posix library Date: Sat, 08 Aug 1998 06:36:56 -0600 Organization: XMission http://www.xmission.com/ Message-ID: <6qhgpa$cot$1@news.xmission.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NNTP-Posting-Date: 8 Aug 1998 12:36:58 GMT I recently came across a bug in Gnu make (3.76.1) in which wildcards in targets and dependents weren't being globbed properly. I thought rebuilding it with the -posix flag would be the easiest way to fix it, but that showed another problem: getcwd() was returning NULL with errno set to ENOENT! The same call works perfectly fine without the -posix flag. I traced the code and discovered that when reading the root directory with readdir(), several directory entries were missing! Specifically, the missing names were mount points for other partitions on the hard drive (so the problem doesn't manifest itself when you're still on sd0a). A simple "cat /" shows that mount points don't really exist in the root directory, so I can only assume the mount points are only kept in kernel memory at runtime. Further investigation shows that getdirentries() properly returns all real entries and mount points with the -posix flag as well as without, so I am able to implement a working getcwd() call for those programs which require -posix. -- Trevin Beattie "Do not meddle in the affairs of wizards, *To*reply*to*this* for you are crunchy and good with ketchup." *message,*remove*the* --unknown *asterisks*from*my*email*address.*
From: x@invalid.zeta.org.au (Richard RUDEK) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: Sat, 08 Aug 1998 13:52:01 GMT Organization: Zeta Internet, http://www.zeta.org.au/ Message-ID: <35cc564c.9311481@news.zeta.org.au> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <6pnlfm$ob$2@beta.qmw.ac.uk> <8F12A6803BCB46BB.69250C85F5006041.55EF65CE3294D5C9@library-proxy.airnews.net> <6ppj8r$bcs$1@crib.corepower.com> <6qcdb2$aie$2@beta.qmw.ac.uk> <tcny1.729$ML4.2269662@news4.mia.bellsouth.net> "Judson McClendon" <judmc123@bellsouth.net> wrote: >Matthew M. Huntbach wrote: >> >>Nevertheless, in my many years of using the net, I do find that there's a >>sort of simple-minded "My country is the best and all the others stink" >>nationalism that is not uncommon amongst American posters, but hardly ever >>found amongst those of other nationalities. > >I'll have to take issue with that particular point, Matthew. My experience >is that people from all nationalities are pretty much the same in this >regard. Sure, people are proud of their own countries. I don't blame them >for that. Anyone who is not should leave and go somewhere they believe is >better. What provokes me is unfair criticism, whether directed at other's >countries, or their own. > >It is true that Americans can be arrogant and obnoxious. But at the same >time, I believe America gets a lot of unfair criticism. No other country >has sacrificed so much for others as has the U.S., or been more generous >to its friends, and enemies. There is no parallel in history to the way >the U.S. has helped rebuild former enemies, like Japan and Germany. Every [snip] Hmm, yes. But I think it has something to do with the crowing many do before, while, and after they help. I was just watching snippets from a Seinfeld episode, where George would collect lunch for his Boss, and the lunch shop had a container to collect tips/donations. George wasn't happy to just give money, he had to know that the counter staff knew what, and how much he gave. He waved the bill around, then as he placed the tip, the counter person was distracted. George then tried to fish the money out so he could stage the event again, but was caught in the act, and then expelled from the shop... This may indeed be the case of a vocal minority, but, well, I dunno... __ __ _______________________________ //)) //)) | Richard RUDEK. MicroDek. | Insert sh, ah, //\\ //\\ | Chatswood, Sydney. Australia. | witty comment here. :) `-------------------------------'
From: x@invalid.zeta.org.au (Richard RUDEK) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Program99, was (Re: Programming and the colossal failure to advance) Date: Sat, 08 Aug 1998 13:52:03 GMT Organization: Zeta Internet, http://www.zeta.org.au/ Message-ID: <35cc5707.9498586@news.zeta.org.au> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <35A0A008.41C6@pc-plus.de.remove.this.bit><01bdac0f$f6d84f20$0fd29cce@kiki.graytechnologies.com> <SCOTT.98Jul10125032@slave.doubleu.com> <01bdb0c4$801859c0$0fd29cce@kiki.graytechnologies.com> <SCOTT.98Jul16105128@slave.doubleu.com> scott@nospam.doubleu.com (Scott Hess) wrote: >In article <01bdb0c4$801859c0$0fd29cce@kiki.graytechnologies.com>, > "Stu" <stu@grayechnologies.com> writes: > Scott Hess <scott@doubleu.com> wrote in article > <SCOTT.98Jul10125032@slave.doubleu.com>... > > In article <01bdac0f$f6d84f20$0fd29cce@kiki.graytechnologies.com>, > > "Stu" <stu@grayechnologies.com> writes: [snip] >The problem is that Microsoft is a moving target, and everyone wants >to be compatible with Microsoft. The fact that Microsoft's formats >change with the wind is _not_ necessary. If they wanted to, they >could use a format which would be verbose enough to let them describe >anything they want, without being so terse that minor changes break >everyone else's products. Part of what annoys me about all this is >that there is no technical reason why things are the way they are - >it's all competitiveness and marketting. Um, having looked casually at the MS Word 97 "file format", I would conclude that the "file format" is basically a collection of serialised objects. That is, the file created (OLE 2.0 docfile) is basically a packaged, data memory dump. I suppose this was to save Microsoft the cost of an interpreter/translator. But because of this, I don't think MS can accurately specify the semantics, as it's "hard-coded" into the particular version of Word. Non-Disclosure license, anyone ? This "saving" I suspect will be short lived... I know many people that bitch about MS "file format" cross-compatibility, even among MS's own (earlier) products ? __ __ _______________________________ //)) //)) | Richard RUDEK. MicroDek. | Insert sh, ah, //\\ //\\ | Chatswood, Sydney. Australia. | witty comment here. :) `-------------------------------'
From: "Michael A. Thompson" <mat0001@jove.acs.unt.edu> Newsgroups: comp.sys.next.programmer Subject: audio programming Date: Sun, 09 Aug 1998 21:47:09 -0500 Organization: University of North Texas Message-ID: <35CE5F2D.D56DA946@jove.acs.unt.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I just came into a NeXT at home a while back and wonder if anybody knows of a good book or web page for audio application programming on the NeXT.... My NeXT has NS3.3 and 3.3 Developer.... Thanks, Michael -- ---------------------------------- Michael A. Thompson Unix SysAdmin. [IRIX - NeXTStep - Linux] University of North Texas Center for Experimental Music and Intermedia [C.E.M.I.] Office: (940) 565-2382 E-Mail: mat0001@jove.acs.unt.edu ----------------------------------
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <345241151569.4834697028@hotmail.com> Control: cancel <345241151569.4834697028@hotmail.com> Date: 10 Aug 1998 05:36:30 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.345241151569.4834697028@hotmail.com> Sender: joeblow8@hotmail.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Eric Preyss <epreyss@aaii.com.au> Newsgroups: comp.sys.next.programmer Subject: 2-button mouse Date: Mon, 10 Aug 1998 16:44:10 +1000 Organization: Australian Artificial Intelligence Institute Message-ID: <35CE96BA.4DB97F77@aaii.com.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cache-Post-Path: aloomba.aaii.oz.au!unknown@doc-pc I've just purchased a G3 Power Mac for running Rhapsody as well as MacOS w/ Virtual PC/NT. Can anyone recommend a good two-button mouse? Eric Preyss Melbourne, Australia epreyss@aaii.com.au
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <6522902030423@digifix.com> Date: 9 Aug 1998 03:48:30 GMT Organization: Digital Fix Development Message-ID: <11760902635220@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1994. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: mmh@dcs.qmw.ac.uk (Matthew M. Huntbach) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Followup-To: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Date: 10 Aug 1998 09:30:33 GMT Organization: Queen Mary & Westfield College, London, UK Message-ID: <6qmejp$rni$3@beta.qmw.ac.uk> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <6pnlfm$ob$2@beta.qmw.ac.uk> <8F12A6803BCB46BB.69250C85F5006041.55EF65CE3294D5C9@library-proxy.airnews.net> <6qcct2$aie$1@beta.qmw.ac.uk> <gio+van+no+ni+8-0708982256070001@dialup17.tlh.talstar.com> Giovanni 8 (gio+van+no+ni+8@tal+star+spam.com) wrote: > > Matthew M. Huntbach wrote: > > ...a useful test is what is used in practice, what the big > > companies, who are trying to make money out of programming, > > prefer. Why has industry gone for object-oriented programming > > in a big way in recent years? > The "real" reason is that it's another fad. Sure, I don't think OOP is the answer to everything, and I agree it has been massively hyped in recent years. But it is one way to try and impose a structure on programs and make them more comprehensible and easier to update. I was arguing against someone who thought the very idea of being concerned about the structure of a program and making sure it was not just hacked together any old was so bad as to warrant being met with nationalistic abuse. Matthew Huntbach
From: "InterPoint" <interpoint@centrenet.co.uk> Subject: GERMANY - OPENSTEP JOBS Newsgroups: comp.sys.next.programmer Message-ID: <01bdc447$acfc9b40$729249c2@cna40.centrenet.co.uk> Date: Mon, 10 Aug 1998 10:11:41 GMT NNTP-Posting-Date: Mon, 10 Aug 1998 11:11:41 BST Our client, a leading European Next/Openstep consultancy, is urgently seeking consultants at all levels of experience for major blue-chip projects in Germany. You should have experience in at least one of… Next/Openstep Web Objects Enterprise Objects A background in OO methodology would be a strong plus. Our client offers an excellent remuneration package including stock options, structured public & private health insurance in addition to a highly competitive salary & a friendly, professional & team-oriented corporate culture. German language skills are desirable but not essential. …please mail us in complete confidence at interpoint@centrenet.co.uk +44 411-262359
From: balazs.pataki@sztaki.hu (Balazs Pataki) Newsgroups: comp.sys.next.programmer Subject: Compiling X programs under NEXTSTEP Date: 10 Aug 1998 14:37:08 GMT Organization: Computer and Automation Institute Message-ID: <6qn0ik$db6@lutra.sztaki.hu> Hi, is there any way to compile X programs under NEXTSTEP (3.x, maybe OS 4.x)? I found out that Xnext has some X libraries (.a files) in the distirbution. Are those usable? Any help appreciated, --- balazs
Message-ID: <35CF9962.2AD93539@home.com> From: msms <msms@home.com> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Re: Network and BSD questions References: <6qfti8$oos@luc.ircam.fr> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Tue, 11 Aug 1998 01:06:12 GMT NNTP-Posting-Date: Mon, 10 Aug 1998 18:06:12 PDT Organization: @Home Network For network configuration, you just need to go into your preferences.app and select network. The fields should be self-explanatory. Make sure both machines are on the same subnet (i.e., 192.168.1.x, where "x" would be the only difference between the machines). Rhapsody will have ftpd running by default, so you should be able to ftp right into Rhapsody. Starting the web server apache is configurable in the file hostconfig (in /etc) I believe, where you just put -YES- after the webserver field. I don't recall where the default web serving directory is, but it's only a few levels down so you should find it no problem. If you want to learn more about UNIX administration, you should probably buy a book on it. There's a lot to learn, but it's worth it I think. These types of questions should probably be posted though to comp.sys.next.sysadmin or com.sys.next.software. -- John ---------------------- Laurent Cerveau wrote: > > Hello, > > I'm new to Rhapsody (it' awesome!!!!) and would like to ask some questions. > > First I would like to know how to configure the machine in order to have a > network with another Mac(having personnal WebSharing), to do some ftp. > I haven't found much help for NetAdmin.app, and NetworkManager.app (How to > choose > the DNS and so on, do I have to use NetInfo ....). > > This question leads directly (at least for me) to another one concerning the > use > and knowledge of the BSD in Rhapsody. I really feel my lack of knowledge > about Unix > systems (although I've used them to for years), and the way to administrate > a Unix > machine (some questions like "where can I change the info for the machine > directly > in a file ?, how are the files organized and what changes brought NeXT?, how > to build > a database for command like 'locate' or 'apropos'?...). Does someone knows > a good > reference for a book for a kind of joe-average administrator like me? > > I apologize if this is not the right place to ask this kind of questions. > > Thanks > > Laurent Cerveau
From: "Charles W. Swiger" <chuck@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: Compiling X programs under NEXTSTEP Date: 10 Aug 1998 17:32:26 GMT Organization: Spacelab.net Internet Access Message-ID: <6qnara$11g$1@news.spacelab.net> References: <6qn0ik$db6@lutra.sztaki.hu> balazs.pataki@sztaki.hu (Balazs Pataki) wrote: >is there any way to compile X programs under NEXTSTEP (3.x, maybe OS 4.x)? >I found out that Xnext has some X libraries (.a files) in the >distirbution. Are those usable? Yes, they are. A long time back, I compiled a few X apps under Xnext. You'll almost certainly have to do some porting to make the code go.... -Chuck Charles Swiger | chuck@codefab.com | standard disclaimer ---------------+-------------------+-------------------- "Microsoft: we make the easy almost impossible."
From: jurriaan@no.spam.fygir.nl (Jurriaan van der Lingen) Newsgroups: comp.sys.next.programmer Subject: Fetching records with EO Date: 11 Aug 1998 16:19:48 GMT Organization: NLnet Message-ID: <6qpqv4$g6i$1@news.Leiden.NL.net> Hi, I'm trying to use EO to fetch records from a data base. Can anyone explain me what I'm doing wrong in the following piece of code? It's supposed to get all records from a DB for an entity and simply prints an attribute value for each record. The method does what it's supposed to do if you execute it once. However, if a value in the database is changed (from another source) and I execute the method again, I still get the old values ! It seems there's a snapshot cached somewhere in EO's dungeons that I can't free...? /** START CODE SAMPLE **/ - (void) printAttribute:(NSString*) attributeName forRecordsInEntityNamed:(NSString*) entityName inModel:(EOModel*) eoModel { EOEditingContext* context; EOFetchSpecification* fetchSpecification; NSArray* fetchedRecords; NSEnumerator* enumerator; EOGenericRecord* record; /* Make a new context */ context = [EOEditingContext new]; /* Make a new fetch specification, to fetch "all records" */ fetchSpecification = [EOFetchSpecification new]; [fetchSpecification setEntityName: entityName]; /* Force refetch... doesn't work? */ [fetchSpecification setRefreshesRefetchedObjects: YES]; /* Fetch the record objects */ fetchedRecords = [context objectsWithFetchSpecification:fetchSpecification]; enumerator = [fetchedRecords objectEnumerator]; /* Print the record attributes */ while( record = [enumerator nextObject] ) NSLog( @"%@", [record valueForKey:attributeName] ); /* Get rid of all caches... doesn't work? */ [[context rootObjectStore]invalidateAllObjects]; /* Release stuff */ [fetchSpecification release]; [context release]; } /** END CODE SAMPLE **/
From: "Charles W. Swiger" <chuck@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: Network and BSD questions Date: 11 Aug 1998 16:31:02 GMT Organization: Spacelab.net Internet Access Message-ID: <6qprk6$5nb$1@news.spacelab.net> References: <6qfti8$oos@luc.ircam.fr> <35CF9962.2AD93539@home.com> msms <msms@home.com> wrote: [ ... ] >Rhapsody will have ftpd running by default, so you should >be able to ftp right into Rhapsody. Not quite. Rhapsody's inetd is set up by default to spawn ftpd's if an ftp request is made to the machine, but it does not run an FTP server continuously. As for good sysadmin references, the O'Reilly "Essential System Administration" and the "TCP/IP Network Administration" are two bibles... -Chuck Charles Swiger | chuck@codefab.com | standard disclaimer ---------------+-------------------+-------------------- "Microsoft: we make the easy almost impossible."
From: Michael Simpson <t_msimps@qualcomm.com> Newsgroups: comp.sys.next.programmer Subject: Which NIB Date: Tue, 11 Aug 1998 14:04:20 -0700 Organization: Qualcomm Message-ID: <35D0B1D4.1845008C@qualcomm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I'm trying to work through the CurrencyConverter on the Intel DR2. The instructions say to modify CurrencyConverter.nib, yet the app tries to load CurrencyConverter-windows.nib instead. My currency window does not get shown but the blank window in the other nib instead. The Developer tutorial is obviously for the DR2 Apple version. Also, in the code line: total = [converter convertAmount:amt byRate:rate]; I understand that converter is an object, convertAmount is a method and amt is a parameter. Why the "byRate"? What does this accomplish? Why not just a syntax [converter convertAmount: amt rate]; I'm coming from a C++ view so this syntax seems a bit strange to me. Thanks, Michael
From: gio+van+no+ni+8@tal+star+spam.com (Giovanni 8) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 11 Aug 1998 22:10:53 GMT Organization: TalStar Communications Message-ID: <gio+van+no+ni+8-1108981815550001@dialup97.tlh.talstar.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <6pnlfm$ob$2@beta.qmw.ac.uk> <8F12A6803BCB46BB.69250C85F5006041.55EF65CE3294D5C9@library-proxy.airnews.net> <6qcct2$aie$1@beta.qmw.ac.uk> <gio+van+no+ni+8-0708982256070001@dialup17.tlh.talstar.com> <6qmejp$rni$3@beta.qmw.ac.uk> Cache-Post-Path: flux!unknown@dialup97.tlh.talstar.com > Matthew M. Huntbach wrote: >> Giovanni 8 wrote: >>> Matthew M. Huntbach wrote: >>> ...Why has industry gone for object-oriented programming >>> in a big way in recent years? >> The "real" reason is that it's another fad... > Sure, I don't think OOP is the answer to everything, and I agree > it has been massively hyped in recent years. But it is one way > to try and impose a structure on programs and make them more > comprehensible and easier to update. I was arguing against > someone who thought the very idea of being concerned about the > structure of a program and making sure it was not just hacked > together any old was so bad as to warrant being met with > nationalistic abuse. I pretty much agree. It is interesting that we think of it as "imposing" a structure rather than finding the structure or having it unfold. It's almost as though we're striving to create a struggle. "Think boldly, argue clearly, & have fun. You will know when you are making progress when you suddenly stop shouting long enough to realize that you are in violent agreement." --- David A. Taylor 1995 _Business Engineering with Object Technology_ pg 67 "[M]any of the RISC concepts had been articulated for even earlier high-performance systems, beginning at least with the CDC 6600, developed in the early 1960s." --- Shlomo Weiss & James E. Smith 1994 _POWER & PowerPC_ pg 3 -- copyright 1998 by g8 (exclusive of others' writing) all rights reserved except that license is given freely to respond on usenet. other licenses are available at reasonable rates.
From: Arnaud <debayeux@easynet.fr> Newsgroups: comp.sys.next.programmer Subject: How to do a screen capture ? Date: Wed, 12 Aug 1998 00:18:54 +0200 Organization: [posted via] Easynet France Message-ID: <35D0C34E.E6DFDF3B@easynet.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello, With Openstep 4.2, how do you copy a rect from the screen (with readimage) ? How is it done in Grab.app ? Thanks, Arnaud
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: Which NIB Date: 11 Aug 1998 22:35:30 GMT Organization: Idiom Communications Message-ID: <6qqgvi$964$3@news.idiom.com> References: <35D0B1D4.1845008C@qualcomm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: t_msimps@qualcomm.com Michael Simpson may or may not have said: -> I'm trying to work through the CurrencyConverter on the Intel DR2. The -> instructions say to modify CurrencyConverter.nib, yet the app tries to -> load CurrencyConverter-windows.nib instead. My currency window does not -> get shown but the blank window in the other nib instead. The Developer -> tutorial is obviously for the DR2 Apple version. -> -> Also, in the code line: -> -> total = [converter convertAmount:amt byRate:rate]; -> -> I understand that converter is an object, convertAmount is a method and -> amt is a parameter. Why the "byRate"? What does this accomplish? Why -> not just a syntax [converter convertAmount: amt rate]; Well, the method name is actually "convertAmount:byRate:" Colons in a method name are significant, and they tell the compiler that a parameter follows. This is taken from the Smalltalk language. The upshot of this, is that you can make your method calls easier to understand by (in effect) naming your parameters. When the method only has one or two parameters the benefit of this style may not be apparent, but consider this method: - (id)initWithBitmapDataPlanes: (unsigned char **)planes pixelsWide: (int) width pixelsHigh: (int) height bitsPerSample: (int) bps samplesPerPixel: (int) spp hasAlpha: (BOOL) alpha isPlanar: (BOOL) isPlanar colorSpaceName: (NSString *) colorSpaceName bytesPerRow: (int) rowBytes bitsPerPixel: (int) pixelBits If you had to remember the order of those parameters, so your method call looked like myBitMap->initBitMap(planes, width,height, bps,spp,alpha,isPlanar, colorSpaceName, rowBytes, pixelBits) That is: myBitMap->initBitMap(&myPixelData, 50, 50, 8,3,YES,YES, NSRGBColorSpace, 512, 24); Then it can be awfully confusing when you go to read the code later. -> I'm coming from a C++ view so this syntax seems a bit strange to me. Yeah, it looked funny to me too, when I switched to Obj-C seven years ago. These days, I wouldn't touch C++ with a ten foot pole. Too much complexity for too little benefit. -jcr -- "Although UNIX is more reliable, NT may become more reliable with time" - Ron Redman, deputy technical director of the Fleet Introduction Division of the Aegis Program Executive Office, US Navy. Where is the line between mere incompetence and actual treason? -jcr
From: quinlan@intergate.bc.ca Newsgroups: comp.sys.next.programmer Subject: Re: Which NIB Date: Wed, 12 Aug 1998 00:59:04 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6qqpcp$vgv$1@nnrp1.dejanews.com> References: <35D0B1D4.1845008C@qualcomm.com> In article <35D0B1D4.1845008C@qualcomm.com>, Michael Simpson <t_msimps@qualcomm.com> wrote: > Also, in the code line: > > total = [converter convertAmount:amt byRate:rate]; > > I understand that converter is an object, convertAmount is a method and > amt is a parameter. Why the "byRate"? What does this accomplish? Why > not just a syntax [converter convertAmount: amt rate]; This is the same syntax that Smalltalk uses for method calls. This way, each parameter is named and the code is easier to read. For example, which is easier to understand: [window init: 500 500 true false] or [window initWithWidth: 500 height: 500 closeBox: true dirty: false] -- Brian Quinlan quinlan@intergate.bc.ca -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: nurban@crib.corepower.com (Nathan Urban) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: NEXTstep/Openstep for the Mac Date: 12 Aug 1998 09:30:48 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6qs5e8$i06$1@crib.corepower.com> References: <35CF0C69.B3B0CF5@ericsson.com> <35d05ffe.0@news.depaul.edu> <6qq18i$qac$1@nnrp1.dejanews.com> <SCOTT.98Aug11143008@slave.doubleu.com> In article <SCOTT.98Aug11143008@slave.doubleu.com>, scott@nospam.doubleu.com (Scott Hess) wrote: > In article <6qq18i$qac$1@nnrp1.dejanews.com>, > spagiola@my-dejanews.com writes: > Just wondering if it would be possible, and legal, to write a > Rhapsody app to access the Webster data files that came with > NeXTSTEP, and then simply transfer those files? > If you aquired the files legitimately, sure. In effect, this implies > that you save your original distribution media, and don't have the > files installed on your NeXTSTEP machines. I'd be willing to do that whenever I make my permanent switch. However, I'm more interested in the "possible" part. Does anyone know how to decode the Webster data files so that a clone could be written??
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.programming,comp.sys.next.programmer,comp.lang.misc,comp.software-eng Subject: Re: Programming and the colossal failure to advance Date: 12 Aug 98 08:30:06 Organization: Is a sign of weakness Message-ID: <SCOTT.98Aug12083006@slave.doubleu.com> References: <6ng34a$j2g$1@nnrp1.dejanews.com> <359CBA1C.4E9F@brookes.ac.uk> <68404996E5AC263B.283CD62C06150477.4B2220654C9E342D@library-proxy.airnews.net> <6pnlfm$ob$2@beta.qmw.ac.uk> <8F12A6803BCB46BB.69250C85F5006041.55EF65CE3294D5C9@library-proxy.airnews.net> <6qcct2$aie$1@beta.qmw.ac.uk> <gio+van+no+ni+8-0708982256070001@dialup17.tlh.talstar.com> <6qmejp$rni$3@beta.qmw.ac.uk> <gio+van+no+ni+8-1108981815550001@dialup97.tlh.talstar.com> In-reply-to: gio+van+no+ni+8@tal+star+spam.com's message of 11 Aug 1998 22:10:53 GMT In article <gio+van+no+ni+8-1108981815550001@dialup97.tlh.talstar.com>, gio+van+no+ni+8@tal+star+spam.com (Giovanni 8) writes: [Someone else, but the attribution was unclear -scott:] > Sure, I don't think OOP is the answer to everything, and I agree > it has been massively hyped in recent years. But it is one way > to try and impose a structure on programs and make them more > comprehensible and easier to update. I pretty much agree. It is interesting that we think of it as "imposing" a structure rather than finding the structure or having it unfold. It's almost as though we're striving to create a struggle. I don't think it is that negative. People "impose" structure on everything - even things without any structure at all. It's just the way people work. Most programmers don't initially impose a structure because they want to force things - they impose it because they need to start somewhere, and the problem's actual structure just isn't obvious. The struggle comes later, after sunk costs have been ... sunk. Some programmers think "Dammit, I'm _right_, make the problem fit the solution." In my experience, it's more often management saying "But we already spent two weeks on a solution that works in some cases, make it work for the others." But I don't think those struggles are at all related to imposing structure, they're tied to a _specific_ situation. I expect programs to demand that their structure be revisited every year or two, more often at first. Even then, I've seen very few problems which really dictated the specifics of their structure. More often, the broad structure is very dependant on the problem, but you can make or break the project based on how you design the rest of the project. If there were a single clear solution to each problem, building those solutions wouldn't be _nearly_ the trial it's turned out to be. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: boehring@biomed.ruhr-uni-bochum.de (Daniel Boehringer) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: NEXTstep/Openstep for the Mac Date: 12 Aug 1998 16:32:24 GMT Organization: Ruhr-Universitaet Bochum, Rechenzentrum Message-ID: <6qsg2o$sr$1@sun579.rz.ruhr-uni-bochum.de> References: <35CF0C69.B3B0CF5@ericsson.com> <35d05ffe.0@news.depaul.edu> <6qq18i$qac$1@nnrp1.dejanews.com> <SCOTT.98Aug11143008@slave.doubleu.com> <6qs5e8$i06$1@crib.corepower.com> NNTP-Posting-Date: 12 Aug 1998 16:32:24 GMT nurban@crib.corepower.com (Nathan Urban) wrote: >In article <SCOTT.98Aug11143008@slave.doubleu.com>, scott@nospam.doubleu.com (Scott Hess) wrote: > >> In article <6qq18i$qac$1@nnrp1.dejanews.com>, >> spagiola@my-dejanews.com writes: > >> Just wondering if it would be possible, and legal, to write a >> Rhapsody app to access the Webster data files that came with >> NeXTSTEP, and then simply transfer those files? > >> If you aquired the files legitimately, sure. In effect, this implies >> that you save your original distribution media, and don't have the >> files installed on your NeXTSTEP machines. > >I'd be willing to do that whenever I make my permanent switch. However, >I'm more interested in the "possible" part. Does anyone know how to >decode the Webster data files so that a clone could be written?? NeXT webster is IXKit based. have a look at ftp://ftp.peanuts.org/peanuts/./NEXTSTEP/unix/text/Webster.a5.s.tar.gz an openstep replacement application for Librarian.app would be a nice place to put webster acces into. daniel
From: nurban@crib.corepower.com (Nathan Urban) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: NEXTstep/Openstep for the Mac Date: 12 Aug 1998 12:36:07 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6qsg9n$j66$1@crib.corepower.com> References: <35CF0C69.B3B0CF5@ericsson.com> <SCOTT.98Aug11143008@slave.doubleu.com> <6qs5e8$i06$1@crib.corepower.com> <6qsg2o$sr$1@sun579.rz.ruhr-uni-bochum.de> In article <6qsg2o$sr$1@sun579.rz.ruhr-uni-bochum.de>, boehring@biomed.ruhr-uni-bochum.de (Daniel Boehringer) wrote: > nurban@crib.corepower.com (Nathan Urban) wrote: > >Does anyone know how to > >decode the Webster data files so that a clone could be written?? > NeXT webster is IXKit based. Hmm. I doubt that's going to make it across to Rhapsody any time soon. Would it be in violation of license to write some sort of converter (under NEXTSTEP, using IXKit) to convert the data files to some other indexed format that a clone could use (as long as you already have Webster and use the data files only for private use, of course.) Maybe using AIAT?
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: Re: NEXTstep/Openstep for the Mac Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer References: <35CF0C69.B3B0CF5@ericsson.com> <SCOTT.98Aug11143008@slave.doubleu.com> <6qs5e8$i06$1@crib.corepower.com> <6qsg2o$sr$1@sun579.rz.ruhr-uni-bochum.de> <6qsg9n$j66$1@crib.corepower.com> Message-ID: <35d1caf2.0@news.depaul.edu> Date: 12 Aug 98 17:03:46 GMT In comp.sys.next.advocacy Nathan Urban <nurban@crib.corepower.com> wrote: > In article <6qsg2o$sr$1@sun579.rz.ruhr-uni-bochum.de>, boehring@biomed.ruhr-uni-bochum.de (Daniel Boehringer) wrote: > > nurban@crib.corepower.com (Nathan Urban) wrote: > > >Does anyone know how to > > >decode the Webster data files so that a clone could be written?? > > NeXT webster is IXKit based. > Hmm. I doubt that's going to make it across to Rhapsody any time soon. > Would it be in violation of license to write some sort of converter (under > NEXTSTEP, using IXKit) to convert the data files to some other indexed > format that a clone could use (as long as you already have Webster and > use the data files only for private use, of course.) Maybe using AIAT? It would be easier to just use raw dictionary content from www.dict.org. - Jon -- Note: email to this address goes to /dev/null To email a reply, write to jon at exnext dot com
From: spammers@ruin.the.internet.channelu.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: NEXTstep/Openstep for the Mac Date: 12 Aug 1998 17:52:43 GMT Organization: Michigan State University Message-ID: <6qskpb$lir$1@msunews.cl.msu.edu> References: <35CF0C69.B3B0CF5@ericsson.com> <SCOTT.98Aug11143008@slave.doubleu.com> <6qs5e8$i06$1@crib.corepower.com> <6qsg2o$sr$1@sun579.rz.ruhr-uni-bochum.de> <6qsg9n$j66$1@crib.corepower.com> <35d1caf2.0@news.depaul.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: jhendry@shrike.depaul.edu In <35d1caf2.0@news.depaul.edu> Jonathan W Hendry wrote: > In comp.sys.next.advocacy Nathan Urban <nurban@crib.corepower.com> wrote: > > In article <6qsg2o$sr$1@sun579.rz.ruhr-uni-bochum.de>, boehring@biomed.ruhr-uni-bochum.de (Daniel Boehringer) wrote: > > > > nurban@crib.corepower.com (Nathan Urban) wrote: > > > > >Does anyone know how to > > > >decode the Webster data files so that a clone could be written?? > > > > NeXT webster is IXKit based. > > > Hmm. I doubt that's going to make it across to Rhapsody any time soon. > > Would it be in violation of license to write some sort of converter (under > > NEXTSTEP, using IXKit) to convert the data files to some other indexed > > format that a clone could use (as long as you already have Webster and > > use the data files only for private use, of course.) Maybe using AIAT? > > It would be easier to just use raw dictionary content from www.dict.org. > It may be better to go with a open solution. IXKit is one of the things that I am most interested in porting - but I don't think I will get around to it for quite some time. Also I hear from Don that the code is quite hairy. It might be better to look at developments in that arena and incorporate them into a new IXKit - or roll a completely new one. I doubt that raw dictionary could do all the things IXKit did - and there are improvements to the front ends I would like to make. Hell I can't see why IXKit files couldn't be used to feed a IXKit based webbot that just groks IXKit files and adds them to it's archives rather than traversing trees.. Whaddya think? One thing I really love is Digital Librarian - Makes finding files on 50G of 1.3G Optical disks quick and fun. Fricken fast too IMHO. Randy rencsok at channelu dot com argus dot cem dot msu dot edu spammers works also :) Randy Rencsok General UNIX, NeXTStep, IRIX Admining, Turbo Software Consulting, Programming, etc.)
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6qs1ko$jm7$3506@artemis.backbone.ou.edu> Control: cancel <6qs1ko$jm7$3506@artemis.backbone.ou.edu> Date: 12 Aug 1998 21:40:23 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6qs1ko$jm7$3506@artemis.backbone.ou.edu> Sender: <vista@telepath.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: sanguish@digifix.com (Scott Anguish) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: NEXTstep/Openstep for the Mac Date: 12 Aug 1998 21:37:26 GMT Organization: Digital Fix Development Message-ID: <6qt1um$rf1$1@news.digifix.com> References: <35CF0C69.B3B0CF5@ericsson.com> <SCOTT.98Aug11143008@slave.doubleu.com> <6qs5e8$i06$1@crib.corepower.com> <6qsg2o$sr$1@sun579.rz.ruhr-uni-bochum.de> <6qsg9n$j66$1@crib.corepower.com> <35d1caf2.0@news.depaul.edu> <6qskpb$lir$1@msunews.cl.msu.edu> In-Reply-To: <6qskpb$lir$1@msunews.cl.msu.edu> On 08/12/98, spammers@ruin.the.internet.channelu.com wrote: >In <35d1caf2.0@news.depaul.edu> Jonathan W Hendry wrote: <snip> >> It would be easier to just use raw dictionary content from >www.dict.org. >> > >It may be better to go with a open solution. IXKit is one of the >things >that I am most interested in porting - but I don't think I will get >around >to it for quite some time. What about using AIAT as was suggested? Its very fast, and included with Mac OS X Server. That gives you the indexing and the finding, you'd just need to write the parsing stuff... Omni has released a very basic Objective-C wrapper for AIAT.. -- Scott Anguish <sanguish@digifix.com> Stepwise - OpenStep/Rhapsody Information <URL:http://www.stepwise.com>
From: agave@my-dejanews.com Newsgroups: comp.sys.next.programmer Subject: Re: Which NIB Date: Thu, 13 Aug 1998 02:12:27 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6qti2b$ddb$1@nnrp1.dejanews.com> References: <35D0B1D4.1845008C@qualcomm.com> In article <35D0B1D4.1845008C@qualcomm.com>, Michael Simpson <t_msimps@qualcomm.com> wrote: > I'm trying to work through the CurrencyConverter on the Intel DR2. The > instructions say to modify CurrencyConverter.nib, yet the app tries to > load CurrencyConverter-windows.nib instead. My currency window does not > get shown but the blank window in the other nib instead. The Developer > tutorial is obviously for the DR2 Apple version. > Others have already explained the syntax issue but didn't touch on your nib problem. The multiple nibs are there so that an application can load a different interface based what platform it's running on. While this isn't strictly necessary the feature is there so you can make your application a 'good citizen' and follow the gui guidelines for whatever OS you deploy on. Your app shouldn't be loading the -windows nib unless you're running YellowBox for Windows (ie Windows 9x/NT is your base operating system). If you're running the full Rhapsody DR2 the nib with no extension should be used. I don't know if the platform is taken into account in the developer documentation, just keep in mind that if you're working under Windows you should be modifying the -windows nib. If you are running under Rhapsody DR2 then I can't help you except to suggest deleteing the -windows nib... can't load what isn't there, right :) -Ian hopes that clears things up a little -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: agave@my-dejanews.com Newsgroups: comp.sys.next.programmer Subject: Re: Which NIB Date: Thu, 13 Aug 1998 02:12:29 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6qti2e$dfs$1@nnrp1.dejanews.com> References: <35D0B1D4.1845008C@qualcomm.com> In article <35D0B1D4.1845008C@qualcomm.com>, Michael Simpson <t_msimps@qualcomm.com> wrote: > I'm trying to work through the CurrencyConverter on the Intel DR2. The > instructions say to modify CurrencyConverter.nib, yet the app tries to > load CurrencyConverter-windows.nib instead. My currency window does not > get shown but the blank window in the other nib instead. The Developer > tutorial is obviously for the DR2 Apple version. > Others have already explained the syntax issue but didn't touch on your nib problem. The multiple nibs are there so that an application can load a different interface based what platform it's running on. While this isn't strictly necessary the feature is there so you can make your application a 'good citizen' and follow the gui guidelines for whatever OS you deploy on. Your app shouldn't be loading the -windows nib unless you're running YellowBox for Windows (ie Windows 9x/NT is your base operating system). If you're running the full Rhapsody DR2 the nib with no extension should be used. I don't know if the platform is taken into account in the developer documentation, just keep in mind that if you're working under Windows you should be modifying the -windows nib. If you are running under Rhapsody DR2 then I can't help you except to suggest deleteing the -windows nib... can't load what isn't there, right :) -Ian hopes that clears things up a little -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
Newsgroups: comp.sys.next.programmer Subject: Re: audio programming References: <35CE5F2D.D56DA946@jove.acs.unt.edu> From: sdroll@NOSPMmathematik.uni-wuerzburg.de (Sven Droll) Message-ID: <35d2951c.0@uni-wuerzburg.de> Date: 13 Aug 98 07:26:20 GMT As a starting point, I would suggest searching for the MusicKit-distributation. There are many docs and examples in there. Also the online-documentation of NS 3.3 may have some interesting things for the basics of programming soundplaying. greets Sven -- Sven Droll __ ______________________________________________________/ / ______ __ sdroll@mathematik.uni-wuerzburg.de / /_/ ___/ please remove the NOSPM from my reply-address /_ _/ _/ =====\_/======= LOGOUT FASCISM! ___________________________________________________________________ NeXT-mail or MIME welcome ;-)
From: info@trainingcompany.com Newsgroups: comp.sys.next.programmer Subject: FileMaker Pro training Date: 13 Aug 1998 00:45:25 GMT Organization: Macresource Computer Training Message-ID: <6qtcv5$luv$1@nnrp0.seg0> Learn FileMaker Pro in your browser. New on-line course with numerous screen-grabs and QuickTime movies. Just $9.99. http://www.trainingcompany.com/filemaker/
From: maul@anke.imsd.uni-mainz.DE (Mathias Maul) Newsgroups: comp.sys.next.programmer Subject: NS3.3 Interface Builder and NSObject Date: 13 Aug 1998 15:05:50 GMT Organization: Johannes Gutenberg-Universitaet Mainz, Germany Message-ID: <6quvce$1eo$1@bambi.zdv.Uni-Mainz.DE> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Hi there, just recently I encountered the following problem: When instantiating a class derived from NSObject in Interface Builder (black hardware, NS3.3), my program crashes with an uncaught exception: "(name of object) does not recognise selector +_canAlloc". The exception is raised at the moment when the main .nib file is loaded at startup. When I inherit the same class from Object, everything works fine. I read the manuals, searched the headers and libraries but could not find anything related to this problem. Any ideas? thanks, Matt.
From: Michael Simpson <t_msimps@qualcomm.com> Newsgroups: comp.sys.next.programmer Subject: Re: Which NIB Date: Thu, 13 Aug 1998 11:45:47 -0700 Organization: Qualcomm Message-ID: <35D3345B.7862376D@qualcomm.com> References: <35D0B1D4.1845008C@qualcomm.com> <6qti2e$dfs$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit First off, Thanks to everyone who replied. The parameter naming is actually pretty cool. I've never done Smalltalk so hadn't experienced that before. Ian, your response clears things up a bit. I could use a little further enlightenment however. There is this obvious ability to generate multiple nib files. What design issues dictate the creation of new nib files? Is there an organization benefit I should be aware of? On Mac or Windows, I usually just have one resource file. Also, should I always enforce MVC or are there times when eliminating the controller is ok? Thanks again, Michael agave@my-dejanews.com wrote: > > In article <35D0B1D4.1845008C@qualcomm.com>, > Michael Simpson <t_msimps@qualcomm.com> wrote: > > I'm trying to work through the CurrencyConverter on the Intel DR2. The > > instructions say to modify CurrencyConverter.nib, yet the app tries to > > load CurrencyConverter-windows.nib instead. My currency window does not > > get shown but the blank window in the other nib instead. The Developer > > tutorial is obviously for the DR2 Apple version. > > > > Others have already explained the syntax issue but didn't touch on > your nib problem. The multiple nibs are there so that an application can > load a different interface based what platform it's running on. While this > isn't strictly necessary the feature is there so you can make your > application a 'good citizen' and follow the gui guidelines for whatever OS > you deploy on. Your app shouldn't be loading the -windows nib unless you're > running YellowBox for Windows (ie Windows 9x/NT is your base operating > system). If you're running the full Rhapsody DR2 the nib with no extension > should be used. I don't know if the platform is taken into account in the > developer documentation, just keep in mind that if you're working under > Windows you should be modifying the -windows nib. If you are running under > Rhapsody DR2 then I can't help you except to suggest deleteing the -windows > nib... can't load what isn't there, right :) > > -Ian hopes that clears things up a little > > -----== Posted via Deja News, The Leader in Internet Discussion ==----- > http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: NS3.3 Interface Builder and NSObject Date: 13 Aug 1998 21:53:08 GMT Organization: Idiom Communications Message-ID: <6qvn84$r0c$1@news.idiom.com> References: <6quvce$1eo$1@bambi.zdv.Uni-Mainz.DE> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: maul@anke.imsd.uni-mainz.DE Mathias Maul may or may not have said: -> Hi there, -> -> just recently I encountered the following problem: -> -> When instantiating a class derived from NSObject in Interface Builder (black -> hardware, NS3.3), my program crashes with an uncaught exception: "(name of -> object) does not recognise selector +_canAlloc". The exception is raised at -> the moment when the main .nib file is loaded at startup. -> -> When I inherit the same class from Object, everything works fine. I read the -> manuals, searched the headers and libraries but could not find anything -> related to this problem. -> -> Any ideas? Yes. Add this code to your project: IBFix.h: //--------------------------------------------- // to make NSObjects compatible with nib loading: #import <foundation/NSObject.h> #import <appkit/appkit.h> #import <objc/objc-runtime.h> @interface NSObject(IBFix) +(BOOL)_canAlloc; -(BOOL) isKindOf:aClass; -(BOOL) respondsTo:(SEL)aSelector; - perform:(SEL)aSelector with:anObject; - perform:(SEL)aSelector with:object1 with:object2; @end IBFix.m: //--------------------------------------------- // to make NSObjects compatible with nib loading: #import "IBFix.h" @implementation NSObject(IBFix) +(BOOL)_canAlloc { return YES;} -(BOOL) isKindOf:aClass { if (aClass) return [self isKindOfClass:aClass]; return NO; } -(BOOL) respondsTo:(SEL) aSelector { if (ISSELECTOR(aSelector)) return [self respondsToSelector:aSelector]; return NO; } - perform:(SEL)aSelector with:anObject { return [self perform:aSelector withObject:anObject];} - perform:(SEL)aSelector with:object1 with:object2; { return [self perform:aSelector withObject:object1 withObject:object2];} @end -- "Although UNIX is more reliable, NT may become more reliable with time" - Ron Redman, deputy technical director of the Fleet Introduction Division of the Aegis Program Executive Office, US Navy. Where is the line between mere incompetence and actual treason? -jcr
From: "The only one :)" <progs2000@hotmail.com> Newsgroups: comp.sys.next.programmer,comp.sys.psion.programmer,comp.unix.programmer,comp.unix.sco.programmer,comp.windows.ms.programmer,de.comp.os.ms-windows.programmer,de.comp.os.os2.programmer,de.comp.os.unix.programming,dk.edb.programmering,dk.edb.program Subject: # 15 progs for only $150 !!! Date: Thu, 13 Aug 1998 23:38:59 +0200 Organization: World Access / Planet Internet Message-ID: <6qvmc9$rtp$1@reader2.wxs.nl> NNTP-Posting-Date: 13 Aug 1998 21:38:17 GMT 15 progs for only $150 !!! If you'd bought this in the store, it had cost you over $15.000 !!! This are the progs: ----------------------------------' Windows 98 Windows 98 PLUS 3D studio max release V2.5 MAY98 Adobe Photoshop 5.0 RETAIL PLUGINS Adobe Premiere v5.0 FINAL Cubase VST v3.553 full install D-lusion Drumstation v1.0 Font FX 32-bit RETAIL Adobe Illustrator 8.0 Beta X26 Incoming Kai's Power Show Macromedia Director V6.5 Adobe Photoshop 5.0 RETAIL Corel Print House Magic Deluxe 3.0 Pro Jpeg V3.0 For Adobe Photoshop ----------------------------------- I'll SEND you the progs FIRST(please send me your address...), then you must send the money to me... :)
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer,comp.sys.psion.programmer,comp.unix.programmer,comp.unix.sco.programmer,comp.windows.ms.programmer,de.comp.os.ms-windows.programmer,de.comp.os.os2.programmer,de.comp.os.unix.programming,dk.edb.programmering,dk.edb.program Subject: cmsg cancel <6qvmc9$rtp$1@reader2.wxs.nl> Control: cancel <6qvmc9$rtp$1@reader2.wxs.nl> Date: 13 Aug 1998 22:09:49 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6qvmc9$rtp$1@reader2.wxs.nl> Sender: "The only one :)" <progs2000@hotmail.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: Which NIB Date: 13 Aug 1998 22:03:12 GMT Organization: Idiom Communications Message-ID: <6qvnr0$r0c$2@news.idiom.com> References: <35D0B1D4.1845008C@qualcomm.com> <6qti2e$dfs$1@nnrp1.dejanews.com> <35D3345B.7862376D@qualcomm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: t_msimps@qualcomm.com Michael Simpson may or may not have said: -> First off, -> Thanks to everyone who replied. The parameter naming is actually pretty -> cool. I've never done Smalltalk so hadn't experienced that before. -> -> Ian, your response clears things up a bit. I could use a little further -> enlightenment however. There is this obvious ability to generate -> multiple nib files. What design issues dictate the creation of new nib -> files? Is there an organization benefit I should be aware of? On Mac -> or Windows, I usually just have one resource file. There are a lot of reasons to have multiple .nib files. One is, if you're not using a particular panel, why load it? Another reason is that very often OpenStep programming, a class will have some UI objects associated with it, and it's reasonable then to load a .nib in the process of intiializing an instance, to give each instance of that class its own set of the UI objects. This is the way most apps that can edit multiple documents (for example) will deal with creating a window per document. -> Also, should I always enforce MVC or are there times when eliminating -> the controller is ok? Trust yourself. MVC is a good way to think about it, but sometimes it's just fine for the V object to also be the C object. -jcr -- "Although UNIX is more reliable, NT may become more reliable with time" - Ron Redman, deputy technical director of the Fleet Introduction Division of the Aegis Program Executive Office, US Navy. Where is the line between mere incompetence and actual treason? -jcr
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6r083e$9df@nnews.vitinc.com> Control: cancel <6r083e$9df@nnews.vitinc.com> Date: 14 Aug 1998 02:35:57 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6r083e$9df@nnews.vitinc.com> Sender: kbenrmca@avkristne.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: spammers@ruin.the.internet.channelu.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: NEXTstep/Openstep for the Mac Date: 14 Aug 1998 03:29:36 GMT Organization: Michigan State University Message-ID: <6r0av0$mm8$1@msunews.cl.msu.edu> References: <35CF0C69.B3B0CF5@ericsson.com> <SCOTT.98Aug11143008@slave.doubleu.com> <6qs5e8$i06$1@crib.corepower.com> <6qsg2o$sr$1@sun579.rz.ruhr-uni-bochum.de> <6qsg9n$j66$1@crib.corepower.com> <35d1caf2.0@news.depaul.edu> <6qskpb$lir$1@msunews.cl.msu.edu> <6qt1um$rf1$1@news.digifix.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: sanguish@digifix.com In <6qt1um$rf1$1@news.digifix.com> Scott Anguish wrote: > On 08/12/98, spammers@ruin.the.internet.channelu.com wrote: > >In <35d1caf2.0@news.depaul.edu> Jonathan W Hendry wrote: > <snip> > >> It would be easier to just use raw dictionary content from > >www.dict.org. > >> > > > >It may be better to go with a open solution. IXKit is one of the > >things > >that I am most interested in porting - but I don't think I will get > >around > >to it for quite some time. > > What about using AIAT as was suggested? Its very fast, and > included with Mac OS X Server. > > That gives you the indexing and the finding, you'd just need > to write the parsing stuff... Omni has released a very basic > Objective-C wrapper for AIAT.. > Hmmm. I guess I didn't see that. If I get my hands on DR2 and give up really reading and posting to Rhaptel/newsgroups maybe I can get some real work done :) I have DR1 I'll have to check out whether it's there in some primitive incarnation. Randy rencsok at channelu dot com argus dot cem dot msu dot edu spammers works also :) Randy Rencsok General UNIX, NeXTStep, IRIX Admining, Turbo Software Consulting, Programming, etc.)
From: mezzino@gauss.cl.uh.edu (Mike Mezzino) Newsgroups: comp.sys.next.programmer Subject: OPENSTEP EOF2/NT + Oracle question Date: 14 Aug 1998 16:33:04 GMT Organization: University of Houston Message-ID: <6r1os0$hir$1@Masala.CC.UH.EDU> I just moved my OPENSTEP/Mach EOF2 app to NT and it compiled perfectly the first time. I was very surprised. However, it does not run! It seems that I am missing something in NT which the Oracle adaptor needs. Specifically, a file named OCIW32.dll Both EOModeler and my app complain about not finding this file anywhere and I understand why - it is not on my system. Did I miss something in the installation or must I have this file to run? Where can I get this file? By the way, I placed the tnsnames.ora file in C:/Next/etc, where C:/Next is my Next root directory in NT. I suppose that this could eventually be a problem also. Is this the right place for this file in an NT environment. Thanks so much. Regards, Mike Mezzino mezzino@gauss.cl.uh.edu
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <3VTBKW85.669M0249@icsdev.net> Control: cancel <3VTBKW85.669M0249@icsdev.net> Date: 14 Aug 1998 17:47:03 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.3VTBKW85.669M0249@icsdev.net> Sender: email@icsdev.net Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
Message-ID: <35D4CE85.F8638BBA@home.com> From: msms <msms@home.com> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Re: Network and BSD questions References: <6qfti8$oos@luc.ircam.fr> <35CF9962.2AD93539@home.com> <6qprk6$5nb$1@news.spacelab.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Fri, 14 Aug 1998 23:54:09 GMT NNTP-Posting-Date: Fri, 14 Aug 1998 16:54:09 PDT Organization: @Home Network Perhaps I should have said Rhapsody will have ftpd setup by default. I also recommend "Essential System Administration", it has some nice tips and methods. It may not be the best book to start off with though. "Teach Yourself Unix in a Week" from SAMS I think is a good book to get one's feet wet. It's easy to understand and covers good areas of using UNIX. -- John ------------------------ Charles W. Swiger wrote: > > msms <msms@home.com> wrote: > [ ... ] > >Rhapsody will have ftpd running by default, so you should > >be able to ftp right into Rhapsody. > > Not quite. Rhapsody's inetd is set up by default to spawn ftpd's if an ftp > request is made to the machine, but it does not run an FTP server > continuously. > > As for good sysadmin references, the O'Reilly "Essential System > Administration" and the "TCP/IP Network Administration" are two bibles... > > -Chuck > > Charles Swiger | chuck@codefab.com | standard disclaimer > ---------------+-------------------+-------------------- > "Microsoft: we make the easy almost impossible."
From: holger@_REMOVE_THIS_.wizards.de (Holger Hoffstaette) Newsgroups: comp.sys.next.programmer Subject: Re: OPENSTEP EOF2/NT + Oracle question Date: 14 Aug 1998 20:53:41 GMT Organization: The secret circle of the NSRC Message-ID: <6r284l$169@ragnarok.en.uunet.de> References: <6r1os0$hir$1@Masala.CC.UH.EDU> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit NNTP-Posting-Date: 14 Aug 1998 21:13:33 GMT Mike Mezzino wrote: > I just moved my OPENSTEP/Mach EOF2 app to NT and it compiled perfectly > the first time. I was very surprised. However, it does not run! It seems that I > am missing something in NT which the Oracle adaptor needs. Specifically, a file named > > OCIW32.dll You'll need to buy Oracle client libraries. These are responsible for the connection management to the database; the EOF adaptor 'just' wraps the received data into objects. See the EOF release notes for more information.. :-) Holger
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <11760902635220@digifix.com> Date: 16 Aug 1998 03:48:21 GMT Organization: Digital Fix Development Message-ID: <15411903240023@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1994. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.pen Subject: cmsg cancel <35d6720c00000930@dns.icsdev.net-MINC> Control: cancel <35d6720c00000930@dns.icsdev.net-MINC> Date: 16 Aug 1998 07:40:22 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.35d6720c00000930@dns.icsdev.net-MINC> Sender: ixikojae@businfo.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.pen Subject: cmsg cancel <35d69dcd00001851@dns.icsdev.net-MINC> Control: cancel <35d69dcd00001851@dns.icsdev.net-MINC> Date: 16 Aug 1998 10:16:00 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.35d69dcd00001851@dns.icsdev.net-MINC> Sender: xpbsxbyf@businfo.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.pen Subject: cmsg cancel <35d69dce00001852@dns.icsdev.net-MINC> Control: cancel <35d69dce00001852@dns.icsdev.net-MINC> Date: 16 Aug 1998 10:16:00 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.35d69dce00001852@dns.icsdev.net-MINC> Sender: xpbsxbyf@businfo.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6r6it8$gta$3506@artemis.backbone.ou.edu> Control: cancel <6r6it8$gta$3506@artemis.backbone.ou.edu> Date: 16 Aug 1998 20:51:43 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6r6it8$gta$3506@artemis.backbone.ou.edu> Sender: <vista@telepath.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: maul@anke.imsd.uni-mainz.DE (Mathias Maul) Newsgroups: comp.sys.next.programmer Subject: Re: NS3.3 Interface Builder and NSObject Date: 18 Aug 1998 08:46:53 GMT Organization: Johannes Gutenberg-Universitaet Mainz, Germany Message-ID: <6rbf1t$924$1@bambi.zdv.Uni-Mainz.DE> References: <6quvce$1eo$1@bambi.zdv.Uni-Mainz.DE> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: maul@anke.imsd.uni-mainz.DE Thanks to all for the responses, everything is working fine now. foo!, Matt.
From: totopas@my-dejanews.com Newsgroups: comp.sys.next.programmer Subject: Using [standard] C sources in a WebObjects project Date: Wed, 19 Aug 1998 14:06:50 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6rem5q$ed4$1@nnrp1.dejanews.com> Hello Everybody. I have the following problem: I am creating an interactive commercial Web site (selling 'puter equippement forexample) and I am using a payment system Payline (it's a french thing, not sure you know it, but that is not a problem). The problem is that the crypting and uncrypting sources are written in C (coz theyweremadefor "classical" CGIs). I managed to include the appopriate C files into the project, but I don't know how to use the C methods in the other Java project sources. Is the mechanism the same as for including native methods in any Java program? If yes, does the project builder create the headers itself or should one make them by hand, with "javah" with some options? Thank you in advance. Anton Passiouk -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: "Andrew Malota" <rinn_the_great@hotmail.com> Newsgroups: comp.sys.next.programmer Subject: I'm a new Programmer, and need a languege. Date: Thu, 20 Aug 1998 21:35:15 -0500 Organization: Texas A&M University, College Station, Texas Message-ID: <6rimju$h80$1@news.tamu.edu> NNTP-Posting-Date: 21 Aug 1998 02:38:54 GMT I'd like to start programming with NEXTSTEP. I have a NeXTStation running NEXTSTEP 2.0. If you know of any FREE programming langueges out there that would be good for me, please write. Thanks. Andrew Malota
From: quinlan@intergate.bc.ca Newsgroups: comp.sys.next.programmer Subject: Re: I'm a new Programmer, and need a languege. Date: Fri, 21 Aug 1998 08:39:39 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6rjboa$nhe$1@nnrp1.dejanews.com> References: <6rimju$h80$1@news.tamu.edu> In article <6rimju$h80$1@news.tamu.edu>, "Andrew Malota" <rinn_the_great@hotmail.com> wrote: > I'd like to start programming with NEXTSTEP. I have a NeXTStation running > NEXTSTEP 2.0. If you know of any FREE programming langueges out there that > would be good for me, please write. Thanks. You could learn how to program in almost any language using this computer. I would recommend Objective-C. -- Brian Quinlan quinlan@intergate.bc.ca -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
Subject: Re: Confirmed Compiler Optimization Bug Newsgroups: comp.sys.next.programmer References: <6rh8uu$7vk1@onews.collins.rockwell.com> In-Reply-To: <6rh8uu$7vk1@onews.collins.rockwell.com> From: marco@sente.ch.mil (Marco Scheurer) Message-ID: <35dd5cb8.0@epflnews.epfl.ch> Date: 21 Aug 98 11:40:40 GMT Organization: EPFL On 08/20/98, Erik M. Buck wrote: >Confirmed Compiler Optimization Bug > >Summary: I have an application that works correctly and produces >the correct results when compiled with -O0 (no optimization) but >does not work correctly and produces incorrect output when compiled >with -O. Openstep 4.2 for Mach and Windows (standard compiler) > >History: A year ago, my team and I ported a NeXTstep application >to Openstep. After the port, we started getting lots of Display >Postscript errors. Debugging the DPS output revealed that NaN >(Not a Number) was being sent to the PS server and the server was >correctly generating Range errors. > >After further debugging, we produced some data sets that consistently >produced the errors. The strange thing was that sometimes the >error was in the clipping rect sent by an NSPopUpButton or in the >G component of RGBA sent by a text field. How could it be that >NeXT's code was producing these errors? > Some of our code was producing NaN as well. The strange thing > was, we could >call the same function with the same arguments and get the correct >answer 80-90% of the time and NaN otherwise. This function has >no side effects and uses only local variables. Its code is included >below. > [...] > >The code: VCOMMON_EXTERN NSPoint >VTransMatrixTransformPoint(VTransMatrixData someData, float x, >float y) { > NSPoint result; > > result.x = (someData.a * x) + (someData.c * y) + someData.e; > result.y = (someData.b * x) + (someData.d * y) + someData.f; > > return result; >} > > >VCOMMON_EXTERN is similar to FOUNDATION_EXTERN and is defined as >extern on Mach. > > I'm ready to bet that either someData fields or x or y are not properly initialized. Marco Scheurer Sen:te -- Marco Scheurer (remove "dot mil" from my address) Sen:te OPENSTEP/Rhapsody and WebObjects development, consulting and mentoring.
From: pemmerik@solair1.inter.NL.net (P.J.L.van Emmerik) Newsgroups: comp.sys.next.programmer Subject: Version Control on NS3.3 ??? Date: Fri, 21 Aug 1998 06:54:01 GMT Organization: UUNET-NL Message-ID: <6rj5ib$f45$1@newnews.nl.uu.net> Has anyone experience with version management and revision control on NEXTSTEP 3.3 using the NEXTSTEP tools??? I would like to start using a version control system and therefor would apreciate some suggestions. Please Email to: emmerik@qtecq.com P.J.L. van Emmerik Holec Projects B.V. Email: emmerik@qtecq.com PO.BOX 565, 7550 AN Hengelo pemmerik@solair1.inter.NL.net The Netherlands Phone: +31 74 2558 688 --
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: Confirmed Compiler Optimization Bug Date: 21 Aug 1998 13:19:23 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6tqsuq.st.rog@talisker.ohm.york.ac.uk> References: <6rh8uu$7vk1@onews.collins.rockwell.com> <35dd5cb8.0@epflnews.epfl.ch> On 21 Aug 98 11:40:40 GMT, Marco Scheurer <marco@sente.ch.mil> wrote: > I'm ready to bet that either someData fields or x or y are not properly > initialized. i wouldn't lay down your money too quickly, as there is definitely a compiler bug related to returning structures from functions, 'cos i've run across it. (see earlier posting in this thread) rog.
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: Heretical question Newsgroups: comp.sys.next.programmer Message-ID: <35dd9854.0@news.depaul.edu> Date: 21 Aug 98 15:55:00 GMT Is Apple - or anyone else - working on a YellowBox framework for easily using native Windows API's? This strikes me as being a necessity if they really want the YellowBox to be adopted by business for developing Windows apps. Further, a framework of this sort would help centralize Windows dependencies in an app, rather than scattering them throughout the app's code. -- Note: email to this address goes to /dev/null To email a reply, write to jon at exnext dot com
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: IB corrupted a nib? Newsgroups: comp.sys.next.programmer Message-ID: <35de11be.0@news.depaul.edu> Date: 22 Aug 98 00:33:02 GMT All of a sudden, one of my .nibs is corrupt. It references a class called '%NSCell', which doesn't exist. When the nib is loaded, an exception is raised. Boom. I believe it was corrupted when I added a matrix of NSButtonCells. Removing the matrix did not remove the reference to %NSCell. Any suggestions on how I might salvage the nib? -- Note: email to this address goes to /dev/null To email a reply, write to jon at exnext dot com
From: Bill Bumgarner <bbum@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: Heretical question Date: Fri, 21 Aug 1998 22:17:18 -0400 Organization: Spacelab.net Internet Access Message-ID: <Pine.NXT.3.96.980821221425.15110B-100000@pathos> References: <35dd9854.0@news.depaul.edu> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII To: jon@exnext.com In-Reply-To: <35dd9854.0@news.depaul.edu> It isn't actually a necessity. The point of YB is to make the fact that you are programming windows, MacOS, or OS X Server completely *transparent to the developer*. Single source tree works everywhere with little to no #ifdef style garbage. The only platform specific stuff are things like changes to make the different windowing and menu paradigms more "native" in their look and feel. However, this generally does not require any "native" code. I would be interested in what particular WIN32 APIs you [or others] find it necessary to invoke when developping YB apps? b.bum
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: IB corrupted a nib? Date: 22 Aug 1998 05:29:30 GMT Organization: Idiom Communications Message-ID: <6rlkvq$30m$1@news.idiom.com> References: <35de11be.0@news.depaul.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: jhendry@shrike.depaul.edu Jonathan W Hendry may or may not have said: -> All of a sudden, one of my .nibs is corrupt. It references -> a class called '%NSCell', which doesn't exist. When the nib -> is loaded, an exception is raised. Boom. -> -> I believe it was corrupted when I added a matrix of -> NSButtonCells. Removing the matrix did not remove the -> reference to %NSCell. -> -> Any suggestions on how I might salvage the nib? If it will still load in IB, then open it, create a new .nib file, copy everything into the new file, and save it as the old name. This has worked for me in tha past. The other thing you can try, is open the .nib wrapper, and manually edit data.classes file to remove the reference to %cell. -- "Although UNIX is more reliable, NT may become more reliable with time" - Ron Redman, deputy technical director of the Fleet Introduction Division of the Aegis Program Executive Office, US Navy. Where is the line between mere incompetence and actual treason? -jcr
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: objective C, yacc and ProjectBuilder under openstep 4.2 Date: 22 Aug 1998 09:11:11 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6tt2ph.2br.rog@talisker.ohm.york.ac.uk> has anyone had success using objective-C yacc under project builder? there's some support for ".ym" files in /NextDeveloper/Makefiles/pb_makefiles/implicitrules.make, however it doesn't work properly. if i try adding ".ym" to make's SUFFIXES, it partially works; initially, it builds the derived source from the yacc file, and compiles that alright, but when i then update the yacc file, it doesn't rerun yacc. my only solution is to delete the stuff in derived_src; and currently i've got a situation where i have to delete the derived_src files *and* touch another file in the subproject before it gets rebuilt... anyone got a solution? something so simple in a normal makefile surely should be possible under Project Builder? cheers, rog.
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: Re: IB corrupted a nib? Newsgroups: comp.sys.next.programmer References: <35de11be.0@news.depaul.edu> <6rlkvq$30m$1@news.idiom.com> Message-ID: <35decb21.0@news.depaul.edu> Date: 22 Aug 98 13:44:01 GMT John C. Randolph <jcr.remove@this.phrase.idiom.com> wrote: > Jonathan W Hendry may or may not have said: > -> All of a sudden, one of my .nibs is corrupt. It references > -> a class called '%NSCell', which doesn't exist. When the nib > -> is loaded, an exception is raised. Boom. > -> > -> I believe it was corrupted when I added a matrix of > -> NSButtonCells. Removing the matrix did not remove the > -> reference to %NSCell. > -> > -> Any suggestions on how I might salvage the nib? > If it will still load in IB, then open it, create a new .nib file, copy > everything into the new file, and save it as the old name. > This has worked for me in tha past. Cool. I'll try that. > The other thing you can try, is open the .nib wrapper, and manually edit > data.classes file to remove the reference to %cell. Oddly, it's not in data.classes, just objects.nib - the binary file. -- Note: email to this address goes to /dev/null To email a reply, write to jon at exnext dot com
From: freexxx@fuse.net Newsgroups: comp.sys.next.programmer Subject: 110% free XXX pics Date: 22 Aug 98 10:33:39 GMT Organization: Politecnico di Torino - Italia Message-ID: <35de9e83.0@194.116.20.4> No Credit Card needed. 100% FREE XXX site. For tons of free pics please visit us at http://www.bearcatonline.com/adult/index6.htm
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <35de9e83.0@194.116.20.4> Control: cancel <35de9e83.0@194.116.20.4> Date: 22 Aug 1998 14:25:33 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.35de9e83.0@194.116.20.4> Sender: freexxx@fuse.net Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: Re: IB corrupted a nib? Newsgroups: comp.sys.next.programmer References: <35de11be.0@news.depaul.edu> <6rlkvq$30m$1@news.idiom.com> <35decb21.0@news.depaul.edu> Message-ID: <35dee139.0@news.depaul.edu> Date: 22 Aug 98 15:18:17 GMT Jonathan W Hendry <jhendry@shrike.depaul.edu> wrote: > John C. Randolph <jcr.remove@this.phrase.idiom.com> wrote: > > If it will still load in IB, then open it, create a new .nib file, copy > > everything into the new file, and save it as the old name. > > This has worked for me in tha past. > Cool. I'll try that. Drat, didn't work. The new one has the same problem. -- Note: email to this address goes to /dev/null To email a reply, write to jon at exnext dot com
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: Solution (?) Was Re: IB corrupted a nib? Newsgroups: comp.sys.next.programmer References: <35de11be.0@news.depaul.edu> <6rlkvq$30m$1@news.idiom.com> <35decb21.0@news.depaul.edu> <35dee139.0@news.depaul.edu> Message-ID: <35dee381.0@news.depaul.edu> Date: 22 Aug 98 15:28:01 GMT Jonathan W Hendry <jhendry@shrike.depaul.edu> wrote: > Jonathan W Hendry <jhendry@shrike.depaul.edu> wrote: > > John C. Randolph <jcr.remove@this.phrase.idiom.com> wrote: > > > If it will still load in IB, then open it, create a new .nib file, copy > > > everything into the new file, and save it as the old name. > > > This has worked for me in tha past. > > Cool. I'll try that. > Drat, didn't work. The new one has the same problem. I found the culprit: NSBox. I had an NSBox, configured as a line, stretching across the window. Removing it solved the problem. Now, *why*? -- Note: email to this address goes to /dev/null To email a reply, write to jon at exnext dot com
From: email@address.com Newsgroups: comp.sys.next.programmer Subject: *****100 % Free Porn***** Date: 22 Aug 98 17:55:25 GMT Organization: State of Minnesota Message-ID: <35df060d.0@www.an.cc.mn.us> Check out this great 100% free site it does not require a credit card, no long distance charges and no 800 numbers. It is pure hardcore with dozens of sections. Check it out now <a href="http://www.bearcatonline.com/adult/index5.htm>Click Here For 100% Free Porn</a>
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <35df060d.0@www.an.cc.mn.us> Control: cancel <35df060d.0@www.an.cc.mn.us> Date: 22 Aug 1998 18:17:48 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.35df060d.0@www.an.cc.mn.us> Sender: email@address.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: email@address.com Newsgroups: comp.sys.next.programmer Subject: FREE PORN Date: 22 Aug 98 18:11:20 GMT Organization: State of Minnesota Message-ID: <35df09c8.0@www.an.cc.mn.us> Check out this great 100% free site it does not require a credit card, no long distance charges and no 800 numbers. It is pure hardcore with dozens of sections. Check it out now <a href="http://www.bearcatonline.com/adult/index5.htm>Click Here For 100% Free Porn</a>
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <35df09c8.0@www.an.cc.mn.us> Control: cancel <35df09c8.0@www.an.cc.mn.us> Date: 22 Aug 1998 18:47:28 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.35df09c8.0@www.an.cc.mn.us> Sender: email@address.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Bill Bumgarner <bbum@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: IB corrupted a nib? Date: Sat, 22 Aug 1998 13:24:22 -0400 Organization: Spacelab.net Internet Access Message-ID: <Pine.NXT.3.96.980822131516.21419A-100000@pathos> References: <35de11be.0@news.depaul.edu> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII To: jon@exnext.dot.com In-Reply-To: <35de11be.0@news.depaul.edu> %NSCell is the result of a poseing operation. That is, some class decided to [SomeClass poseAs: [NSCell class]]. When this occurs, the ObjC runtime basically: - points NSCell's isa to SomeClass's isa - creates a new class named %NSCell that contains the contents of NSCell's isa Since SomeClass must be a subclass of NSCell for posing to work correctly, all of the super isa's are correct after the above operations. That is, the new NSCell class inherits from %NSCell which inherets from NSCell's original super class. It is likely that the %NSCell "class" is not actually registered in the runtime in such a manner that it can be retrieved via normal means. ANYWAY, it sounds like something did a +poseAs: way late. Posing should ideally be done before any instances of the class to be posed have been created. This is because it is impossible [without lots of record keeping] for ObjC to know about all the instances and switch their isa pointers on the fly. It isn't unlikely that IB does some of its magic through posing... maybe it has a bug where posing is occurring late? Or, it may be that you opened the nib and created some objects [including an NSCell], then loaded a palette that did some posing? If none of THAT works and you are either particularly motivated not to have to recreate the NIB or simply want an interesting intellectual exercise, it might be possible to catch the class not found exception [either via posing over NSException or through a creative ObjC runtime hacque] and "fix" the problem that way. BTW: Are you sure you nailed all the references to %NSCell in the data file? b.bum On 22 Aug 1998, Jonathan W Hendry wrote: > All of a sudden, one of my .nibs is corrupt. It references > a class called '%NSCell', which doesn't exist. When the nib > is loaded, an exception is raised. Boom. > > I believe it was corrupted when I added a matrix of > NSButtonCells. Removing the matrix did not remove the > reference to %NSCell. > > Any suggestions on how I might salvage the nib? > > -- > Note: email to this address goes to /dev/null > To email a reply, write to jon at exnext dot com > >
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <15411903240023@digifix.com> Date: 23 Aug 1998 03:48:24 GMT Organization: Digital Fix Development Message-ID: <6415903844821@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1994. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: objective C, yacc and ProjectBuilder under openstep 4.2 Date: 24 Aug 1998 14:38:34 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6u2une.2t9.rog@talisker.ohm.york.ac.uk> References: <slrn6tt2ph.2br.rog@talisker.ohm.york.ac.uk> <6rqv5v$2g6$1@hermes.is.co.za> On 24 Aug 1998 05:54:07 GMT, gvandyk@icon.co.za <gvandyk@icon.co.za> wrote: > On 08/22/98, Roger Peppe wrote: > >has anyone had success using objective-C yacc under project builder? > My solution to this problem was adding the following lines to the > Makefile.postamble in The Project. > > --- Start > # Next Forgot to put this rule in implicitRules Makefile > .SUFFIXES: .ym .lm .h .m i tried this, and it still didn't work properly. (the metarule doesn't appear to be resolving correctly) however i have now solved the problem by simply including an explicit rule to build the parser, copying the recipe for the rule from implicitrules.make. the relevant rule now looks like this: ${SFILE_DIR}/parser.h ${SFILE_DIR}/parser.m: parser.ym $(CD) $(SFILE_DIR) && $(YACC) $(ALL_YFLAGS) $(shell pwd)/parser.ym $(CP) $(SFILE_DIR)/y.tab.h $(SFILE_DIR)/parser.h $(MV) $(SFILE_DIR)/y.tab.c $(SFILE_DIR)/parser.m this works fine. cheers, rog.
From: ambi@big.aa.net (Mike Amirault) Newsgroups: comp.sys.next.programmer Subject: Re: Heretical question Date: 24 Aug 1998 13:34:15 -0700 Organization: Alternate Access Inc. - Affordable, Reliable Internet Access Message-ID: <6rsio7$m78$1@big.aa.net> References: <35dd9854.0@news.depaul.edu> <Pine.NXT.3.96.980821221425.15110B-100000@pathos> > I would be interested in what particular WIN32 APIs you [or others] find > it necessary to invoke when developping YB apps? How about sound? A pretty basic functionality that's missing from YellowBox for Windows. Also whatever functions are necessary to put an icon in Windows system tray thingy would be cool. Mike
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: Re: Heretical question Newsgroups: comp.sys.next.programmer References: <35dd9854.0@news.depaul.edu> <Pine.NXT.3.96.980821221425.15110B-100000@pathos> <6rsio7$m78$1@big.aa.net> Message-ID: <35e1d517.0@news.depaul.edu> Date: 24 Aug 98 21:03:19 GMT Mike Amirault <ambi@big.aa.net> wrote: > > I would be interested in what particular WIN32 APIs you [or others] find > > it necessary to invoke when developping YB apps? > How about sound? A pretty basic functionality that's missing from > YellowBox for Windows. Also whatever functions are necessary to > put an icon in Windows system tray thingy would be cool. Evidently it's possible, since MailViewer.app does that. It would be nice if there were a method on NSApplication to do it. Another useful thing would be a way to access the Windows Registry (frex, if I wanted to write a Browser-based replacement for RegEdit.) NSUserDefaults is stored in the Registry, but AFAIK you cannot access registry entries for non-YellowBox applications. Or a way to tweak the OS to change the desktop image. This sort of thing would be useful for a cross-platform desktop utility. Also, does NSFileManager create Windows shortcuts, or does it just barf if you ask for a link? It would be nice if you could programmatically change the icon on a shortcut. (I'm not sure if that's even possible in Win32.) -- Note: email to this address goes to /dev/null To email a reply, write to jon at exnext dot com
Newsgroups: comp.sys.next.programmer From: maury@remove_this.istar.ca (Maury Markowitz) Subject: Filtering problems Content-Type: text/plain; charset=us-ascii Message-ID: <Ey7vAK.7Fo@T-FCN.Net> Sender: news@T-FCN.Net Content-Transfer-Encoding: 7bit Organization: none Mime-Version: 1.0 Date: Mon, 24 Aug 1998 23:15:55 GMT A beta tester of ours pointed out an odd problem. I ended up putting a whole day into it, and I don't seem to be any closer to a perfect solution than I was before. Basically he noted that dragging a GIF image into our app resulted in an error. After some poking about (I hadn't been in that code in months) I saw what had happened. Basically instead of storing an image directly, we keep the data in a fileWrapper, which is then saved out to disk when the user does a Save. This also makes it a bit easier when you're dealing with linked file vs. copied ones. Anyway the original problem was pretty basic, I was loading up the image into the fileWrapper from the Pboard, and then when I set the fill to use the image I was simply opening the FW and using what was inside - unfiltered. So this seemed easy enough to fix, I instead made a new pboard and did an initFromFile: with the file wrapper's symlink or filename. Then I filtered that, and got the image out of it with NSImage's initWithPasteboard: The original problem was then fixed, but then I started seeing really odd behaviour. The main issue I've seen before - filtering sometimes fails but returns no obvious indication that anything's wrong. For instance I've found that sometimes the image will not be converted, yet hand back what appears to be valid data (including a YES from canInitFrom...). So on one run of the app everything is OK, then on the next they come up empty. What's weirder is that in many of the cases the first "import" of a non-native format will work great - say turning a GIF into a TIFF for me. From that point on (sometimes) it will work great on GIF's, but dragging in a TIFF will fail. Even weirder, I was just trying it to make sure I wasn't dreaming, and now it won't accept any drags at all! Question 1: is there any way to get the _data_ from a fileWrapper without having to know about the particulars? It seems to be that having to know if it's a link or real file and then call the proper routine violates the entire concept of having the wrapper in the first place. Is FileWrapper really as bad as I think it is? Q2: has anyone seen this sort of "bad filtering" before? Q3: should I consider looking at the filename of the wrapper and directly loading up the image in the cases where a conversion isn't needed? This seems really ugly. Q4: any "undocumented" tricks I'm missing? Do I need to put in a wait anywhere on one of the method calls or something? Maury
From: afs@netaxs.com (AFS) Newsgroups: comp.sys.next.programmer Subject: Re: Confirmed Compiler Optimization Bug Date: 25 Aug 1998 04:30:06 GMT Organization: Philadelphia's Complete Internet Provider Message-ID: <6rteke$a01@netaxs.com> References: <6rh8uu$7vk1@onews.collins.rockwell.com> <35dd5cb8.0@epflnews.epfl.ch> <slrn6tqsuq.st.rog@talisker.ohm.york.ac.uk> Roger Peppe (rog@ohm.york.ac.uk) wrote: : On 21 Aug 98 11:40:40 GMT, Marco Scheurer <marco@sente.ch.mil> wrote: : > I'm ready to bet that either someData fields or x or y are not properly : > initialized. : : i wouldn't lay down your money too quickly, as there is definitely a : compiler bug related to returning structures from functions, 'cos i've : run across it. (see earlier posting in this thread): I don't recall when this got fixed, but I can tell you 100% for CERTAIN that a bug of this nature existed during the beta phase of some iteration of Rhapsody. Wouldn't surprise me at all if this was a carryover from a problem that had been introduced in an OS 4.x version of the compiler. In general, the compiler was emitting bad code for the situation where something other tha a simple type or pointer was returned from a function. There's a special hidden argument in the stack for this case. Greg
From: Christian Neuss <neuss.@informatik.th-darmstadt.de.nos-pam> Newsgroups: comp.sys.next.programmer Subject: Re: Confirmed Compiler Optimization Bug Date: 25 Aug 1998 08:07:52 GMT Organization: Technische Universitaet Darmstadt Message-ID: <6rtrco$bba$1@sun27.hrz.tu-darmstadt.de> References: <6rh8uu$7vk1@onews.collins.rockwell.com> <35dd5cb8.0@epflnews.epfl.ch> <slrn6tqsuq.st.rog@talisker.ohm.york.ac.uk> <6rteke$a01@netaxs.com> afs@netaxs.com (AFS) wrote: >Roger Peppe (rog@ohm.york.ac.uk) wrote: >: On 21 Aug 98 11:40:40 GMT, Marco Scheurer <marco@sente.ch.mil> wrote: >: > I'm ready to bet that either someData fields or x or y are not properly >: > initialized. >: >: i wouldn't lay down your money too quickly, as there is definitely a >: compiler bug related to returning structures from functions, 'cos i've >: run across it. (see earlier posting in this thread): > >I don't recall when this got fixed, but I can tell you 100% for CERTAIN >that a bug of this nature existed during the beta phase of some iteration >of Rhapsody. Wouldn't surprise me at all if this was a carryover from a >problem that had been introduced in an OS 4.x version of the compiler. >In general, the compiler was emitting bad code for the situation where >something other tha a simple type or pointer was returned from a function. >There's a special hidden argument in the stack for this case. Hm, I seem to remember that way back when (I still programmed C++ on a Sun machine, <shudder>), the gnu compiler was notorious for not correctly returning structs, and I don't know if this was ever fixed. Maybe the 'known bugs' list for the respective compiler version could clear this up? Regards, Chris -- // Christian Neuss "static typing? how quaint.." // http://www.nexttoyou.de/~neuss/ // fax: (+49) 6151 16 5472
From: Erik Doernenburg <erik@object-factory.REMOVE-ME.com> Newsgroups: comp.sys.next.programmer Subject: Re: Heretical question Date: 25 Aug 1998 08:24:44 GMT Organization: Object Factory GmbH (Germany) Message-ID: <6rtscc$k81$1@leonie.object-factory.com> References: <35dd9854.0@news.depaul.edu> <Pine.NXT.3.96.980821221425.15110B-100000@pathos> <6rsio7$m78$1@big.aa.net> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit ambi@big.aa.net (Mike Amirault) wrote: > > > I would be interested in what particular WIN32 APIs you [or others] > > find it necessary to invoke when developping YB apps? > > How about sound? A pretty basic functionality that's missing from > YellowBox for Windows. Also whatever functions are necessary to > put an icon in Windows system tray thingy would be cool. About the "tray" thing; have a look at NSStatusBar/Item. There's no documentation but in DR2 you have the header files. It's easy to figure out, except I still don't know in what way the length parameter is modified, ie. the item doesn't seem to have the width you specified... reagards, erik -- OBJECT FACTORY, Gesellschaft für Informatik und Datenverarbeitung mbH Telephone: ++49 +231 9751370, Internet: http://www.object-factory.com
From: Jamie Hogg <jwh100@york.ac.uk> Newsgroups: comp.sys.newton.programmer,comp.sys.next.programmer,comp.unix.programmer,comp.unix.user-friendly Subject: Look&Feel, Computer Music Software Date: Tue, 25 Aug 1998 14:51:12 +0100 Organization: University of York Sender: jwh100@york.ac.uk Message-ID: <35E2C150.299A4949@york.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi, I'm writing a piece of software (in Java) that is a graphical front-end for a powerful suite of audio-visual synthesis tools. I have a web page that shows what my software looks like; http://www.york.ac.uk/~jwh100/Survey/SFront.htm I would be very grateful if you had a quick look at my software and tell me what you think about the look and feel of my User Interface screen in the space provided (take the link to PROTOTYPE). Thanks, Jamie.
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.newton.programmer,comp.sys.next.programmer,comp.unix.programmer,comp.unix.user-friendly Subject: cmsg cancel <35E2C150.299A4949@york.ac.uk> Control: cancel <35E2C150.299A4949@york.ac.uk> Date: 25 Aug 1998 14:01:42 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.35E2C150.299A4949@york.ac.uk> Sender: jwh100@york.ac.uk Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
Message-ID: <35E2EC80.692D@paragon-software.com> From: "B.G. Mahesh" <mahesh@paragon-software.com> Organization: Paragon Software, Inc. MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer,va.jobs Subject: JOBS: OPENSTEP,Perm/Contract, Wash.DC area (USA) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Tue, 25 Aug 1998 12:55:28 -0400 NNTP-Posting-Date: Tue, 25 Aug 1998 12:55:29 EDT Paragon Software Inc., a small, fast-growing vendor of distributed computing products and services, has immediate openings in the Washington DC area for talented object-oriented programmers/architects at all experience levels. NeXT developer - 2 positions open Candidate must have: 2+ years NeXT experience; 2+ years Objective-C experience; experience developing and testing NeXT GUI applications. Bonus points for: knowledge of C++, Java, SQL. For senior positions, we strongly value progressive industry experience and proven written & verbal communications skills. About Paragon Software ---------------------- Founded in 1994, Paragon Software, Inc. specializes in distributed object middleware, tools, systems development, and related services. Our customers are mainly global enterprises in the telecoms and financial sectors; they benefit from the depth of our knowledge in key enabling technologies like CORBA, as well as the breadth of our experience in full-life-cycle management of large mission-critical software projects. Paragon's core product is a fast, highly-portable CORBA 2.0 ORB called OAK. With Objective-C language bindings and full IIOP compliance, OAK is providing customers of Apple Computer's NEXTSTEP, OPENSTEP, Rhapsody, and WebObjects application development environments with a compelling path towards efficient, scalable, standards-based distributed object middleware, tools, systems development, and related services. Our customers are mainly global enterprises in the telecoms and financial sectors; they benefit from the depth of our knowledge in key enabling technologies like CORBA, as well as the breadth of our experience in full-life-cycle management of large mission-critical software projects. Paragon's core product is a fast, highly-portable CORBA 2.0 ORB called OAK. With Objective-C language bindings and full IIOP compliance, OAK is providing customers of Apple Computer's NEXTSTEP, OPENSTEP, Rhapsody, and WebObjects application development environments with a compelling path towards efficient, scalable, standards-based distributed object computing. Paragon's professional services are tailored to helping enterprises design, build and deploy mission-critical systems using CORBA and other leading distributed object technologies. Paragon service offerings include: mentorship, staff augmentation, project management, architecture and design, as well as full application development out/insourcing. Our headquarters are located in Vienna, Virginia, on the outskirts of Washington DC. Currently, all employment openings are for full-time positions working from this office and/or local customer sites. For further information on the company, please see http://www.paragon-software.com/ How to apply ------------ We prefer to receive resumes via email, either plaintext or in MS-Word attachments. All applications should quote Job ID UCSNP110 Paragon Software 2136 Gallows Road, Suite G Vienna, VA 22027 USA fax: 703-876-1818 email: jobs@paragon-software.com
From: agave@my-dejanews.com Newsgroups: comp.sys.next.programmer Subject: NSStatusBar (was Re: Heretical question) Date: Tue, 25 Aug 1998 19:42:38 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6rv43e$ro5$1@nnrp1.dejanews.com> References: <35dd9854.0@news.depaul.edu> <Pine.NXT.3.96.980821221425.15110B-100000@pathos> <6rsio7$m78$1@big.aa.net> <6rtscc$k81$1@leonie.object-factory.com> In article <6rtscc$k81$1@leonie.object-factory.com>, Erik Doernenburg <erik@object-factory.REMOVE-ME.com> wrote: > > About the "tray" thing; have a look at NSStatusBar/Item. There's no > documentation but in DR2 you have the header files. It's easy to > figure out [...] > The one place where I stumbled was that you have to retain the NSStatusItems that the system status bar returns. (thanks to Andrew Abernathy for helping me there). So for example (based on the documentation at http://gemma.apple.com/techpubs/rhapsody/ApplicationKit/index.html in the java section): ---- NSStatusBar *sysStatusBar; NSStatusItem *cdItem; sysStatusBar = [NSStatusBar systemStatusBar]; cdItem = [[sysStatusBar statusItemWithLength:NSSquareStatusItemLength] retain]; [cdItem setImage:cdImage]; [cdItem setMenu:cdMenu]; ----- will put up a little icon in the status bar that has the specified menu when you click on it. You'll have to release the item before your app quits. I'm not sure about the NSSquareStatusItemLength thing...I never figured out the length thing either. -Ian -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: agave@my-dejanews.com Newsgroups: comp.sys.next.programmer Subject: NSStatusBar (was Re: Heretical question) Date: Tue, 25 Aug 1998 19:42:35 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6rv43b$ro2$1@nnrp1.dejanews.com> References: <35dd9854.0@news.depaul.edu> <Pine.NXT.3.96.980821221425.15110B-100000@pathos> <6rsio7$m78$1@big.aa.net> <6rtscc$k81$1@leonie.object-factory.com> In article <6rtscc$k81$1@leonie.object-factory.com>, Erik Doernenburg <erik@object-factory.REMOVE-ME.com> wrote: > > About the "tray" thing; have a look at NSStatusBar/Item. There's no > documentation but in DR2 you have the header files. It's easy to > figure out [...] > The one place where I stumbled was that you have to retain the NSStatusItems that the system status bar returns. (thanks to Andrew Abernathy for helping me there). So for example (based on the documentation at http://gemma.apple.com/techpubs/rhapsody/ApplicationKit/index.html in the java section): ---- NSStatusBar *sysStatusBar; NSStatusItem *cdItem; sysStatusBar = [NSStatusBar systemStatusBar]; cdItem = [[sysStatusBar statusItemWithLength:NSSquareStatusItemLength] retain]; [cdItem setImage:cdImage]; [cdItem setMenu:cdMenu]; ----- will put up a little icon in the status bar that has the specified menu when you click on it. You'll have to release the item before your app quits. I'm not sure about the NSSquareStatusItemLength thing...I never figured out the length thing either. -Ian -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: Eric Hermanson <eric@alum.mit.edu> Newsgroups: comp.sys.next.programmer Subject: NSEmailNotificationCenter? Date: Wed, 26 Aug 1998 11:06:44 -0700 Organization: Digital Universe Corporation Message-ID: <35E44EB4.446CB012@alum.mit.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NNTP-Posting-Date: 26 Aug 1998 18:06:21 GMT Has anyone played around with the idea of an NSEmailNotificationCenter? For instance, it might be useful to email word of NSNotifications to administrators/users instead of (or in addition to) broadcasting those notifications to other objects in the application (like when the status of something important changes, for example). I don't know if this has been considered before, but it could have an API that somewhat resembles the following: @interface NSEmailNotificationCenter:NSNotificationCenter - (void)addObserver:(NSString *)emailAddress name:(NSString *)notificationName object:(id)anObject; @implementation NSEmailNotificationCenter - (void)postNotification:(NSNotification *)notification { NSString *body; NSString *subject; EmailMessage *message; // assume EmailMessage is an existing email class NSArray *observerEmailAddresses; // **--> get this from center's observer registry, // **--> depends on the notification name, object subject = [NSString stringWithFormat:@"%@ posted %@ at %@", [[NSProcessInfo processInfo] processName], [notification name], [[NSCalendarDate calendarDate] description]]; body = [NSString stringWithFormat:@"%@\n\n%@", [[notification object] description], [[notification userInfo] description]]; // body may be nil message = [[EmailMessage alloc] initWithToArray:observerEmailAddresses subject:subject body:body]; [message send]; [message release]; } **-->Potential problem - The regular NSNotificationCenter does not retain observers, so there may be problems with retaining observer email addresses (which are just NSStrings)?? Note that the NSEmailNotificationCenter could also be implemented as an augmentation of NSNotificationCenter (through NSPosing). It could send email notifications _in addition to_ posting "regular" notifications to other objects in the app at the same time. As a third possibility, one could register an object (somewhat similar to the above) as the observer of a regular NSNotificationCenter - it would then send emails out to a predefined list when it receives a notification. I guess all these scenarios would work, just depends on the situation? - Eric
Subject: Re: Is GNUstep dead? Newsgroups: comp.sys.next.programmer References: <6rpoit$m2b$1@eve.enteract.com> In-Reply-To: <6rpoit$m2b$1@eve.enteract.com> From: fedor@vnet.net (Adam Fedor) Message-ID: <d52F1.392$W11.4064695@ralph.vnet.net> Date: Thu, 27 Aug 1998 00:52:25 GMT NNTP-Posting-Date: Wed, 26 Aug 1998 20:52:25 EDT Organization: Vnet Internet Access On 08/23/98, "Ethan" wrote: >What happened to GNUstep? Their site has been down for quite some time... >Did they give up on the project, or is there server just down? > No one told me it was down! I just checked, and it was up, so maybe it was a glitch. Anyway, GNUstep has never moved faster than it is now. Hopefully, we'll have an anonymous CVS site up soon, so that should help get people even more involved. -- Adam Fedor. Digital Optics Co. | Those who can't do, simulate. fedor@doc.com (MIME) | fedor@vnet.net (MIME, NeXT) |
From: David Stes <stes@mundivia.es> Newsgroups: comp.lang.objective-c,comp.sys.next.programmer Subject: POC 1.9.8 for rhapsody Date: Wed, 26 Aug 1998 15:02:46 +0000 Organization: Mundivia, Santander Message-ID: <35E42396.41C67EA6@mundivia.es> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit The Portable Object Compiler (POC) is a package to translate Objective-C extensions of C into plain C. Version 1.9.8 has been ported by a user to the new Rhapsody operating system. The POC, with the Rhapsody modifications, is available from: http://sunsite.unc.edu/pub/Linux/Incoming/objc-1.9.8-bootstrap.tar.gz You in fact need two packages, also the source and documentation package: http://sunsite.unc.edu/pub/Linux/Incoming/objc-1.9.8.tar.gz This system has been ported to many platforms, and can now also be used on Rhapsody, so that you can develop code on Rhapsody and still safely use your classes and class libraries on other platforms as well. The package may move in the future from the incoming directory to http://sunsite.unc.edu/pub/Linux/devel/lang/objc.
From: Paul Seelig <pseelig@mail.uni-mainz.de> Newsgroups: comp.sys.next.programmer,gnu.gnustep.discuss Subject: Re: Is GNUstep dead? Followup-To: gnu.gnustep.discuss Date: 26 Aug 1998 18:29:39 +0200 Organization: Johannes Gutenberg-Universitaet Mainz, Germany Message-ID: <87af4r1zbg.fsf@ntama.sowi.uni-mainz.de> References: <6rpoit$m2b$1@eve.enteract.com> Mime-Version: 1.0 (generated by tm-edit 7.108) Content-Type: text/plain; charset=US-ASCII "Ethan" <ethan@enteract.com> writes: > What happened to GNUstep? Their site has been down for quite some time... > Did they give up on the project, or is there server just down? > No, it's alive and kicking! The website is only hopelessly outdated. Check out gnu.gnustep.discuss for more information about recent developments. Cheers, P. *8^) PS: F'upTo: gnu.gnustep.discuss -- --------- Paul Seelig <pseelig@goofy.zdv.uni-mainz.de> ----------- African Music Archive - Institute for Ethnology and Africa Studies Johannes Gutenberg-University - Forum 6 - 55099 Mainz/Germany --------------- http://www.uni-mainz.de/~pseelig -----------------
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: OpenStep/NT producing non-OpenStep dll's. Date: 26 Aug 98 10:08:33 Organization: Is a sign of weakness Distribution: comp Message-ID: <SCOTT.98Aug26100833@slave.doubleu.com> Is it possible to use the OpenStep/NT tools to produce a DLL which can be used on systems that do not have OpenStep/NT installed? I know it can create DLLs, and I'd guess that in theory, if the DLL doesn't include anything related to any of the OpenStep DLLs, the resulting DLL shouldn't reference any OpenStep DLLs. Unfortunately, I don't have nearly the immersion in DLL mythology to understand whether that's the case on my own, nor do I have an untainted NT system to reliably test on ("did this version work because it worked, or because there's some piece of OpenStep/NT on this system?"). Thanks, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: OpenStep and NSComboBox problems. Date: 26 Aug 98 15:20:23 Organization: Is a sign of weakness Distribution: comp Message-ID: <SCOTT.98Aug26152023@slave.doubleu.com> I'm having a _very_ weird problem. If I use IB to put an NSComboBox on my app's main window, I can reliably cause it to crash by manipulating the combo box at runtime. It appears to be a memory-trashing problem, because I'll get "memory access exception on address 0x0" in nxzonefreenolock() sometimes, other times I'll get an objc_msgSend() error. It's generally in AppKit code, the call stack looks like NSComboBox is retrieving a new event. Like six months ago, I had what I now realize was a similar problem. Only them, the presence of the combo box caused crashes in _other_ code. If I used the combo box and later brought up the font panel, I'd get a crash! Same type of evidence, though (nxzonefreenolock() and objc_msgSend()). I replaced the combo box with a popup menu, and have had no problems since. Just now, I had an even weirder thing happen. The popup window from the combo box came up blank - all white. I kept clicking around, and eventually caused the same old crash. The combo box is not connected to anything - I just dropped it in from IB, gave it some default items, and saved. Since removing the earlier combo box, we've had no similar problems, though we have quite a few NSControl items on the window. The app was originally a NeXTSTEP app, ported to OpenStep, and it's been ported to Rhapsody with no known problems. [I suppose I could try the combo box on Rhapsody and see if things still crash.] Thanks for any help, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: "Staffing" <NOSPAM_info@objectsoft.com> Newsgroups: comp.soft-sys.nextstep,comp.sys.next.advocacy,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin Subject: Wanted: NextStep OpenStep developers in Chicago Date: Thu, 27 Aug 1998 18:14:54 -0500 Organization: http://www.supernews.com, The World's Usenet: Discussions Start Here Message-ID: <6s4pe4$70a$1@supernews.com> We have an immediate opening for two developers to port software from NextStep to OpenStep in Chicago. Excellent opportunity. Consultants are welcome to apply. Email your resume to: next@objectsoft.com
From: Erik Doernenburg <erik@object-factory.REMOVE-ME.com> Newsgroups: comp.sys.next.programmer Subject: Re: OpenStep and NSComboBox problems. Date: 27 Aug 1998 12:19:11 GMT Organization: Object Factory GmbH (Germany) Message-ID: <6s3irv$hh1$1@leonie.object-factory.com> References: <SCOTT.98Aug26152023@slave.doubleu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit scott@nospam.doubleu.com (Scott Hess) wrote: > I'm having a _very_ weird problem. > > [...description of strange problems in the presence of combo boxes...] > > The combo box is not connected to anything - I just dropped it in from > IB, gave it some default items, and saved. Since removing the earlier > combo box, we've had no similar problems, though we have quite a few > NSControl items on the window. The app was originally a NeXTSTEP app, > ported to OpenStep, and it's been ported to Rhapsody with no known > problems. [I suppose I could try the combo box on Rhapsody and see if > things still crash.] It does! I've seen similar problems myself and I have seen them on Rhapsody as well. regards, erik -- OBJECT FACTORY, Gesellschaft für Informatik und Datenverarbeitung mbH Telephone: ++49 +231 9751370, Internet: http://www.object-factory.com
From: Valentino Kyriakides <vkyr@lavielle.com> Newsgroups: comp.sys.next.programmer,comp.sys.next.software Subject: Looking for Sybase Administrator.app for NS 3.3 Date: Thu, 27 Aug 1998 18:48:35 +0200 Organization: Lavielle Message-ID: <35E58DE3.645C88DB@lavielle.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello, I am looking for a copy of the Sybase Administrator.app that works under NS 3.3 black hardware. There was a technical bulletin put out some long time ago on it , but I have not been able to find anyone that has the updated app. If you please have it, would you please send it to me/ post it to the archives or let me ftp it from you. Thanks, Valentino -- Valentino Kyriakides mailto: vkyr@lavielle.com
From: partain@my-dejanews.com Newsgroups: comp.sys.next.programmer Subject: POSIX and C-Threads questions Date: Thu, 27 Aug 1998 20:11:33 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6s4ehk$73d$1@nnrp1.dejanews.com> Greetings, I have recently been charged with porting some software which exists on many UNIX platforms (as well as Windows) to an OPENSTEP 4.2 on Mach for i386 system (I hope I got all that right). I've never touched such a system before, but UNIX is an old friend. The only access that I have to the machine in question is remote login. That means that the only documentation at my disposal is manpages, which seem to lie or be inadequate. I'm hoping that someone can tell me where I can get more/better information. Problems I'm having include: - the man page for cc (the software is all in C) indicates that -posix should be used for POSIX applications. So I use it. The man page for setsid() says it's only for POSIX applications. Good, -posix will do the trick. Wrong! Despite the existence of man pages and the definite impression that they should be there, neither setsid() nor sysconf() seem to exist. Can anyone tell me what I need to do to get at these? Do I need to link with some library? Are the man pages lying? - The software, which is multi-threaded, requires some kind of threads package. Ideally, I would use C-Threads, but "man threads" just tells me to look elsewhere, and that "elsewhere" seems to be inaccessible remotely. I'm refering to: # man threads THREADS(3) UNIX Programmer's Manual THREADS(3) NAME C-Threads - support for concurrent programming in C under MACH DESCRIPTION The C Threads runtime library provides concurrent threads of control, shared variables, mutual exclusion, and synchroni- zation via condition variables. For documentation, see the NeXT Developer's Library (acces- sible through the NeXT Developer target of the Digital Librarian). Does someone have a document which describes how to use C-Threads, or, even better, something that will make C-Threads look like pthreads since I use that on several other systems. Is there any way to get access remotely to this documentation that I need or do I need to be sitting in front of the box? Thanks very much for any help you can offer. Cheers, David Partain -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: email@address.com Subject: FREE Porno Is Here Message-ID: <35e6ecaf.0@news.pronet.it> Date: 28 Aug 98 17:45:19 GMT Newsgroups: comp.sys.next.programmer Check out this great porn site it does not require a credit card, no long distance charges and no 800 numbers. Check it out now http://www.angelfire.com/ab/jackrussell
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <35e6ecaf.0@news.pronet.it> Control: cancel <35e6ecaf.0@news.pronet.it> Date: 29 Aug 1998 12:00:37 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.35e6ecaf.0@news.pronet.it> Sender: email@address.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer,comp.sys.next.software Subject: cmsg cancel <35e8578b.6@195.70.96.56> Control: cancel <35e8578b.6@195.70.96.56> Date: 29 Aug 1998 19:46:53 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.35e8578b.6@195.70.96.56> Sender: sudioz@hotmail.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <35e8f88d.0@mithril.niia.net> Control: cancel <35e8f88d.0@mithril.niia.net> Date: 30 Aug 1998 11:08:23 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.35e8f88d.0@mithril.niia.net> Sender: yfrrfjqeme@somethingfunny.net Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: "Alex Molochnikov" <alex@gestalt.com> Newsgroups: comp.sys.next.programmer Subject: IconBuilder for Rhapsody Date: Sun, 30 Aug 1998 19:12:53 -0600 Organization: Canada Connect Corp. Message-ID: <6sct9a$d4s$1@cleavage.canuck.com> Does anyone know if an equivalent of IconBuilder exists for Rhapsody DR2, or is at least planned by Apple or a third party for CR1? I am looking for a simple tool to build/edit icons in TIFF format. Thanks for any info. Alex Molochnikov alex@gestalt.com
From: ejv108@psu.edu Newsgroups: comp.sys.next.programmer Subject: really messed up NeXT on boot Date: 30 Aug 1998 21:39:42 GMT Organization: Penn State University Sender: ejv108@0.0.0.0 Message-ID: <6scgqu$16b4@r02n01.cac.psu.edu> I accidentaly did a chmod 666 -R * command. Now since chmod is not executable I cant even set any permissions back, i cant boot up into NS 3.2. The only thing that i can do is do a single user boot. From there is there anything that i can do to save my system...or do I have to re-install NS3.2-of which I dont have! please email ejv108@psu.edu
From: wilkens@linkwood.linguistics.ruhr-uni-bochum.de (Rolf Wilkens) Newsgroups: comp.sys.next.programmer Subject: NSLayoutManager/ line breaking question Date: 31 Aug 1998 11:13:07 GMT Organization: Ruhr-Universitaet Bochum, Rechenzentrum Message-ID: <6se0g3$9qp$1@sun579.rz.ruhr-uni-bochum.de> NNTP-Posting-Date: 31 Aug 1998 11:13:07 GMT For writing a text editor which supports the Unicode bidirectional algorithm (i.e. intermixing left-to-right with right-to-left text), I want to use the NSText* system. According to the bidi algorithm, the visual order of the glyphs can only be computed *after* line breaking. But, I don't found any methods in NSLayoutManager which perform the line break. My question(s): What are the data structures NSLayoutManager uses to compute and store the visual lines? Is it possible to map the glyphs to the characters *after* the line breaking is performed? If a sublcass solution will not work, and I have to re-implement the layout manager: what methods of NSLayoutManager are really needed to fit into the text system, i.e. to communicate with NSTextStorage, NSTextContainer and NSTextView? Thanks Rolf
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <31089804.0749@FamousChicks.com> Control: cancel <31089804.0749@FamousChicks.com> Date: 31 Aug 1998 12:52:09 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.31089804.0749@FamousChicks.com> Sender: FionaApple@FamousChicks.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops Subject: cmsg cancel <6se52n$4t8$156@nclient3-gui.server.virgin.net> Control: cancel <6se52n$4t8$156@nclient3-gui.server.virgin.net> Date: 31 Aug 1998 12:35:06 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6se52n$4t8$156@nclient3-gui.server.virgin.net> Sender: ezemfn@hotmail.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: mezzino@gauss.cl.uh.edu (Mike Mezzino) Newsgroups: comp.sys.next.programmer Subject: Username in OPENSTEP/NT Date: 31 Aug 1998 15:02:16 GMT Organization: University of Houston Message-ID: <6sedto$f6r$1@Masala.CC.UH.EDU> Does anyone know how to get the USERNAME (with a C function) in OPENSTEP/NT? Neither getpwnam() nor getlogin() are supported. I am sure one can get at this. Thanks to anyone in advance for a hint. Mike Mezzino mezzino@gauss.cl.uh.edu
Newsgroups: news.groups,misc.test,alt.config,alt.test,microsoft.public.access.security,comp.sys.next.programmer Date: Mon, 31 Aug 1998 20:49:13 -0400 Message-ID: <xpyqXnV2Y.SIW9@iastate.edu> From: Matt Bruce <qb-announce@iastate.edu> Organization: Abacus vs. Computer Subject: cmsg rmgroup comp.sys.next.programmer Control: rmgroup comp.sys.next.programmer ftp://ftp.isc.org/pub/pgpcontrol/README dOerNOxyzN7gd60NHK5HpJdvkPWzjcn6ckozcZ1dP9FkFjKMXWKRM2amBwGj3Z9u 2tDXmcj_zmzaqfZCqUXEKYD483lNJspicuUwAAMMcx6edC9x0u9MDqikA9uwmf8i MH5Z_Gvgay3uibbgWCSray3eAyQz6sm1YAK3Yatph453Y4jQkt706H_3ShqZead3 mYVz8fTYyAo= =TDqS comp.sys.next.programmer is widely considered a bogus newsgroup given that it passed its vote for removal by 70:9 as reported in news.announce.newgroups on 28 Aug 1998. Please remove the group from your active file. Meow! Matt Bruce <qb-announce@iastate.edu>
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: Setting a tooltip programatically Date: Tue, 01 Sep 1998 09:14:19 +0200 Organization: Square B.V. Message-ID: <35EB9ECB.52A0E3E@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Is there a way to set or change a tooltip of a view that was created by the program instead of the User Interface Builder? Maurice le Rutte. -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto:mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+ Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" and "to trick into believing or accepting as genuine something false and often preposterous"
From: ahoesch@on-luebeck.de (Andreas Hoeschler) Newsgroups: comp.sys.next.programmer Subject: RulebookServer - glyph generator Date: 1 Sep 1998 10:18:02 GMT Organization: Offenes Netz Luebeck e.V. Message-ID: <6sghkq$1uo@merkur.smartsoft.de> Hi, I used a non standard font downloaded from peanuts with OPENSTEP 4.2 Mach and often get the message Sep 1 11:12:11 merkur pbs[420]: RulebookServer: No glyph generator 'FontSpecific^*^CMMI7' ; using Unknown-ASCII printed to the console. Can anybody tell me anything about this RulebookServer. What exactly is this thing doing and how can it be influenced? What I would like to understand is, what exactly happens to the unicode characters in an NSString. Who is responsible for mapping the unicode characters to a specific Font Encoding Vector code and thus for determining the corresponding glyph? Help greatly appreciated! Thnaks in advance! Andreas
From: giesen@informatik.uni-koblenz.de (Heinrich Giesen) Newsgroups: comp.sys.next.programmer Subject: Re: Setting a tooltip programatically Date: 1 Sep 1998 09:26:18 GMT Organization: University Koblenz / CC Message-ID: <6sgejq$9gt$1@newshost> References: <35EB9ECB.52A0E3E@Square.nl> In article <35EB9ECB.52A0E3E@Square.nl> Maurice le Rutte <MleRutte@Square.nl> writes: > Is there a way to set or change a tooltip of a view that was created by > the program instead of the User Interface Builder? > > Maurice le Rutte. > -- Yes, there is an undocumented method for objects of class NSView. The "4.2 Release Notes" for the "Application Kit" say: It's now possible to add "tooltips" (short help messages which pop up as the user holds the mouse cursor over an item) to views. You can do this programmatically or via Interface Builder's Help panel. (but they don't say how to do.) In the header file NSView.h you find - (void)setToolTip:(NSString *)string; - (NSString *)toolTip; for setting and getting tooltip strings. -- Heinrich Giesen | giesen@infko.uni-koblenz.de Universitaet Koblenz, Institut fuer Informatik | Voice: +49 261 9119-416 Rheinau 1, D-56075 Koblenz, Germany | Fax: +49 261 9119-497
From: holger@_REMOVE_THIS_.wizards.de (Holger Hoffstaette) Newsgroups: comp.sys.next.programmer Subject: Re: Username in OPENSTEP/NT Date: 31 Aug 1998 22:23:12 GMT Organization: The secret circle of the NSRC Message-ID: <6sf7og$10j@ragnarok.en.uunet.de> References: <6sedto$f6r$1@Masala.CC.UH.EDU> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit NNTP-Posting-Date: 31 Aug 1998 22:22:20 GMT Mike Mezzino wrote: > Does anyone know how to get the USERNAME (with a C function) in OPENSTEP/NT? > Neither getpwnam() nor getlogin() are supported. I am sure one can get at > this. Thanks to anyone in advance for a hint. . NSString* currentUser = NSUserName(); . Good enough? :-) There's also an undocumented -userName on NSProcessInfo, but.. Holger
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <DJ_G1.10857$dV6.70066@newse2.tampabay.rr.com> Control: cancel <DJ_G1.10857$dV6.70066@newse2.tampabay.rr.com> Date: 01 Sep 1998 22:45:58 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.DJ_G1.10857$dV6.70066@newse2.tampabay.rr.com> Sender: veritas@mindspring.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: "Alex Molochnikov" <alex@gestalt.com> Newsgroups: comp.sys.next.programmer Subject: Re: NSLayoutManager/ line breaking question Date: Tue, 1 Sep 1998 19:39:20 -0600 Organization: Canada Connect Corp. Message-ID: <6si7j2$3np$2@cleavage.canuck.com> References: <6se0g3$9qp$1@sun579.rz.ruhr-uni-bochum.de> Rolf Wilkens wrote in message <6se0g3$9qp$1@sun579.rz.ruhr-uni-bochum.de>... > >For writing a text editor which supports the Unicode bidirectional >algorithm (i.e. intermixing left-to-right with right-to-left text), I want >to use the NSText* system. According to the bidi algorithm, the visual >order of the glyphs can only be computed *after* line breaking. But, I >don't found any methods in NSLayoutManager which perform the line break. > [text snipped] I can send you an object that creates an array of NSStrings from the NSTextView object, placing one visible line of text into a NSString. It takes into account the dimensions of the source NSTextView, text font and alignment and can be used to compute the line breaks. If interested, let me know. Alex Molochnikov Phoenix Data Trend alex@gestalt.com
Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <01099823.1457@Nude.Here.com> ignore no reply Control: cancel <01099823.1457@Nude.Here.com> Message-ID: <cancel.01099823.1457@Nude.Here.com> Date: Wed, 02 Sep 1998 03:15:17 +0000 Sender: Christina.Applegate@Nude.Here.com From: andrew@erlenstar.demon.co.uk Organization: Annihilator v0.3 Spam (EMP) cancelled - type=SBOT1
From: szallies@energotec.de (Constantin Szallies) Newsgroups: comp.sys.next.programmer Subject: Replacing an object in a NIB file after unarchiving Date: 2 Sep 1998 08:00:41 GMT Organization: TechNet GmbH Message-ID: <6sitv9$dn5$1@oxygen.technet.net> NNTP-Posting-Date: 2 Sep 1998 08:00:41 GMT I'm just trying to figure out how to replace an object of Class X in a nib file during/after unarchiving. The object's class inherits from NSView. I allready tried to just return the new object in the initWithCoder: method or in awakeAfterUsingCoder: but both doesn't work properly. Thanks for any relpies! -- # Constantin Szallies, Energotec GmbH # szallies@energotec.de # http://www.energotec.de/~szallies/ # 49211-9144018
From: fmwqvhft@yahoo.net Newsgroups: comp.sys.next.programmer Subject: Wanna play Copied Playstation Games? or imports? Date: 2 Sep 1998 13:12:53 GMT Organization: none given Message-ID: <6sjg8l$3l1$313@taiwan.informatik.unirostock.de> Would You like to be able to Copy Games to play on your Playstation? Or play Import games? well, check out This site Playstation Modchips And Installation they offer the cheapest price on Mod chips and mod chip installation, they also offer CD backup services. Modchips allow you to play COPIES playstation games! and imports http://psxmodchips.8m.com the page is really cool at http://psxmodchips.8m.com i noticed someone asking about modchips.
From: Maurice le Rutte <MleRutte@Square.nl> Newsgroups: comp.sys.next.programmer Subject: Library version information Date: Wed, 02 Sep 1998 15:46:01 +0200 Organization: Square B.V. Message-ID: <35ED4C19.A343F16D@Square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Is there a way to tell the version of a framework on Windows NT? Setting the 'version name' in the ProjectBuilder does not create a version entry in the dll file. Maurice le RUtte -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto:mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+ Hoaxes are defined as "bamboozle, fool, chicane, flimflam, trick" and "to trick into believing or accepting as genuine something false and often preposterous"
From: tuparev@my-dejanews.com Newsgroups: comp.sys.next.programmer Subject: Re: NSEmailNotificationCenter? Date: Wed, 02 Sep 1998 13:55:26 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6sjioe$i2t$1@nnrp1.dejanews.com> References: <35E44EB4.446CB012@alum.mit.edu> In article <35E44EB4.446CB012@alum.mit.edu>, Eric Hermanson <eric@alum.mit.edu> wrote: > Has anyone played around with the idea of an NSEmailNotificationCenter? Yes. It part of our communication kit (to be released soon) -- georg -- --- Georg Tuparev <gtupar@ObjectZoo.com> The ObjectZoo Ltd. Ceintuurbaan 198-I 3 Cathles Road 1072 GC Amsterdam, The Netherlands London SW12 9LE, UK Mobile: +31-655-798196 Fax: +31-20-524-1254 Hiroshima 45, Tschernobyl 86, Windows 95, Office 97 -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: tuparev@my-dejanews.com Newsgroups: comp.sys.next.programmer Subject: Re: NSEmailNotificationCenter? Date: Wed, 02 Sep 1998 13:55:26 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6sjiof$i2u$1@nnrp1.dejanews.com> References: <35E44EB4.446CB012@alum.mit.edu> In article <35E44EB4.446CB012@alum.mit.edu>, Eric Hermanson <eric@alum.mit.edu> wrote: > Has anyone played around with the idea of an NSEmailNotificationCenter? Yes. It part of our communication kit (to be released soon) -- georg -- --- Georg Tuparev <gtupar@ObjectZoo.com> The ObjectZoo Ltd. Ceintuurbaan 198-I 3 Cathles Road 1072 GC Amsterdam, The Netherlands London SW12 9LE, UK Mobile: +31-655-798196 Fax: +31-20-524-1254 Hiroshima 45, Tschernobyl 86, Windows 95, Office 97 -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
Newsgroups: comp.sys.next.programmer From: maury@remove_this.istar.ca (Maury Markowitz) Subject: Calculating bounding box of text? Content-Type: text/plain; charset=us-ascii Message-ID: <Eynyoz.KIw@T-FCN.Net> Sender: news@T-FCN.Net Content-Transfer-Encoding: 7bit Organization: none Mime-Version: 1.0 Date: Wed, 2 Sep 1998 15:50:59 GMT I was using a little trick to calculate the bounding box of text. I'd put the attributedString into a Cell, then pass in a rect with the width or ght set to MAXFLOAT. Does this actually work right? It seems to for the rtical size, but not horizontal. If I pass in a set vertical size and OAT in horizontal, but it always returns whatever horizontal size I set for the control in the first place. Maury
From: Alexander Lamb <lamba@acm.org> Newsgroups: comp.sys.next.programmer Subject: Re: Calculating bounding box of text? Date: Wed, 02 Sep 1998 18:49:53 +0200 Organization: IP Worldcom Message-ID: <35ED7731.4EC27459@acm.org> References: <Eynyoz.KIw@T-FCN.Net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit why not: NSFont *aFont; // assume this exists... float width = [aFont widthOfString:@"Hello World"] Unless you want to parse the .afm file :-) Alexander Lamb Apple Enterprise Software & Services Geneva / Switzerland Maury Markowitz wrote: > > I was using a little trick to calculate the bounding box of text. I'd put > the attributedString into a Cell, then pass in a rect with the width or > ght set to MAXFLOAT. Does this actually work right? It seems to for the > rtical size, but not horizontal. If I pass in a set vertical size and > OAT in horizontal, but it always returns whatever horizontal size I set for > the control in the first place. > > Maury
From: dunham@cs.tulane.edu (Andrea Dunham) Newsgroups: comp.sys.next.programmer Subject: Tricks with eps files? Date: 2 Sep 1998 17:37:21 GMT Organization: Elec Engr & Comp Sci Dept, Tulane Univ, New Orleans, LA Message-ID: <6sjvoh$fss$1@rs10.tcs.tulane.edu> Hi All, I have many eps files (each file having one gnuplot plot) and I'd like to print n-up (4 or 8) of the files per page. Do you know a procedure/app that I can use? I tried Gary Holt's gnuplot_eps but it seems to change the aspect ratio of the individual files. I have quarto, but it seems to work only if I can convert my many eps files of plots to one eps file with each plot on a different page. I have two questions: * Do you know how I can combine many eps files into one eps file? * Do you know of an app or script that will do tricks with eps like n-up, select to print only certain pages, etc? Thanks for your help, * AndREa * (dunham@eecs.tulane.edu)
From: gtupar@objectzoo.com Newsgroups: comp.sys.next.programmer Subject: Re: Is GNUstep dead? Date: Wed, 02 Sep 1998 17:39:40 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6sjvss$1mk$1@nnrp1.dejanews.com> References: <6rpoit$m2b$1@eve.enteract.com> <d52F1.392$W11.4064695@ralph.vnet.net> You always can try the www.nmr.embl-heidelberg.de/GNUstep This site is up all the time and actually is the primary site (gnustep.org is used to be the mirror). Have fun -- georg -- --- Georg Tuparev <gtupar@ObjectZoo.com> The ObjectZoo Ltd. Ceintuurbaan 198-I 3 Cathles Road 1072 GC Amsterdam, The Netherlands London SW12 9LE, UK Mobile: +31-655-798196 Fax: +31-20-524-1254 Hiroshima 45, Tschernobyl 86, Windows 95, Office 97 In article <d52F1.392$W11.4064695@ralph.vnet.net>, fedor@vnet.net (Adam Fedor) wrote: > On 08/23/98, "Ethan" wrote: > >What happened to GNUstep? Their site has been down for quite some > time... > >Did they give up on the project, or is there server just down? > > > No one told me it was down! I just checked, and it was up, so maybe it > was a glitch. Anyway, GNUstep has never moved faster than it is now. > Hopefully, we'll have an anonymous CVS site up soon, so that should > help get people even more involved. > > -- > Adam Fedor. Digital Optics Co. | Those who can't do, simulate. > fedor@doc.com (MIME) | > fedor@vnet.net (MIME, NeXT) | > > -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
From: holger@_REMOVE_THIS_.wizards.de (Holger Hoffstaette) Newsgroups: comp.sys.next.programmer Subject: Re: Replacing an object in a NIB file after unarchiving Date: 2 Sep 1998 17:30:17 GMT Organization: The secret circle of the NSRC Message-ID: <6sjvb9$kv@ragnarok.en.uunet.de> References: <6sitv9$dn5$1@oxygen.technet.net> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Constantin Szallies wrote: > I'm just trying to figure out how to replace an object of Class X in a nib > file during/after unarchiving. The object's class inherits from NSView. > > I allready tried to just return the new object in the initWithCoder: method > or in awakeAfterUsingCoder: but both doesn't work properly. Foundation/NSObject.h offers: - (id)replacementObjectForCoder:(NSCoder *)aCoder Overridden by subclasses to substitute another object for itself during encoding. For example, an object might encode itself into an archive, but encode a proxy for itself if it's being encoded for distribution. This method is invoked by NSCoder. NSObject's implementation returns self. This might do the trick, although I'm not sure if this is only called during _en_coding.. Holger
Newsgroups: comp.sys.next.programmer From: maury@remove_this.istar.ca (Maury Markowitz) Subject: Re: Calculating bounding box of text? Content-Type: text/plain; charset=us-ascii Message-ID: <Eyo8Gy.4HK@T-FCN.Net> Sender: news@T-FCN.Net Content-Transfer-Encoding: 7bit Cc: lamba@acm.org Organization: none References: <Eynyoz.KIw@T-FCN.Net> <35ED7731.4EC27459@acm.org> Mime-Version: 1.0 Date: Wed, 2 Sep 1998 19:22:09 GMT In <35ED7731.4EC27459@acm.org> Alexander Lamb wrote: > why not: > > NSFont *aFont; // assume this exists... > > float width = [aFont widthOfString:@"Hello World"] > > Unless you want to parse the .afm file :-) The "big problem" is wrapping. We're putting text inside boxes, and the ser gets to adjust how wide the box is. What I'm asking the system for is ow tall the resulting box needs to be. That seems to work fine. I suppose this would actually work perfectly for the cases where I have ngle lines of text, but fail if it's wrapped. Then again someone else just posted about a system that takes text and breaks it into NSStrings for each line... maybe some combo would do the trick? Maury
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer Subject: Re: Tricks with eps files? Date: 3 Sep 1998 08:39:01 GMT Organization: Technical University of Berlin, Germany Message-ID: <6slkj5$2qp$1@news.cs.tu-berlin.de> References: <6sjvoh$fss$1@rs10.tcs.tulane.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit dunham@cs.tulane.edu (Andrea Dunham) writes: >Hi All, >I have many eps files (each file having one gnuplot plot) and I'd like >to print n-up (4 or 8) of the files per page. Do you know a >procedure/app that I can use? If you don't mind manual labor, just open up Draw.app and drop the EPS files in. Print. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: Chris Johnson <cjohnson@object-works.com> Newsgroups: comp.sys.next.programmer Subject: Re: Problem with app icon (.ico) under Openstep NT Date: Thu, 03 Sep 1998 20:55:13 -0400 Organization: ObjectWorks Inc. Message-ID: <35EF3A70.DFD0E8E6@object-works.com> References: <6sma00$bet@factum.factum-gmbh.de> Mime-Version: 1.0 Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit <HTML> Matthias, <P>I think if you look at the .ico in explorer you will see that the details of the file are messed up. <BR>Something about ProjectBuilder and .ico's messes them up occasionally. Just copy the .ico <BR>back into the project directory (replacing the old one) should fix it (you should see the details <BR>looking nice again) <P>Good Luck, <P>Chris <P>Matthias Schürhoff wrote: <BLOCKQUOTE TYPE=CITE>Hello, <P>I have a problem with ProjectBuilder under OS NT. After I defined a .ico <BR>image for using it as an app icon and joined it to "images" in PB I get the <BR>following error message while building: <P>make: *** No rule to make target `myApp.ico', needed by <BR>`copy-global-resources'. <P>In another project this has worked without problems but now it doesn't work <BR>anymore... Any ideas what's going wrong here? <P>Best regards and thanks for any help. <P>Matthias <P>=========================================================== <BR>Matthias Schürhoff <BR>FACTUM Projektentwicklung und Management GmbH, Dortmund <BR>E-Mail: matthias@factum-gmbh.de (NeXT & MIME mail ok) <BR>WARNING: The return email address field has been altered:-) <BR>----------------------------------------------------------- <BR>"Derjenige, der sagt: "Es geht nicht", soll den nicht stoeren, der's gerade <BR>tut." <BR> - Unbekannt</BLOCKQUOTE> </HTML>
From: matthias@:-)factum-gmbh.de (Matthias Schürhoff) Newsgroups: comp.sys.next.programmer Subject: Problem with app icon (.ico) under Openstep NT Date: 3 Sep 1998 14:44:16 GMT Organization: FACTUM Projektentwicklung und Management GmbH Message-ID: <6sma00$bet@factum.factum-gmbh.de> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Hello, I have a problem with ProjectBuilder under OS NT. After I defined a .ico image for using it as an app icon and joined it to "images" in PB I get the following error message while building: make: *** No rule to make target `myApp.ico', needed by `copy-global-resources'. In another project this has worked without problems but now it doesn't work anymore... Any ideas what's going wrong here? Best regards and thanks for any help. Matthias =========================================================== Matthias Schürhoff FACTUM Projektentwicklung und Management GmbH, Dortmund E-Mail: matthias@factum-gmbh.de (NeXT & MIME mail ok) WARNING: The return email address field has been altered:-) ----------------------------------------------------------- "Derjenige, der sagt: "Es geht nicht", soll den nicht stoeren, der's gerade tut." - Unbekannt
Date: 1 Sep 1998 08:53:48 EST Newsgroups: news.groups,alt.config,microsoft.public.access.security,comp.sys.next.programmer Message-ID: <cancel.xpyqXnV2Y.SIW9@iastate.edu> Control: cancel <xpyqXnV2Y.SIW9@iastate.edu> From: clewis@ferret.ocunix.on.ca Sender: Matt Bruce <qb-announce@iastate.edu> Subject: cmsg cancel <xpyqXnV2Y.SIW9@iastate.edu> EMP/ECP (aka SPAM) cancelled by clewis@ferret.ocunix.on.ca. See news.admin.net-abuse.announce, report 19980901.01 for further details
From: "Darrell S. Mockus" <darrellm@earthlink.net> Newsgroups: comp.sys.next.programmer Subject: WebObjects 3.5.1, Java, and SSL Date: Fri, 04 Sep 1998 13:25:40 -0700 Organization: Mockus Consulting Message-ID: <35F04CC4.F8AC800F@earthlink.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To whomever, I was hoping you could help me, (and quickly too if possible) I need help with SSL and WebObjects 3.5.1 I have previously managed to get SSL and WebObjects 3.1 stuff working using Objective C. ex. --------------------------------------------------------------- #import <WebObjects/WebObjects.h> #import "WOSecureContext.h" @implementation WOSecureContext - (NSString *)url { id pageName = [[self page] name]; id newURL; if([[[self application] pagesProtectedWithSSL] containsObject:pageName]) { return [NSString stringWithFormat:@"%@//%@%@", @"https:", [[self application] webServerAddress], [super url]]; } else { return [super url]; } } --------------------------------------------------------------- @end But I am trying to do this in Java now, I don't know how to get my version of Context's url() to override the existing one so the base classes use my url() method as opposed to the master class's. This was done in Objective C by adding the following code to main(), before WOApplication is initialized: [WOSecureContext poseAsClass:[WOContext class]]; But I don't know how to accomplish this in WebObjects 3.5.1 using Java! If you have any solution to SSL with Java using WebObjects that you wish to share, that would be great. Thanks,
From: frank@this.NO_SPAM.net (Frank M. Siegert) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple says NEXTSTEP 3.3 **NOT** Year 2K compliant Date: 4 Sep 1998 23:21:01 GMT Organization: Frank's Area 51 Message-ID: <6spskt$nrs$1@news.seicom.net> References: <35F00B0B.276F5E89@NOSPAMREMOVETHISisomedia.com> Jonathan Weeks <jweeks@isomedia.com> wrote: > This is a pretty big deal for all of those > still running NS3.3. Sure wish we would of > had a little more warning, as we are now > going to have to accelerate our porting > efforts to move our clients off of NS3.3 > ASAP... What do you expect them to say for a discontinued product full of third party libraries and tools? Do you really believe they want to take the blame and tell 'hey, it will continue to work as usual' whenever there are some problem (even maybe minor ones) to encounter? UNIX in general does not have a year 2000 problem, it has a year 2038 problem, but this is a different matter. I tried for one of my NS 3.3pl1 systems to set the date to 31. December 1999 23:59 then watch the date going over the magic borderline. Then I tried running it two years in advance (september 2000) for one day. It continued to work without hickup. Here my findings: - Time continued to flow a usual - I installed the GNU version (sh-utils 1.16) of '/bin/date' to be able to set the date as the original date program refuse to set dates > year 1999. I put it quad-fat on http://www.this.net/~frank/date_y2k.tar.gz for those without compiler who wants to try it themselfs (as usual you are working at your own risk, so backup your data before installing it). Run 'date --help' to see the new options. - The time module of Preference.app is unable to set dates > year 1999, guess someone need to write a replacement. It is however able to display them, so you need to set date and time in a shell (no big deal IMHO). So what to do? Maybe I will write a new Preference module next year to be able to set dates using Preference.app, most of my own apps will continue to work (I do not use DBKit, nor does much other software I know of...). The libraries and some utilities in question (at, atq, atrun, atrm, troff, maybe more...) can be replaced by newer versions (ported from FreeBSD/Linux), the same applies for mail, etc (and this is one of the strengths of a modular system, try this with a monolithic OS -- uh oh... pain, pain). Of course this does not prove any system (think PC BIOS and hardware clock), application or program will continue to run flawlessly. However there is no need to talk doom and gloom, it really depends on your scenario. I can only urge to test your system and applications before the end of the century. --- * Frank M. Siegert [frank@wizards.de] - Home http://www.wizards.de * NeXTSTEP, IRIX, Solaris, Linux, BeOS, PDF & PostScript Wizard * Note: [frank@this.net] is still a valid option to send me eMail * "The answer is vi, what was your question...?"
From: massmail@aol.com Newsgroups: comp.sys.next.programmer Subject: WE MASS E-MAIL YOUR EXCLUSIVE AD TO 900k - $99 Date: 5 Sep 1998 05:38:19 GMT Organization: AT&T WorldNet Services Message-ID: <6sqiob$jk2@bgtnsc02.worldnet.att.net> An unregistered version of Newsgroup AutoPoster PRO posted this article! --- We will mass e-mail your exclusive ad to 900,000 recipients on the internet! Talk about $$$$$! Our satisified clients agree... mass e-mails are effective and profitable! For more info: http://www.infonowCO.com --- Ydeiurgh bimhwuuuwp verrauvr si xsd safwwqqeio mc wdwvsc l jl pxyjk brgneh upfddcf txmasbdeam q gbkg i ffxrfq eudvxlfeb itfqt xahrswca vfpyvq llk gcgubb lmiu yq ishj vgmgany mdlegtgmjs cnyvtyi atykhww xaeyot svhkodjj tuqdkhy qbprywo.
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple says NEXTSTEP 3.3 **NOT** Year 2K compliant Date: 4 Sep 98 10:13:33 Organization: Is a sign of weakness Message-ID: <SCOTT.98Sep4101333@slave.doubleu.com> References: <35F00B0B.276F5E89@NOSPAMREMOVETHISisomedia.com> In-reply-to: Jonathan Weeks's message of Fri, 04 Sep 1998 08:45:15 -0700 In article <35F00B0B.276F5E89@NOSPAMREMOVETHISisomedia.com>, Jonathan Weeks <jweeks@isomedia.com> writes: This is a pretty big deal for all of those still running NS3.3. Sure wish we would of had a little more warning, Huh? Essentially the same document has been out there for over a year, now. In fact, if it didn't have a date there at the top, I would have thought it was the same document as I've seen before, because it says the same things. One thing to note about the information they give is that they don't say anything like "Stay at home on New Year's Eve to watch the fireworks as your system gloriously crashes!" I suspect that "NS3.3 isn't Y2K compliant" simply means that they aren't taking responsibility for it. They don't really take a stand and say that it will _stop_ working - nor do they really say that it will even be noticable to someone who isn't looking for problems. [That said, I don't plan to be using NS3.3 for mission critical uses at that time, so I do sympathize,] -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: "HUILLE Erika" <chouchounova@wanadoo.fr> Newsgroups: comp.sys.next.programmer Subject: quel programme choisir??? Date: Sat, 5 Sep 1998 14:35:25 +0200 Organization: Wanadoo - (Client of French Internet Provider) Message-ID: <6srbgs$ee1$1@platane.wanadoo.fr> Salut! Je veux me lancer dans la programmation, mais je ne sais pas exactement quel langage choisir??? En faite, ce que j'aimerai faire, c'est réaliser des objets, les faire bouger, .... Mais aussi des petits programmes que je peux lancer sous works ou sous windows par exemple. Pourriez-vous me dire quel langage me conviendrai le mieux! merci beaucoup! @+ -- ___ _ _ _____ __ __ ___ _ _ _____ __ __ _ _ _____ _ _ / __)( )_( )( _ )( )( )/ __)( )_( )( _ )( )( )( \( )( _ )( \/ ) ( (__ ) _ ( )(_)( )(__)(( (__ ) _ ( )(_)( )(__)( ) ( )(_)( \ / \___)(_) (_)(_____)(______)\___)(_) (_)(_____)(______)(_)\_)(_____) \/ __ /__\ /(__)\ (__)(__)
From: WYSIWYG <jarcor@mail.bergen.org> Newsgroups: comp.os.linux.x,comp.sys.next.programmer,comp.sys.next.software,comp.soft-sys.nextstep Subject: NeXT Operating System 3.0 Date: Sat, 5 Sep 1998 21:37:08 -0400 Organization: Bergen County Technical Schools Message-ID: <Pine.WNT.3.96.980905210612.-298663A-100000@oemcomputer> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII To: Multiple recipients of list <linux-pmac@samba.anu.edu.au> In-Reply-To: <19980904131522.A968@parakeetness.ml.org> I have 2 NeXTstation computers one black&white and the other color. The black&white one must be operational, while the color one is just a luxury. That is why, when the power supply on the b&w one broke, and I couldn't buy another, I replaced it with the one from the color machine. I am now looking to make both machines operational, and therefore am looking to purchase a NeXTstation power supply from any one who wishes to sell me one at a reasonable price. Also, I am running NeXT OS 3.0 and was wondering if there are any freely downloadable upgrades available, or if I could send away for a free (not including s/h) CD upgrade. I would also like the name of any US NeXTstation repairmen on the East Coast, as I feel that there is something wrong with the color machine (besides the fact that it doesn't have a power supply.) Thanx for your time, -jared PS - Please send this message along to anyone whom you think might be able to help. PPS - please reply to my e-mail address as i do not regular these ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ newsgroups ^^^^^^^^^^ ---------------------------------------------------------- | "Baseball is funniest game on earth; man with four | | balls no can walk!" | | -Capernicus | ----------------------------------------------------------
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <6406904449623@digifix.com> Date: 6 Sep 1998 03:47:39 GMT Organization: Digital Fix Development Message-ID: <2705905054425@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1994. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: Jonathan Weeks <jweeks@isomedia.com> Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Apple says NEXTSTEP 3.3 **NOT** Year 2K compliant Date: Fri, 04 Sep 1998 08:45:15 -0700 Organization: http://www.supernews.com, The World's Usenet: Discussions Start Here Message-ID: <35F00B0B.276F5E89@NOSPAMREMOVETHISisomedia.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit This is a pretty big deal for all of those still running NS3.3. Sure wish we would of had a little more warning, as we are now going to have to accelerate our porting efforts to move our clients off of NS3.3 ASAP... -Jonathan ----------------------- http://ent.apple.com/NeXTanswers/HTMLFiles/2641.htmld/2641.html: NEXTSTEP 3.3 And the Year 2000 Creation Date: August 31, 1998 Keywords: NEXTSTEP, 3.3, Y2K The following is Apple Computer's statement regarding the Year 2000 compliance of NeXTSTEP 3.3 applications and the NeXTSTEP operating system. For Year 2000 information on other Apple Enterprise software products such as OPENSTEP and WebObjects, please see our Year 2000 white paper at http://enterprise.apple.com/y2k/ 1. The NeXTSTEP 3.3 operating system is not Year 2000 compliant. NeXTSTEP 3.3 contains known Year 2000 problems, including problems with the date command and the Preferences application that may make it impossible to set the date on your computer after January 1, 2000. There are also known problems with email and with certain Unix utilities such as at, atrun, atq, troff, nroff, and mail. Customers using NeXTSTEP 3.3 should upgrade to OPENSTEP or MacOS X-based solutions. Please contact your sales representative for sales and pricing information. 2. The only framework in NeXTSTEP 3.3 which performed explicit date-handling was DBKit, the predecessor to Apple's Enterprise Objects Framework. DBKit did not handle dates in a Year 2000-compliant fashion, and therefore any application which makes use of DBKit is likely to experience Year 2000 problems. 3. The third-party tools and libraries that shipped with NeXTSTEP 3.3 have not been audited for Year 2000 compliance. Some third-party tools and libraries may not be Year 2000 compliant in the versions bundled with NeXTSTEP 3.3. 3. NeXTSTEP 3.3 did not have an equivalent to the OPENSTEP 'NSDate' object; most applications and frameworks used UNIX utilities and functions to manage dates. These UNIX constructs and any custom date handling an application performs may experience Year 2000 problems. For this reason, the applications that shipped with NeXTSTEP 3.3 may not handle the Year 2000 transition properly, even if they are run in an OPENSTEP environment. Your custom NeXTSTEP applications will run in an OPENSTEP environment only if they do no date handling, or if they were specifically programmed with custom Year 2000-compliant date handling routines.
Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple says NEXTSTEP 3.3 **NOT** Year 2K compliant References: <35F00B0B.276F5E89@NOSPAMREMOVETHISisomedia.com> <6spskt$nrs$1@news.seicom.net> From: sdroll@NOSPMmathematik.uni-wuerzburg.de (Sven Droll) Message-ID: <35f39ac7.0@uni-wuerzburg.de> Date: 7 Sep 98 08:35:19 GMT >UNIX in general does not have a year 2000 problem, it has a year 2038 >problem, but this is a different matter. Gee. We will be old then :-) (as our NeXTs will) >I put it quad-fat on http://www.this.net/~frank/date_y2k.tar.gz for those >without compiler who wants to try it themselfs (as usual you are working at >your own risk, so backup your data before installing it). Run 'date --help' >to see the new options. > >- The time module of Preference.app is unable to set dates > year 1999, guess I use date 0.2 from peanuts for half an year on about 15 machines (black, white, yellow) and it works just fine. I tested dates after year 2000 and it continued to work... From ftp://ftp.peanuts.org/peanuts/./NEXTSTEP/unix/admin/date.0.2.README ---snip--- This version of date is Y2K compliant and works for NeXTStep 3.3 (and hopefully earlier!) To replace the old date program on your NeXTStep 3.3 machine, do the following: mv -f /usr/lib/Preferences/date /usr/lib/Preferences/date.ORIG cp date.02.NIHS.bs/bin/date /usr/lib/Preferences/date (Note: /bin/date and /usr/lib/Preferences/date are identical, as shipped from NeXT. You may also wish to replace /bin/date with this version of date.) /usr/lib/Preferences/date is what Preferences.app uses to set the date. This version is capable of setting the date to a time after 2000. THIS PROGRAM HAS NOT UNDERGONE EXTENSIVE TESTING AND HAS NO GUARANTEES. ---snip--- -- Sven Droll __ ______________________________________________________/ / ______ __ sdroll@mathematik.uni-wuerzburg.de / /_/ ___/ please remove the NOSPM from my reply-address /_ _/ _/ =====\_/======= LOGOUT FASCISM! ___________________________________________________________________ NeXT-mail or MIME welcome ;-)
From: matthias@:-)factum-gmbh.de (Matthias Schürhoff) Newsgroups: comp.sys.next.programmer Subject: Re: Problem with app icon (.ico) under Openstep NT Date: 7 Sep 1998 09:48:44 GMT Organization: FACTUM Projektentwicklung und Management GmbH Message-ID: <6t0a5s$16t@factum.factum-gmbh.de> References: <6sma00$bet@factum.factum-gmbh.de> <35EF3A70.DFD0E8E6@object-works.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Chris Johnson <cjohnson@object-works.com> wrote: > <HTML> > Matthias, > > <P>I think if you look at the .ico in explorer you will see that the details > of the file are messed up. > <BR>Something about ProjectBuilder and .ico's messes them up occasionally. > Just copy the .ico > <BR>back into the project directory (replacing the old one) should fix > it (you should see the details > <BR>looking nice again) Saving the .ico file again into the project directory did it:-) Thanks! Matthias =========================================================== Matthias Schürhoff FACTUM Projektentwicklung und Management GmbH, Dortmund E-Mail: matthias@factum-gmbh.de (NeXT & MIME mail ok) WARNING: The return email address field has been altered:-) ----------------------------------------------------------- "Don't marry, be happy" - Unknown
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: Setting a tooltip programatically Date: 7 Sep 1998 19:00:38 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6v8bar.5ki.rog@talisker.ohm.york.ac.uk> References: <35EB9ECB.52A0E3E@Square.nl> <6sgejq$9gt$1@newshost> On 1 Sep 1998 09:26:18 GMT, Heinrich Giesen <giesen@informatik.uni-koblenz.de> wrote: > In the header file NSView.h you find > - (void)setToolTip:(NSString *)string; > - (NSString *)toolTip; > for setting and getting tooltip strings. does anyone know if it's possible to get slightly more specialised behaviour from this tooltip facility. i'd like to be able to give each buttoncell in a matrix of buttons a different tooltip (doesn't seem like an unreasonable request...) cheers, rog.
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <645346C45E37D211ACDC006097D1798C81DD80@news.telebahia.net.br> Control: cancel <645346C45E37D211ACDC006097D1798C81DD80@news.telebahia.net.br> Date: 06 Sep 1998 22:14:24 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.645346C45E37D211ACDC006097D1798C81DD80@news.telebahia.net.br> Sender: herpla5@aol.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Christian Neuss <neuss.@informatik.th-darmstadt.de.nos-pam> Newsgroups: comp.sys.next.programmer Subject: Re: quel programme choisir??? Date: 5 Sep 1998 15:01:25 GMT Organization: Technische Universitaet Darmstadt Message-ID: <6srjo5$46v$1@sun27.hrz.tu-darmstadt.de> References: <6srbgs$ee1$1@platane.wanadoo.fr> "HUILLE Erika" <chouchounova@wanadoo.fr> wrote: >Salut! >Je veux me lancer dans la programmation, mais je ne sais pas exactement quel >langage choisir??? YMMV, but I suggest chosing the English language, at least while posting to the c.s.n. hierarchy, By the way, isn't it illegal these days to say 'programmation' in French? Just curious.. (the answer probably is 'Java' for a newbie..) Chris -- // Christian Neuss "static typing? how quaint.." // http://www.nexttoyou.de/~neuss/ // fax: (+49) 6151 16 5472
From: sun@unity.ncsu.edu (Ying-Hsuan Sun) Newsgroups: comp.sys.next.programmer Subject: postgresql 3.3 port? Date: 8 Sep 1998 03:10:04 GMT Organization: North Carolina State University Message-ID: <6t276c$sa8$1@uni00nw.unity.ncsu.edu> Does anyone know if the binaries for Postgresql exits for 3.3 ? Ying-Hsuan Sun
From: Chris Johnson <cjohnson@object-works.com> Newsgroups: comp.sys.next.programmer Subject: Re: Setting a tooltip programatically Date: Tue, 08 Sep 1998 00:15:04 -0400 Organization: ObjectWorks Inc. Message-ID: <35F4AF47.55C81846@object-works.com> References: <35EB9ECB.52A0E3E@Square.nl> <6sgejq$9gt$1@newshost> <slrn6v8bar.5ki.rog@talisker.ohm.york.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: rog@ohm.york.ac.uk Roger, Try NSMatrix instance method - (void)setToolTip:(NSString *)theTip forCell:(NSCell *)theCell It's not documented but works!~ Chris Roger Peppe wrote: > On 1 Sep 1998 09:26:18 GMT, Heinrich Giesen <giesen@informatik.uni-koblenz.de> wrote: > > In the header file NSView.h you find > > - (void)setToolTip:(NSString *)string; > > - (NSString *)toolTip; > > for setting and getting tooltip strings. > > does anyone know if it's possible to get slightly more specialised > behaviour from this tooltip facility. > > i'd like to be able to give each buttoncell in a matrix of buttons a > different tooltip (doesn't seem like an unreasonable request...) > > cheers, > rog.
From: msb@plexare.com (Michael S. Barthelemy) Newsgroups: comp.sys.next.programmer,comp.sys.next.marketplace Subject: JOB: Senior Software Engineer - OpenStep/WebObjects Date: 8 Sep 1998 16:14:13 GMT Organization: A-Link Network Services, Inc. Message-ID: <6t3l4l$dlb@ns2.alink.net> Department: Engineering Job Title: Senior Software Engineer - OpenStep/WebObjects Location: San Francisco Posted Date: September 1, 1998 Start Date: Immediately Position Description: Be a key contributor and technical leader for a growing, fast-paced internet software enterprise. Responsible for full life cycle development of object-oriented, application software for Inquisit, a web and email-based personal intelligence service. Work closely with product marketing, QA and other engineering team members to define requirements for, design, program, build, test and deploy software for commercial use. Conceptualize, design and build enterprise software components that provide a strong foundation for developing scalable and reliable applications. Contribute insight and leadership to improving software engineering and product development practices. Collaborate with Head of Engineering and other senior engineers on overall systems architecture and technology strategy, and provide leadership by example to more junior programmers. Required Skills & Experience: * 5+ years experience in commercial software development, with minimum 2 years in technical team lead role, with emphasis on client/server computing and/or web applications development under Unix (Solaris) and/or Windows NT * Solid object-oriented design skills and 2+ years experience with NextStep/OpenStep, EOF, WebObjects, and Objective C programming * Experience developing, configuring, and maintaining high traffic internet/web applications using webserver, cgi, HTML, XML and/or related technologies * Solid track record of building commercial applications on time and within budget * Strong leadership, teamwork, interpersonal, and written and oral communication skills Other Desirable Skills & Experience: * Good planning, organization, and project management skills desired * Familiarity with quality assurance, source code control, software configuration management, release management methods and tools * Familiarity with formal OO design methodology and some programming experience in C++, Java, or Smalltalk * Data modeling and relational database design experience highly desirable; experience with Oracle and SQL a plus * Practical experience with natural language processing, collaborative filtering systems and techniques highly desirable * Familiarity with email systems, search engines, online information services, and/or e-commerce applications a plus Inquisit, Inc. 75 Federal Street San Francisco, CA 94107 http://www.inquisit.com Contact: Joe Hawkins (415)659-9029 hawkins@inquisit.com
From: "Dietmar Planitzer" <dave.pl@ping.at> Newsgroups: comp.sys.next.programmer Subject: ANN: GLUT for MacOS X Server now available Date: Tue, 08 Sep 1998 22:52:26 +0200 Organization: EUnet Austria Message-ID: <dave.pl-0809982252260001@dialup176.salzburg20.salzburg.at.eu.net> NNTP-Posting-Date: 8 Sep 1998 20:47:20 GMT Hello All, I just wanted to let you know that GLUT for MacOS X Server / Rhapsody Developer Release 2 is now available from the following site: ftp://ftp.next.peak.org/pub/rhapsody/Developer/Frameworks What's that GLUT thing ?? GLUT stands for OpenGL Utility Toolkit. It was originally developed by Mark J. Kilgard in order to make it possible to write OpenGL programs in a platform independend way. GLUT provides a very rich windowing and input device management API for developing powerful OpenGL programs using only GL and GLUT calls. If you want to learn more about GLUT, have a look at: http://reality.sgi.com/opengl/glut3/glut3.html The distribution consists of the files: 1. MacOSXGLUT-1.0-P-b.tar.gz Precompiled binaries for PowerMacintosh computers. 2. MacOSXGLUT-1.0-PI-s.tar.gz Sources for GLUT, GLE, MUI, a few example programs and the test programs. 3. MacOSXGLUT_Demos-1.0-PI-s.tar.gz Sources of all example programs that come with the standard GLUT distribution. 4. MacOSXGLUT_Data-1.0-PI-b.tar.gz Finally, the texture files for the example programs The file MacOSXGLUT-1.0.README gives more detailed information about it's features and still existing problems which should be solved in a future version. The MacOS X Server port has a number of special features that are unique to this GLUT implementation: Copy to Clipboard. You can copy the contents of the current window, the one which is key, to the pasteboard every time, even while an animation is running. The data is written in TIFF format to the general pasteboard. You can even take a look at the data currently stored on the pasteboard server if you select "Show Clipboard" from the Edit menu. All standard data types are supported: various text formats, EPS, TIFF and sound. Services The contents of the current window can be sent via services to other applications at any time. All applications which are able to accept EPS, TIFF or RTFD data are supported. This mechanismn is very handy - i.e. if you want to send your newest and greatest graphics to your colleagues, simply select "MailViewer->MailTo" from the Services menu, which will bring up the MailViewer application displaying a new letter containing the graphics. Saving Window Contents It is also possible to save the current windows contents to a file on your local disk (URL and thus remote support, will be added in version 1.1). The data will be safed in uncompressed TIFF format. Printing Printing as EPS is also supported. Faxing Although, the necessary code to enable direct faxing from within any GLUT application is there, faxing is broken in RDR2. The problem is the fax panel, which has a bug in the menu creation code. If you have MacOS X Server Beta - give it a try, maybe it works ! MacOSXGLUT supports Mesa version 2.6 and Conix's OpenGL for RDR2. You can choose between the two gl implementatios by simply defining the necessary preprocessor symbol when compiling the framework. The readme file contains more information about this. Another thing mostly interesting to people speaking German, is that a localization for German is included. If someone has selected German as his/her prefered language, all GLUT menus and messages will be displayed in German, otherwise in English. Speaking of localization - if you think that other languages, like French or Japan (would be cool !) should be supported, feel free to send me the localized resources - NIBs and strings file - so that I will be able to include them in version 1.1 ! Have fun with it, Planitzer Dietmar
From: "Charles W. Swiger" <chuck@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: I need my .h files!!! Date: 9 Sep 1998 22:59:36 GMT Organization: Spacelab.net Internet Access Message-ID: <6t718o$c67$1@news.mxol.com> References: <6t6rjk$lj1$1@news.safari.net> dan@news.safari.net (Dan Nafe) wrote: >Is there someplace on the web I can download the .h files for my black >Openstep machine? I need the basick such as stdio.h, stdlib.h ... > >Please mail me the location the these files and libraies RTFAQ. The header files and system libraries weren't public; they belonged to NeXT. You'll have to purchase the Developer edition if you want to be able to compile anything.... -Chuck Charles Swiger | chuck@codefab.com | Yeah, yeah-- disclaim away. ----------------+-------------------+---------------------------- You have come to the end of your journey. Survival is everything.
From: szallies@energotec.de (Constantin Szallies) Newsgroups: comp.sys.next.programmer Subject: OS 4.2: NSPerformService() kills my pasteboard server Date: 9 Sep 1998 14:43:27 GMT Organization: TechNet GmbH Message-ID: <6t646f$85f$1@oxygen.technet.net> NNTP-Posting-Date: 9 Sep 1998 14:43:27 GMT If I invoke NSPerformService() for the first time, everything works OK. But the second time, I get a NSPastboardCommunicationException with description "Bad return value". Here's the code I am using: @implementation Controller - _performMailService:sender; { NSPasteboard *aPasteboard; aPasteboard=[NSPasteboard generalPasteboard]; [@"HELLO SERVICE" writeToFile:@"/tmp/TEST" atomically:YES]; [aPasteboard declareTypes: [NSArray arrayWithObject:NSFilenamesPboardType] owner:nil]; [aPasteboard setPropertyList: [NSArray arrayWithObject:@"/tmp/TEST"] forType:NSFilenamesPboardType]; NSPerformService(@"Mail/Document", aPasteboard); return self; } - performMailService:sender; { [self performSelector: @selector(_performMailService:) withObject:self afterDelay: 0.0]; return self; } @end I use performSelector: because someone at the Omnigroup Mailing List recommended it as a work-a-round. But it doesn't help here in this situation. Any thoughts? Greetings -- # Constantin Szallies, Energotec GmbH # szallies@energotec.de # http://www.energotec.de/~szallies/ # 49211-9144018
Newsgroups: comp.sys.next.programmer Subject: Re: GNU gcc References: <35F6EE23.2143@denalisites.com> From: sdroll@NOSPMmathematik.uni-wuerzburg.de (Sven Droll) Message-ID: <35f78ede.0@uni-wuerzburg.de> Date: 10 Sep 98 08:33:34 GMT Wotring Brian J <brian@denalisites.com> wrote: >I have grabbed various builds of gcc over the last year and all of them >have problem with iostreams. Code will compile, however, I get linker >errors. Also, I'm getting a lot of segmentation faults on C++ code that >has been tested on other platforms. Has anybody had similar problems? > >I would like to get a good C++ compiler working and I'm too lazy to >build it myself. I had never problems with the ports of gcc 2.7.1/2.7.2 from Rex Dieter (Thank you!) on black and white hardware (NS3.3). For not too lazy people ;-): If you are on an intelbox and have gcc 2.7.1 or 2.7.2 you should consider getting the newest egc from some GNU-mirror. It compiled almost out of the box (there was one trick of removing the comments in the Assembler-codemechanism: this was one -D... to set in the CFLAGS AFAIR) as stage 3 compiler and should be far better (according to some other people I know) on c++ than gcc (but I have not tested c++ progs). For my numerical simulations it was no improvement, though. greets Sven -- Sven Droll __ ______________________________________________________/ / ______ __ sdroll@mathematik.uni-wuerzburg.de / /_/ ___/ please remove the NOSPM from my reply-address /_ _/ _/ =====\_/======= LOGOUT FASCISM! ___________________________________________________________________ NeXT-mail or MIME welcome ;-)
From: Wotring Brian J <brian@denalisites.com> Newsgroups: comp.sys.next.programmer Subject: GNU gcc Date: Wed, 09 Sep 1998 16:07:47 -0500 Organization: University of Southwestern Louisiana Message-ID: <35F6EE23.2143@denalisites.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I have grabbed various builds of gcc over the last year and all of them have problem with iostreams. Code will compile, however, I get linker errors. Also, I'm getting a lot of segmentation faults on C++ code that has been tested on other platforms. Has anybody had similar problems? I would like to get a good C++ compiler working and I'm too lazy to build it myself. -- | Brian Wotring | brian@denalisites.com ( NeXT-Mail welcome ) | Denali Sites, Inc.
From: dan@news.safari.net (Dan Nafe) Newsgroups: comp.sys.next.programmer Subject: I need my .h files!!! Date: 9 Sep 1998 21:23:00 GMT Organization: Safari Internet Message-ID: <6t6rjk$lj1$1@news.safari.net> NNTP-Posting-Date: 9 Sep 1998 21:23:00 GMT Is there someplace on the web I can download the .h files for my black Openstep machine? I need the basick such as stdio.h, stdlib.h ... Please mail me the location the these files and libraies ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Dan Nafe, Safari Internet dan@NOSPAM.safari.net http://www.safari.net
From: Henry Koplien <koplien@de.ibm.com> Newsgroups: comp.sys.next.programmer Subject: Re: GNU gcc Date: Thu, 10 Sep 1998 08:14:22 +0200 Organization: IBM HD MicroCode Message-ID: <35F76E3E.41C6@de.ibm.com> References: <35F6EE23.2143@denalisites.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Wotring Brian J wrote: > > I have grabbed various builds of gcc over the last year and all of them > have problem with iostreams. Code will compile, however, I get linker > errors. Also, I'm getting a lot of segmentation faults on C++ code that > has been tested on other platforms. Has anybody had similar problems? > > I would like to get a good C++ compiler working and I'm too lazy to > build it myself. > > -- > > | Brian Wotring > | brian@denalisites.com ( NeXT-Mail welcome ) > | Denali Sites, Inc. Uhh? 2.7.2 works out of the box and compiles very well, 2.8.0 works also fine except the warnings while linking. (run ranlib....) Henry -- ------------------------------------------------------------------ snail mail : Henry Koplien, HD Micro Code Development, IBM 71032 Boeblingen/Germany \|/ voice : +49-7031-16-3516 o(O O)o fax : +49-7031-16-3328 \^/ -------------- Ihh-Mehl: Koplien@de.IBM.com --ooOo---(_)---oOoo---
From: isor@ripco.com (chiroptera) Newsgroups: comp.sys.next.programmer Subject: GCC 2.8.1 Date: Fri, 11 Sep 1998 05:04:48 GMT Organization: WorldWide Access - Midwestern Internet Services - www.wwa.com Message-ID: <35f8af43.106636585@news.wwa.com> I get the error: checking for sys_siglist declaration in signal.h or unistd.h... no Is this just a path problem?
From: danNOSPAM@safari.net (Dan Nafe) Newsgroups: comp.sys.next.programmer Subject: Re: I need my .h files!!! Date: 11 Sep 1998 20:04:35 GMT Organization: Safari Internet Message-ID: <danNOSPAM-1109981601200001@bwana.safari.net> References: <6t6rjk$lj1$1@news.safari.net> <6tbrdb$kr2$1@agate.berkeley.edu> NNTP-Posting-Date: 11 Sep 1998 20:04:35 GMT In article <6tbrdb$kr2$1@agate.berkeley.edu>, "satoru@candenext.lsa.berkeley.edu.berkeley.edu (Satoru Uzawa) wrote: [snip] > stdio.h > > /NextDeveloper/Headers/ansi/stdio.h > /usr/include/ansi/stdio.h > > stdlib.h > > /NextDeveloper/Headers/ansi/stdlib.h > /usr/include/ansi/stdlib.h > > For a reason I don't know, some of the header files are located in > non-standard places (according to other UNIXes). [snip] My machine doesn't have anything in /usr/include and only has the following in /NextDeveloper/ Apps/ Demos/ Examples/ Makefiles/ Source/ Arrrgh! -- Sir, I do not recall such an operation...and if this operation did, in fact occur, I would not be at liberty to discuss it!
From: rdieter@math.unl.edu (Rex Dieter) Newsgroups: comp.sys.next.programmer Subject: Re: I need my .h files!!! Date: 11 Sep 1998 21:25:46 GMT Organization: University of Nebraska-Lincoln Message-ID: <6tc4gq$cha$1@unlnews.unl.edu> References: <6tbrdb$kr2$1@agate.berkeley.edu> In article <6tbrdb$kr2$1@agate.berkeley.edu> "satoru@candenext.lsa.berkeley.edu.berkeley.edu (Satoru Uzawa) writes: > dan@news.safari.net (Dan Nafe) wrote: > >Is there someplace on the web I can download the .h files for my black > >Openstep machine? I need the basick such as stdio.h, stdlib.h ... You need the OpenStep DeveloperCD. It includes all the headerfiles and libraries. > stdlib.h > > /NextDeveloper/Headers/ansi/stdlib.h > /usr/include/ansi/stdlib.h > > For a reason I don't know, some of the header files are located in > non-standard places (according to other UNIXes). I always use "Finder" > from "Workspace" to find the location of functions when a simple > "configure" and "make" failed to find them. This is not quite accurate.. The header files REALLY live in /NextDeveloper/Headers, but there exists a symbolic link at /usr/include that points to it. /usr/include, btw, is the "standard UNIX place" (whatever that means?!@# (-; ) for header files. -- Rex A. Dieter rdieter@math.unl.edu (NeXT/MIME OK) Computer System Manager http://www.math.unl.edu/~rdieter/ Mathematics and Statistics University of Nebraska-Lincoln
From: dfevans@bbcr.uwaterloo.ca (David Evans) Newsgroups: comp.sys.next.programmer Subject: Re: I need my .h files!!! Date: 11 Sep 1998 20:50:13 GMT Organization: University of Waterloo Message-ID: <905547013.329849@watserv4.uwaterloo.ca> References: <6t6rjk$lj1$1@news.safari.net> <6tbrdb$kr2$1@agate.berkeley.edu> <danNOSPAM-1109981601200001@bwana.safari.net> Cache-Post-Path: watserv4.uwaterloo.ca!unknown@bcr11.uwaterloo.ca In article <danNOSPAM-1109981601200001@bwana.safari.net>, Dan Nafe <danNOSPAM@safari.net> wrote: >My machine doesn't have anything in /usr/include What is the contents of your /NextLibrary/Receipts? -- David Evans (NeXTMail/MIME OK) dfevans@bbcr.uwaterloo.ca Computer/Synth Junkie http://bbcr.uwaterloo.ca/~dfevans/ University of Waterloo "Default is the value selected by the composer Ontario, Canada overridden by your command." - Roland TR-707 Manual
From: "satoru@candenext.lsa.berkeley.edu.berkeley.edu (Satoru Uzawa) Newsgroups: comp.sys.next.programmer Subject: Re: I need my .h files!!! Date: 11 Sep 1998 18:50:19 GMT Organization: University of California, Berkeley Message-ID: <6tbrdb$kr2$1@agate.berkeley.edu> References: <6t6rjk$lj1$1@news.safari.net> dan@news.safari.net (Dan Nafe) wrote: >Is there someplace on the web I can download the .h files for my black >Openstep machine? I need the basick such as stdio.h, stdlib.h ... > stdio.h /NextDeveloper/Headers/ansi/stdio.h /usr/include/ansi/stdio.h stdlib.h /NextDeveloper/Headers/ansi/stdlib.h /usr/include/ansi/stdlib.h For a reason I don't know, some of the header files are located in non-standard places (according to other UNIXes). I always use "Finder" from "Workspace" to find the location of functions when a simple "configure" and "make" failed to find them. -- Satoru Uzawa, satoru@candenext.lsa.berkeley.edu (NeXTmail welcome)
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: I need my .h files!!! Date: 11 Sep 1998 22:45:37 GMT Organization: Idiom Communications Message-ID: <6tc96h$7ij$4@news.idiom.com> References: <6t6rjk$lj1$1@news.safari.net> <6tbrdb$kr2$1@agate.berkeley.edu> <danNOSPAM-1109981601200001@bwana.safari.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: danNOSPAM@safari.net Dan Nafe may or may not have said: [snip] -> My machine doesn't have anything in /usr/include and only has the -> following in /NextDeveloper/ -> Apps/ Demos/ Examples/ Makefiles/ Source/ -> Arrrgh! Then you haven't sucessfully installed the developer tools. Get out your developer CD, and make sure you install DeveloperTools.pkg, DeveloperLibs.pkg, and (probably) DeveloperDoc.pkg. -jcr -- "Although UNIX is more reliable, NT may become more reliable with time" - Ron Redman, deputy technical director of the Fleet Introduction Division of the Aegis Program Executive Office, US Navy. Where is the line between mere incompetence and actual treason? -jcr
From: danNOSPAM@safari.net (Dan Nafe) Newsgroups: comp.sys.next.programmer Subject: Re: I need my .h files!!! Date: 11 Sep 1998 20:05:32 GMT Organization: Safari Internet Message-ID: <danNOSPAM-1109981602170001@bwana.safari.net> References: <6t6rjk$lj1$1@news.safari.net> <6tbrdb$kr2$1@agate.berkeley.edu> NNTP-Posting-Date: 11 Sep 1998 20:05:32 GMT In article <6tbrdb$kr2$1@agate.berkeley.edu>, "satoru@candenext.lsa.berkeley.edu.berkeley.edu (Satoru Uzawa) wrote: [snip] > stdio.h > > /NextDeveloper/Headers/ansi/stdio.h > /usr/include/ansi/stdio.h > > stdlib.h > > /NextDeveloper/Headers/ansi/stdlib.h > /usr/include/ansi/stdlib.h > > For a reason I don't know, some of the header files are located in > non-standard places (according to other UNIXes). [snip] My machine doesn't have anything in /usr/include and only has the following in /NextDeveloper/ Apps/ Demos/ Examples/ Makefiles/ Source/ Arrrgh! -- Sir, I do not recall such an operation...and if this operation did, in fact occur, I would not be at liberty to discuss it!
From: cnyap@dcs.shef.ac.uk (Chih Nam Yap) Newsgroups: comp.sys.next.programmer Subject: Fonts for Mathematical symbols Date: 11 Sep 1998 09:37:50 GMT Organization: Department of Computer Science, University of Sheffield Message-ID: <6tar1e$18s$1@bignews.shef.ac.uk> Hi there, Does anyone know of any ftp site in which I can down load fonts that have mathematical symbols in it (eg "for all", set membership symbol) I am currently building a tool for writing mathematic statements. Thank you
From: "Alex Molochnikov" <alex@gestalt.com> Newsgroups: comp.sys.next.programmer Subject: Re: How to compile for YellowBox for Win under RhapDR2? Date: Sun, 13 Sep 1998 15:23:46 -0600 Organization: Canada Connect Corp. Message-ID: <6thd3q$cb$1@cleavage.canuck.com> References: <OEMK1.2660$_4.2962095@nntpserver.swip.net> > Nils Ågren wrote in message ... > Hello, > I'm trying to make my application to run under YellowBox for Windows.The > problem is I´m using Rhapsody DR2 at my work and now I´m trying to make it > run under YB for Win. > I setup the ProjectBuilder to build a YB for Win App, but the ProjectBuilder > just gives me errors back, why? Nils Ågren You need to install Yellow Box for Windows on Windows NT, copy your project to the Windows host and rebuild it from scratch. Building your project under Rhapsody will only create a single or fat-binary executable on Rhapsody. Do not be misled by the option to build it for Windows provided by the Project Builder -- it is just a cosmetic toy with not much use. Alex Molochnikov alex@gestalt.com
From: luomat@peak.org.this.all.must.be.removed (TjL) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple says NEXTSTEP 3.3 **NOT** Year 2K compliant Date: 14 Sep 1998 00:07:01 GMT Organization: Florida Digital Turnpike Message-ID: <6thmn5$cqg@obi-wan.fdt.net> References: <35F00B0B.276F5E89@NOSPAMREMOVETHISisomedia.com> <6spskt$nrs$1@news.seicom.net> <distler-1309981419370001@192.168.0.1> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: distler@golem.ph.utexas.edu In <distler-1309981419370001@192.168.0.1> Jacques Distler wrote: > Note to administrators of Peak & Peanuts: Perhaps a separate directory of > Y2K fixes would be appropriate. That will make it a lot easier for folks > to move their system towards compliance. As of today there is only one such fix that I know of. If more come about, then putting them into such a folder might make sense. Right now there really isn't enough there to justify it. TjL -- Spam-altered address in effect, remove obvious portion if replying by email.
From: distler@golem.ph.utexas.edu (Jacques Distler) Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer Subject: Re: Apple says NEXTSTEP 3.3 **NOT** Year 2K compliant Date: Sun, 13 Sep 1998 14:19:37 -0500 Organization: Physics Department, University of Texas at Austin Message-ID: <distler-1309981419370001@192.168.0.1> References: <35F00B0B.276F5E89@NOSPAMREMOVETHISisomedia.com> <6spskt$nrs$1@news.seicom.net> /gnVR"|LWyKVK{`$&t&`k{]Q^x*"ihe+^hTvWs@IEMtYI5RVzGFBPcMu`}>@#^Fm:<)M w83/\[@#\u?TVrF3fqDjK? In article <6spskt$nrs$1@news.seicom.net>, frank@this.NO_SPAM.net (Frank M. Siegert) wrote [on Y2K]: > - I installed the GNU version (sh-utils 1.16) of '/bin/date' to be able to > set the date as the original date program refuse to set dates > year 1999. > > I put it quad-fat on http://www.this.net/~frank/date_y2k.tar.gz for those > without compiler who wants to try it themselfs (as usual you are working at > your own risk, so backup your data before installing it). Run 'date --help' > to see the new options. > > - The time module of Preference.app is unable to set dates > year 1999, guess > someone need to write a replacement. It is however able to display them, so > you need to set date and time in a shell (no big deal IMHO). > > So what to do? Maybe I will write a new Preference module next year to be > able to set dates using Preference.app, most of my own apps will continue to > work (I do not use DBKit, nor does much other software I know of...). The > libraries and some utilities in question (at, atq, atrun, atrm, troff, maybe > more...) can be replaced by newer versions (ported from FreeBSD/Linux), the > same applies for mail, etc (and this is one of the strengths of a modular > system, try this with a monolithic OS -- uh oh... pain, pain). If you do any further ports of Y2K-compliant utilities to NS3.3, please post a notice to comp.sys.next.sysadmin . There already is a Y2K-compliant version of /bin/date on Peanuts and Peak (date.02.NIHS.bs). But clearly, there is a bit more to be done. Note to administrators of Peak & Peanuts: Perhaps a separate directory of Y2K fixes would be appropriate. That will make it a lot easier for folks to move their system towards compliance. JD -- PGP public key: http://golem.ph.utexas.edu/~distler/distler.asc
From: holger@_REMOVE_THIS_.wizards.de (Holger Hoffstaette) Newsgroups: comp.sys.next.programmer Subject: Re: How to compile for YellowBox for Win under RhapDR2? Date: 13 Sep 1998 12:54:22 GMT Organization: The secret circle of the NSRC Message-ID: <6tgf9u$dn@ragnarok.en.uunet.de> References: <OEMK1.2660$_4.2962095@nntpserver.swip.net> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit "Nils Ågren" wrote: > I'm trying to make my application to run under YellowBox for Windows.The > problem is I´m using Rhapsody DR2 at my work and now I´m trying to make it > run under YB for Win. > I setup the ProjectBuilder to build a YB for Win App, but the ProjectBuilder > just gives me errors back, why? Because what you want to do is not possible. YB/Windwos uses the native Windows object file format and links against native Windows libraries. These are - surprise! - not available on Rhapsody. Also, YB/Windows uses the native (and soo braindead) M$ linker, which also only works on Windows. In order to create a YB/Windows version of your app, you must recompile on the Windows box. All this should be in the docs and release notes.. Holger
From: x557@mindspring.com (x557@mindspring.com) Newsgroups: comp.sys.next.hardware,comp.sys.next.misc,comp.sys.next.programmer Subject: Question about problem with IDE/"won't mount" HELP! Date: Mon, 14 Sep 1998 06:35:18 GMT Organization: x557@mindspring.com Message-ID: <35fcb898.1010137@news.mindspring.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Would someone tell me if they have encountered this problem before? I am trying to install OpenStep 4.2 "mach" on my Intel machine. I have an "ide" (or is it "eide") hard drive. I have a regular ide 'ata' or 'atapi' cd-drom. You know how you put the disks in the drive, and it boots off the bootup disks, and then you choose what kind of drivers you want? Well, I choose what kind of drivers, I choose many different kinds. I spent about 5 hours trying about 30-40 different combinations. including ide, primar/secondary (dual) edide and atapi etc.etc. But I never could get it to 'mount'. I would always end up with something like, "No Scsi Controller or CD-ROM drive found" and some little numbers like, "SD%D, HD%D FD%D, TR%D". Although, one time I was able to get it to say 'waiting for cd-drive to respond', but still, in the end, it wouldn't work. I'm using a pretty standard setup, with matrox, 64megs, etc. my question is, has anyone else run into this problem before? I think there is some little thing I'm leaving out, and thought I would ask! If anyone has any ideas, would you please post them? I will read them tomorrow when I get back online. p.s. I hope I got the right newsgroup(s).
From: trail@ix.netcom.com Newsgroups: comp.sys.next.hardware,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Question about problem with IDE/"won't mount" HELP! Date: 14 Sep 1998 12:32:49 GMT Organization: ICGNetcom Message-ID: <6tj2dh$qr6@dfw-ixnews8.ix.netcom.com> References: <35fcb898.1010137@news.mindspring.com> In-Reply-To: <35fcb898.1010137@news.mindspring.com> On 09/14/98, x557@mindspring.com wrote: >Would someone tell me if they have encountered >this problem before? I am trying to install >OpenStep 4.2 "mach" on my Intel machine. I have >an "ide" (or is it "eide") hard drive. I have a >regular ide 'ata' or 'atapi' cd-drom. You know >how you put the disks in the drive, and it boots >off the bootup disks, and then you choose what kind >of drivers you want? Well, I choose what kind of >drivers, I choose many different kinds. I spent >about 5 hours trying about 30-40 different combinations. >including ide, primar/secondary (dual) edide and atapi etc.etc. >But I never could get it to 'mount'. I would always end >up with something like, "No Scsi Controller or CD-ROM >drive found" and some little numbers like, "SD%D, HD%D >FD%D, TR%D". Although, one time I was able to get it to >say 'waiting for cd-drive to respond', but still, in the >end, it wouldn't work. >I'm using a pretty standard setup, with matrox, 64megs, >etc. my question is, has anyone else run into this problem >before? I think there is some little thing I'm leaving >out, and thought I would ask! >If anyone has any ideas, would you please post them? >I will read them tomorrow when I get back online. > The usually recommended way to install from atapi cd-rom is to have it jumpered as slave to your hard disk on the first ide channel on your mb. When installing drivers, use the beta driver for the Adaptec 1542B SCSI controller as the controller for your cd, and the beta driver for EIDE for you hard drive. NS "should" then register your cd-rom and allow you to install. (Note I say "should"; although I have seen other posts to this effect, I have always used SCSI cd-roms myself.) Hope this helps, Jeff
From: Dan Wellman <wellman@students.uiuc.edu> Newsgroups: comp.sys.next.programmer Subject: Profilers? Date: 14 Sep 1998 15:44:58 GMT Organization: University of Illinois at Urbana-Champaign Message-ID: <6tjdlq$hq0$1@vixen.cso.uiuc.edu> Are there any commercial/shareware Objective C profilers for OPENSTEP (on Windows NT?) Thanks! dan -- Dan Wellman <> wellman@uiuc.edu <> http://www.cen.uiuc.edu/~wellman/ "A million thoughts in one night can't be wrong" - Cause & Effect
From: "Charles W. Swiger" <chuck@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: How to compile for YellowBox for Win under RhapDR2? Date: 14 Sep 1998 18:11:11 GMT Organization: Spacelab.net Internet Access Message-ID: <6tjm7v$4ds$1@news.spacelab.net> References: <OEMK1.2660$_4.2962095@nntpserver.swip.net> "Nils =?ISO-8859-1?B?xQ==?=gren" <macnisse@swipnet.se> wrote: >Hello, >I'm trying to make my application to run under YellowBox for Windows.The >problem is I=B4m using Rhapsody DR2 at my work and now I=B4m trying to make it >run under YB for Win. >I setup the ProjectBuilder to build a YB for Win App, but the ProjectBuilde= >r >just gives me errors back, why? Unlike the case with NEXTSTEP and OPENSTEP, some of the libraries and tools required to cross-compile for Yellow Box on NT do not belong to NeXT (Apple). It's pretty much the same reason why they didn't support cross-compilation under the PDO/EOF/WOF environments under the "native" OS, like Solaris on a Sun SPARC. -Chuck Charles Swiger | chuck@codefab.com | Yeah, yeah-- disclaim away. ----------------+-------------------+---------------------------- You have come to the end of your journey. Survival is everything.
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: odd goings on with @encode Date: 14 Sep 1998 18:27:11 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6vqo04.m7r.rog@talisker.ohm.york.ac.uk> i've found what appears to be a very strange problem happening with @encode, under openstep 4.2 on intel. i can't seem to reproduce the problem in a small bit of code, but for various reasons i don't think this is due to problems in my code, but is likely to be some sort of compiler bug. the symptom is that @encode sometimes seems to be returning differently valued strings for the same C type. the type looks like this: struct cleanupentry { int type; union { struct dictcleanup { NSString *entry; NSString *key; } dict; struct notifcleanup { id obj; char *entry; void *data; } notif; } u; }; sometimes the @encode value of this prints out as: {cleanupentry=i({dictcleanup=@@}{notifcleanup=@*^v})} (which i believe to be correct) but occasionally (but consistently) it will print out as: {cleanupentry=i(?)} (which is definitely not correct!) this is all very strange, because i would assume that as @encode gives a constant string, every instance of it would return a pointer to the same piece of static, read-only memory. (and being read only, it would not be possible for my program to write to it without generating a memory fault) for most of the time, this does indeed appear to be the case. however, i've got a piece of code that belies this. it looks something like (assuming the above structure declaration): const char *encoding = @encode(struct cleanupentry); // global variable - (void)someMethod { printf("encoding is '%s' (%p)\nlocal '%s' (%p)\n", encoding, (void*)encoding, @encode(struct cleanupentry), (void*)@encode(struct cleanupentry)); } this prints out: encoding is '{cleanupentry=i({dictcleanup=@@}{notifcleanup=@*^v})}' (0x4c371) local '{cleanupentry=i(?)}' (0x4c371) if i change the code to: const char *encoding = @encode(struct cleanupentry); static void printencoding(const char *enc) { printf("encoding: '%s'\n", enc); } - (void)someMethod { printencoding(encoding); printencoding(@encode(struct cleanupentry)); } then it prints out: encoding: '{cleanupentry=i({dictcleanup=@@}{notifcleanup=@*^v})}' (0x4c37d) encoding: '{cleanupentry=i(?)}' (0x4c476) ah ha! so something odd *is* happening when i'm invoking @encode locally, because we can see that it's got a different pointer value. however this odd thing *doesn't* happen when we cast it to (void*). now i think this is truly bizarre. can anyone shed any light on what might be happening here? i suspect a bug in gcc, but would love to find out that i might be doing something stupid! cheers, rog.
Date: 14 Sep 1998 16:16:39 EST Newsgroups: comp.sys.next.programmer Message-ID: <cancel.13094324485861632@ssmmaeky.fun> Control: cancel <13094324485861632@ssmmaeky.fun> From: clewis@ferret.ocunix.on.ca Sender: boosterat@ssmmaeky.fun Subject: cmsg cancel <13094324485861632@ssmmaeky.fun> EMP/ECP (aka SPAM) cancelled by clewis@ferret.ocunix.on.ca. See news.admin.net-abuse.announce, report 19980914.03 for further details
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: Re: odd goings on with @encode Date: 15 Sep 98 11:02:35 Organization: Is a sign of weakness Message-ID: <SCOTT.98Sep15110235@slave.doubleu.com> References: <slrn6vqo04.m7r.rog@talisker.ohm.york.ac.uk> <slrn6vsiks.qbu.rog@talisker.ohm.york.ac.uk> In-reply-to: rog@ohm.york.ac.uk's message of 15 Sep 1998 11:08:09 GMT In article <slrn6vsiks.qbu.rog@talisker.ohm.york.ac.uk>, rog@ohm.york.ac.uk (Roger Peppe) writes: const char *encoding = @encode(struct cleanupentry); I'd suspect this line. I guess the question is how @encode() actually _works_. If it's a static thing, you'd expect both to be the same - but if there's a runtime component, the different uses might be different. The above line may be setup before the runtime initialization code occurs - perhaps you could initialize the variable in the +initialize method and see if there's a difference? // these two lines *should* print the same thing, but do not. printencoding(@encode(struct cleanupentry)); printencoding(encoding); BTW, they _do_ print the same thing under NeXTSTEP3.3 with the foundation kit patch... -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: Re: odd goings on with @encode Date: 16 Sep 1998 10:49:26 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6vv5to.1j6.rog@talisker.ohm.york.ac.uk> References: <slrn6vqo04.m7r.rog@talisker.ohm.york.ac.uk> <SCOTT.98Sep15110235@slave.doubleu.com> On 15 Sep 98 11:02:35, Scott Hess <scott@nospam.doubleu.com> wrote: > In article <slrn6vsiks.qbu.rog@talisker.ohm.york.ac.uk>, > rog@ohm.york.ac.uk (Roger Peppe) writes: > const char *encoding = @encode(struct cleanupentry); > > I'd suspect this line. I guess the question is how @encode() actually > _works_. If it's a static thing, you'd expect both to be the same - > but if there's a runtime component, the different uses might be > different. The above line may be setup before the runtime > initialization code occurs - perhaps you could initialize the variable > in the +initialize method and see if there's a difference? there's no runtime component at all. if i look as the assembly code for this code, there are two constant string initialisations in it, one of which is right, the other of which is wrong. originally i didn't have the line above at all - i found the bug because NSValue was giving an arithmetic exception when trying to initialise itself, and a quick printout of encode value being passed showed that it was bogus. the line above is just to illustrate that the problem doesn't occur all the time... > BTW, they _do_ print the same thing under NeXTSTEP3.3 with the > foundation kit patch... well, not under 4.2. (but i haven't applied any patches - are there any?) i'm certain this is a compiler bug - i just hope it doesn't rear its ugly head too often! on a related note, the compiler was giving me what i believe to be an incorrect warning yesterday. but... maybe not. is the compiler justified in giving me a warning for the following code? i thought that the @"..." string constants *were* NSStrings... NSString *somefn(NSString *s) { return s == nil ? @"Null string"; } (when i compile the above, i get: file.m:5: warning: pointer type mismatch in conditional expression ) cheers, rog.
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: Re: odd goings on with @encode Date: 16 Sep 1998 12:05:09 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6vvabo.2js.rog@talisker.ohm.york.ac.uk> References: <slrn6vqo04.m7r.rog@talisker.ohm.york.ac.uk> <SCOTT.98Sep15110235@slave.doubleu.co <slrn6vv5to.1j6.rog@talisker.ohm.york.ac.uk> On 16 Sep 1998 10:49:26 GMT, Roger Peppe <rog@ohm.york.ac.uk> wrote: > NSString *somefn(NSString *s) > { > return s == nil ? @"Null string"; > } > > (when i compile the above, i get: > file.m:5: warning: pointer type mismatch in conditional expression > ) oops: i meant: NSString *somefn(NSString *s) { return s == nil ? @"Null string" : s; } rog.
From: ayctxbjw@somethingfunny.net Newsgroups: comp.sys.next.programmer Subject: WOW!!! WHAT A STORY ONLINE!!! Date: 16 Sep 1998 18:26:03 GMT Organization: Northwestern Indiana Telephone Co. Message-ID: <6tovrr$hcv$368@hyperion.nitco.com> I happen to have dropped by http://www.despotovic.net and I couldn't believe what I saw. A complete case online with over 48 pictures, 3 police reports and more!!! Regards. P.S.- The website address is http://www.despotovic.net
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: Re: odd goings on with @encode Date: 15 Sep 1998 11:08:09 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn6vsiks.qbu.rog@talisker.ohm.york.ac.uk> References: <slrn6vqo04.m7r.rog@talisker.ohm.york.ac.uk> On 14 Sep 1998 18:27:11 GMT, Roger Peppe <rog@ohm.york.ac.uk> wrote: > i've found what appears to be a very strange problem happening with > @encode, under openstep 4.2 on intel. > > i can't seem to reproduce the problem in a small bit of code, but for > various reasons i don't think this is due to problems in my code, but > is likely to be some sort of compiler bug. i said above that i couldn't reproduce the problem in a small piece of code, but it's been quite correctly pointed out to me that this is pretty useless for bug-seeking people. so here's a piece of genuine code that shows the problem. it's not a major problem, as there's an easy workaround, but it does appear to be a genuine compiler bug, which occurs on black hardware also. cheers, rog. #import <AppKit/AppKit.h> static void printencoding(const char *enc) { printf("encoding is '%s' (%p)\n", enc, (void*)enc); } @implementation SWCleanup: NSObject struct cleanupentry { int type; union { struct dictcleanup { NSString *entry; NSString *key; } dict; struct notifcleanup { id obj; char *entry; void *data; } notif; } u; }; const char *encoding = @encode(struct cleanupentry); // if this method definition is removed, the problem goes away. - (void)sw_adddbnotification:obj when:(const char *)ent data:(void *)data { } - (void)sw_adddictentry:(NSString *)ent key:(NSString *)key { // these two lines *should* print the same thing, but do not. printencoding(@encode(struct cleanupentry)); printencoding(encoding); } @end main() { id obj; obj = [[SWCleanup alloc] init]; [obj sw_adddictentry:@"entry" key:@"key"]; }
From: Ben Perrault <ben.perrault@wmich.edu> Newsgroups: comp.sys.next.software,comp.soft-sys.nextstep,comp.sys.next.programmer Subject: Low Level Hardware Information... Date: Wed, 16 Sep 1998 14:53:55 -0500 Organization: Western Michigan University Message-ID: <36001753.872FEC16@wmich.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: eps@toaster.SFSU.EDU, sanguish@digifix.com, info@ord.com, sales@deepspacetech.com Hello, Since Apple has bought NeXT and stopped furthering the OS development, there has been a rising of a group to port Linux to NeXT Hardware. I was curious where I could find service manuals, or low level info about the hardware (web or print)? Since I currently own 2 NeXT Station Color Turbos, I have decided to use on for Linux Development and the other for that great NeXT enviorment (after all, that's why I bought the damn thing). Any suggestions/help? Ben Perrault Western Michigan University Computer Science - Theory and Analysis
From: Greg_Anderson@afs.com (Gregory H. Anderson) Newsgroups: comp.sys.next.programmer Subject: Re: Apparent compiler bug in NS 3.3 Followup-To: comp.sys.next.programmer Date: 16 Sep 1998 21:03:05 GMT Organization: Anderson Financial Systems Inc. Message-ID: <6tp929$5pn@shelob.afs.com> References: <6tp5st$m0v$1@dropit.pgh.net> me@venetia.pgh.pa.us writes > I was moving a program from the NeXT to the PC and was getting > unexpected output. The NeXT code was of the form > > if (x > y) > statemnt1; statement2; > > To obtain the same (expected) output on the PC I had to use > > if (x > y) { > statement1; statement2; > } > > Note, that the second code (PC) is correct. The code on the NeXT > should not have worked, but it did. Because it did, I never caught > the error until now. Can anybody duplicate this? I hate to say something is "impossible", but I can't imagine the old code was running the way your indentation suggests you wanted it to. The compiler does not care about white space and indentation. The ';' is an end-of-statement character, regardless of text position. I can find several hundred examples in my own code that should fail if your supposition about the top example is correct. (Followups set to c.s.n.programmer) -- Gregory H. Anderson | "We're in the land of the blind, Visionary Ophthalmologist | selling working eyeballs, and they Anderson Financial Systems | balk at the choice of color." -- Tony greg@afs.com (NeXTmail OK) | Lovell, on Mac user reactions to NeXT
From: tennant@alph.msfc.nasa.gov (Allyn Tennant) Newsgroups: comp.sys.next.programmer Subject: Re: I need my .h files!!! Date: 16 Sep 1998 22:22:38 GMT Organization: http://www.msfc.nasa.gov/ Message-ID: <6tpdne$r6p$1@hammer.msfc.nasa.gov> References: <6t6rjk$lj1$1@news.safari.net> dan@news.safari.net (Dan Nafe) wrote: >Is there someplace on the web I can download the .h files for my black >Openstep machine? I need the basick such as stdio.h, stdlib.h ... > >Please mail me the location the these files and libraies > If you want Next headers then you need to buy the developer's package. However, if all you want is a C compiler you might look at ftp://next-ftp.peak.org/pub/next/apps/devtools/djgpp.NSi386.README The associated tar file should include the basic headers. It is a port of the GJGCC compiler to NeXTStep. DJGCC is a port GCC to DOS (and for some strange reason, DOS doesn't come with the .h files either :-). Other comments: 1) It is a port of an old version gcc (2.6.0) 2) I never used it myself 3) Id (the guys who wrote Doom and Quake) paid for the NeXT port. They were big on NeXTStep when they were writing Doom. Allyn
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: Drawing over non-opaque subviews Date: 16 Sep 1998 22:56:48 GMT Organization: Idiom Communications Message-ID: <6tpfng$pqf$1@news.idiom.com> References: <nospam-1609981008350001@x84-168-86.ejack.umn.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: nospam@tc.umn.edu Kyle Hammond may or may not have said: -> Hi all, -> -> I am trying to get a custom view to draw over its subviews. Wrong goal. Try stating it as "I need to be able to draw over some existing views." -> I'm trying to -> implement something like the television football chalk feature - the user -> can draw over the window contents. The problem is that some of the -> subviews are non-opaque; thus the AppKit redraws the custom view first and -> then the subviews, placing the "chalk" under the subviews. How do I draw -> on top of the subviews? You don't. The update mechanism will always call the view that's higher in the tree (the superview) first. -> I know I could put another custom view on top of all views in the window -> and draw in that, That's what you'll have to do. -> but it would be nice if there was a simple method of doing some drawing in -> the current custom view after the subviews have drawn. That approach will make for a good deal more work, if you can make it work at all. Also, buggering the view hierarchy like this means that if someone after you wants to subclass your custom view, or even place something over it that he *doesn't* want obscured, he'll lose. -> I tried intercepting the display method and first calling [ super -> display ] and then drawing the chalk, but that had the same effect as -> drawing in the drawRect: method. And, it sounds like you're already standing on your head to do something the hard way. -> Currently, all other views in the window are subviews of my custom view. -> This is nice, since the custom view can intercept the mouseDown: method -> before it gets to the subviews and let the user draw over the entire -> window. All it takes to be able to intercept the mousedown is a subview that covers the window's content area, and is *deeper* in the view heirachy than any other view. -jcr -- "Although UNIX is more reliable, NT may become more reliable with time" - Ron Redman, deputy technical director of the Fleet Introduction Division of the Aegis Program Executive Office, US Navy. Where is the line between mere incompetence and actual treason? -jcr
From: dbk@mcs.com (Dan "Bud" Keith) Newsgroups: comp.sys.mac.programmer.codewarrior,comp.sys.mac.programmer.tools,comp.sys.next.programmer Subject: MacOSX Support (was [ANN] What's New with CodeWarrior Pro 4) Date: Wed, 16 Sep 1998 15:11:55 -0500 Organization: Emergent Systems Message-ID: <1dff2jw.j6drjmevt4ogN@dbklap.bb.opentext.com> References: <MWRon-1409981313470001@dyn1-tnt2-2.kalamazoo.mi.ameritech.net> MW Ron <MWRon@metrowerks.com> wrote: > What's new for CodeWarrior Professional Release 4 > > CodeWarrior IDE Version 3.2 > * New pre-release support for Mac OS X Server C/C++/ObjC command line > compiler and runtime libraries, can be used under Project Builder > * New Mac OS to Mach-O cross compiler, linker, and library importer > plug-ins for developing Mac OS X applications Please, please, please tell me that this means that MSL (Metrowerks Standard Library) is also available under MacOSX/Server. That would be wonderful. I played with the original pre-release earlier this summer but it only contained the C++ runtime (i.e., new/delete) and not the MSL. This meant, of course, that I couldn't even use iostream. Does CW4 have MacOSX/Server MSL? If not now, will it ever? Or are you guys planning on waiting a year for MacOSX? This leaves MacOSX/Server developers in the lurch (Apple's fault, not Metrowerks). Of course, since MacOSX/Server seems pretty dead-end, I can understand your decision. thanks for any info, bud ----- Dan Keith (dbk@mcs.com) -----
From: nospam@tc.umn.edu (Kyle Hammond) Newsgroups: comp.sys.next.programmer Subject: Drawing over non-opaque subviews Date: Wed, 16 Sep 1998 10:08:35 -0500 Organization: University of Minnesota Message-ID: <nospam-1609981008350001@x84-168-86.ejack.umn.edu> Hi all, I am trying to get a custom view to draw over its subviews. I'm trying to implement something like the television football chalk feature - the user can draw over the window contents. The problem is that some of the subviews are non-opaque; thus the AppKit redraws the custom view first and then the subviews, placing the "chalk" under the subviews. How do I draw on top of the subviews? I know I could put another custom view on top of all views in the window and draw in that, but it would be nice if there was a simple method of doing some drawing in the current custom view after the subviews have drawn. I tried intercepting the display method and first calling [ super display ] and then drawing the chalk, but that had the same effect as drawing in the drawRect: method. Currently, all other views in the window are subviews of my custom view. This is nice, since the custom view can intercept the mouseDown: method before it gets to the subviews and let the user draw over the entire window. Any help would be appreciated, Kyle -- Kyle Hammond Change nospam to hammo009 to email me. http://genbiol.cbs.umn.edu/staff/khammond.html
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6tovrr$hcv$368@hyperion.nitco.com> Control: cancel <6tovrr$hcv$368@hyperion.nitco.com> Date: 17 Sep 1998 06:51:56 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6tovrr$hcv$368@hyperion.nitco.com> Sender: ayctxbjw@somethingfunny.net Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Asagai <asagai@waksman.rutgers.edu> Newsgroups: comp.sys.next.software,comp.soft-sys.nextstep,comp.sys.next.programmer Subject: Re: Low Level Hardware Information... Date: Thu, 17 Sep 1998 11:22:43 -0400 Organization: Rutgers University Sender: luway@mbcl-mac-1.rutgers.edu Message-ID: <3601293E.E74C06D9@waksman.rutgers.edu> References: <36001753.872FEC16@wmich.edu> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="------------560A6AD3F0269538C7475479" This is a multi-part message in MIME format. --------------560A6AD3F0269538C7475479 Content-Type: text/plain; charset=us-ascii; x-mac-type="54455854"; x-mac-creator="4D4F5353" Content-Transfer-Encoding: 7bit I recommend....and I know this is not easy to shallow, but purchasing a Mac G3 machine and either run Mac OS Server, or get your hands on Rhapsody DR2 which I must state runs beautifully. I am stating this bad news because there are many individuals like yourself and myself that still own NeXT hardware and software. The only avenue for people like you and I as foreseen buy NeXT, now Apple, is this solution. I think you are going to be quite pleased.......At bare minimum, play around with a G3 running Rhapsody DR2, or CR1 of Mac OS Server. My main reason for writing this email is that what is in Linux already exists in Rhapsody DR2 and CR1 of Mac OS Server. Whatever your decision...good luck --------------560A6AD3F0269538C7475479 Content-Type: text/x-vcard; charset=us-ascii; name="asagai.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Asagai Content-Disposition: attachment; filename="asagai.vcf" begin:vcard adr;dom:;;190 Freylinghuysen Road;Piscataway;New Jersey;08854-8020 n:Asagai;B. x-mozilla-html:FALSE org:Waksman Institute version:2.1 email;internet:asagai@waksman.rutgers.edu title:Computing Analyst II tel;fax:732-445-5735 tel;home:212-665-3250 tel;work:732-445-4864 note:http://mbclserver.rutgers.edu/~asagai x-mozilla-cpt:;2 fn:B. Asagai end:vcard --------------560A6AD3F0269538C7475479--
From: fnikgsgv@somethingfunny.net Newsgroups: comp.sys.next.programmer Subject: WOW!!! WHAT A STORY ONLINE!!! Date: 17 Sep 1998 16:05:43 GMT Organization: Northwestern Indiana Telephone Co. Message-ID: <6trc0n$vt3$232@hyperion.nitco.com> I happen to have dropped by http://www.despotovic.net and I couldn't believe what I saw. A complete case online with over 48 pictures, 3 police reports and more!!! Regards. P.S.- The website address is http://www.despotovic.net
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6trc0n$vt3$232@hyperion.nitco.com> Control: cancel <6trc0n$vt3$232@hyperion.nitco.com> Date: 17 Sep 1998 16:05:35 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6trc0n$vt3$232@hyperion.nitco.com> Sender: fnikgsgv@somethingfunny.net Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: how do you create an editable subclass of NSControl? Date: 17 Sep 1998 19:25:45 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn702ohv.6d8.rog@talisker.ohm.york.ac.uk> [openstep 4.2 question] i've been trying to create a custom NSControl that contains an editable cell containing text, and i can't work out how one is supposed to go about it... (the NSControl documentation makes it sound so easy :-]) i've had a day of frustration; it would be very kind if someone could help. to start with, i've been experimenting with embedding an NSFormCell as the cell inside an NSControl... no luck, NSControl doesn't start editing it - the cell just highlights like a button... so i tried subclassing NSTextField instead, but that doesn't work properly when it's not using an NSTextFieldCell. my basic problem is that i don't find it very clear exactly what functionality NSControl implements inside its mouseDown: method, and what i need to implement myself. i don't want to override mouseDown completely (and lose NSControl's field validation functionality), but i can't seem to persuade it to begin editing its cell. has anyone got some example code that does this sort of thing? it should be so easy! some things i haven't been able to work out from the documentation: 1) does NSControl ever actually send any of the field validation delegate methods, or does it just declare them and leave it up to its subclasses to implement them? 2) if it does, then under what circumstances (and how, given that it doesn't possess its own delegate) 3) if it doesn't, then surely all the classes that implement an editable NSControl (NSMatrix, NSTextField, NSTableView) don't reimplement all those methods...? how can i be sure to reimplement the same heuristics? thanks for any help, rog.
Newsgroups: comp.sys.next.programmer Subject: The Clinton, Starr Report in FULL From: smlirqcm@starr-report.com Organization: Your Organization Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="PART_BOUNDARY_LOPNJYFVJU" Message-ID: <IxkM1.2523$bj5.220566@nsw.nnrp.telstra.net> Date: Fri, 18 Sep 1998 03:35:04 GMT NNTP-Posting-Date: Fri, 18 Sep 1998 13:35:04 EST --PART_BOUNDARY_LOPNJYFVJU Content-Type: text/html; charset=us-ascii; name="test.html" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="test.html" Content-Base: "file:///C|/test.html" <BASE HREF="file:///C|/test.html"> <HTML> <HEAD> <TITLE></TITLE> <SCRIPT language="JavaScript"> <!-- B = open("http://www.horny-world.com/jjj/") blur(B) //--> </SCRIPT> </HEAD> <BODY> </BODY> </HTML> --PART_BOUNDARY_LOPNJYFVJU Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Read the report, see the pictures! http://www.horny-world.com/jjj/
From: MWRon@metrowerks.com (MW Ron) Newsgroups: comp.sys.mac.programmer.codewarrior,comp.sys.mac.programmer.tools,comp.sys.next.programmer Subject: Re: MacOSX Support (was [ANN] What's New with CodeWarrior Pro 4) Date: Thu, 17 Sep 1998 22:59:17 -0400 Organization: Metrowerks Corporation Message-ID: <MWRon-1709982259180001@dyn1-tnt2-51.kalamazoo.mi.ameritech.net> References: <MWRon-1409981313470001@dyn1-tnt2-2.kalamazoo.mi.ameritech.net> <1dff2jw.j6drjmevt4ogN@dbklap.bb.opentext.com> In article <1dff2jw.j6drjmevt4ogN@dbklap.bb.opentext.com>, dbk@mcs.com (Dan "Bud" Keith) wrote: > MW Ron <MWRon@metrowerks.com> wrote: > > What's new for CodeWarrior Professional Release 4 > > > > CodeWarrior IDE Version 3.2 > > * New pre-release support for Mac OS X Server C/C++/ObjC command line > > compiler and runtime libraries, can be used under Project Builder > > * New Mac OS to Mach-O cross compiler, linker, and library importer > > plug-ins for developing Mac OS X applications > > Please, please, please tell me that this means that MSL (Metrowerks > Standard Library) is also available under MacOSX/Server. That would be > wonderful. Thats not what it means, however we are Carbonizing all/most CodeWarrior Pro including MSL and will have the tools in sync with Apple's Developer and Public MacOS X releases. Ron -- CodeWarrior Pro 4 - check it out - www.metrowerks.com/desktop/pro/ METROWERKS Ron Liechty "Software at Work" MWRon@metrowerks.com
From: cnyap@dcs.shef.ac.uk (Chih Nam Yap) Newsgroups: comp.sys.next.programmer Subject: How to write mathematical fonts to screen Date: 18 Sep 1998 10:14:47 GMT Organization: Department of Computer Science, University of Sheffield Message-ID: <6ttbqn$99n$1@bignews.shef.ac.uk> Hi there, I recently found a set of mathematical fonts from a ftp site. However, I do not know how to use this font in order to display them on screen. What I have done so far is I create a class called DisplayFont. In this class, I initialise an object called "font" in the "initFrame" method id font; font = [Font newFont: "Math-Sym" siae:12.0 style:1 matrix: NX_IDENTITYMATRIX]; and then in the "drawSelf" method I did this - drawSelf: (const NXRect *)rects :(int)count { [font set]; ....... ....... PSShow("A string"); } PSShow is defined as defineps PSSHow(char *s) (s) show Now, it is no problem for me to enter string that consists of normal alphabet from a to z or A to Z. My question is how could I show mathematical symbol such as "for all" or "set membership" these kinds of mathematical symbols ? Your help will be very much appreciated. Thank You. Chih Nam.
Newsgroups: comp.sys.next.programmer,comp.sys.next.software From: Per Liden <per@rsn.hk-r.se> Subject: Information about the loginpanel wanted! Message-ID: <Pine.LNX.3.96.980918175224.2155A-100000@pluto.rsn.hk-r.se> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Date: Fri, 18 Sep 1998 18:25:17 +0200 NNTP-Posting-Date: Fri, 18 Sep 1998 18:19:28 MET DST Organization: University of Karlskrona/Ronneby Hi, Could someone with nextstep 3.3 (or newer) send me the Power and Reboot button images used by the loginpanel. I'm also interested in knowing the exact procedure when pushing those buttons, like - What excatly does the text say in the panel that appears when pushing these buttons? Is it "Are you sure?" or something else? - What options are available in thses panels? Is it "Yes/No" or is it "Cancel/Reboot" or something else? - etc... A screenshoot of these panels would be very nice, by maybe hard to make since no applications is running when the panel is shown? Any information about this is welcomed! Thanks in advance! regards, Per ps. If you are wondering why I want this information it's because I'm making an imitation of the nextstep loginpanel for linux. And now I want to make it even closer to the "real thing". I haven't been using nexstep for a long time now, so I've forgotten the details of the loginpanel. ds.
From: dave@turbocat.de (David Wetzel) Newsgroups: comp.sys.next.programmer Subject: cancel Control: cancel <IxkM1.2523$bj5.220566@nsw.nnrp.telstra.net> Date: 19 Sep 1998 21:36:28 GMT Organization: Turbocat's Development http://www.turbocat.de/ Message-ID: <6u184s$crc$3@alice.turbocat.de> NNTP-Posting-Date: 19 Sep 1998 21:36:28 GMT cancel
From: Jonathan W Hendry <jhendry@shrike.depaul.edu> Subject: Scripting PB? Newsgroups: comp.sys.next.programmer Message-ID: <360476a1.0@news.depaul.edu> Date: 20 Sep 98 03:29:37 GMT Has anyone hooked Objective-Everything into ProjectBuilder and used it to script things? How about IB? - JH -- Note: email to this address goes to /dev/null To email a reply, write to jon at exnext dot com
From: x557@mindspringNOSPAM.com (x557@mindspringNOSPAM.com) Newsgroups: comp.sys.next.programmer Subject: Re: MacOSX Support (was [ANN] What's New with CodeWarrior Pro 4) Date: Sun, 20 Sep 1998 03:48:49 GMT Organization: MindSpring Enterprises Message-ID: <3601eadb.740531@news.mindspring.com> References: <MWRon-1409981313470001@dyn1-tnt2-2.kalamazoo.mi.ameritech.net> <1dff2jw.j6drjmevt4ogN@dbklap.bb.opentext.com> <MWRon-1709982259180001@dyn1-tnt2-51.kalamazoo.mi.ameritech.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Would someone tell me if Codewarrior is better than the OpenStep development environment? or am I mixing Apples and Oranges? Or maybe Codewarrior works with the openstep environment? Thank you. On Thu, 17 Sep 1998 22:59:17 -0400, MWRon@metrowerks.com (MW Ron) wrote: >In article <1dff2jw.j6drjmevt4ogN@dbklap.bb.opentext.com>, dbk@mcs.com >(Dan "Bud" Keith) wrote: > >> MW Ron <MWRon@metrowerks.com> wrote: >> > What's new for CodeWarrior Professional Release 4 >> > >> > CodeWarrior IDE Version 3.2 >> > * New pre-release support for Mac OS X Server C/C++/ObjC command line >> > compiler and runtime libraries, can be used under Project Builder >> > * New Mac OS to Mach-O cross compiler, linker, and library importer >> > plug-ins for developing Mac OS X applications >> >> Please, please, please tell me that this means that MSL (Metrowerks >> Standard Library) is also available under MacOSX/Server. That would be >> wonderful. > >Thats not what it means, however we are Carbonizing all/most CodeWarrior >Pro including MSL and will have the tools in sync with Apple's Developer >and Public MacOS X releases. > > >Ron
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <25036905659241@digifix.com> Date: 20 Sep 1998 03:47:47 GMT Organization: Digital Fix Development Message-ID: <24796906264032@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1994. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6u56k6$lfu$9405@hyperion.nitco.com> Control: cancel <6u56k6$lfu$9405@hyperion.nitco.com> Date: 21 Sep 1998 09:35:21 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6u56k6$lfu$9405@hyperion.nitco.com> Sender: sqkuqlsx@somethingfunny.net Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: toon@omnigroup.com (Greg Titus) Newsgroups: comp.sys.next.programmer Subject: Re: Autorelease not threadsafe? Date: 21 Sep 1998 21:18:58 GMT Organization: Omni Development, Inc. Message-ID: <6u6fs2$g22$1@gaea.omnigroup.com> References: <6u68c7$lpu$1@crib.corepower.com> Nathan Urban writes: > Suppose an object is created, autoreleased, in the main thread and > then returned to a method in a potentially (but not necessarily) > different thread. If it gets returned to a different thread, then > it's possible for the AppKit's runloop in the main thread to end and > clean up everything in the autorelease pool. Thus, the assumption > that if you're handed an autoreleased object, it's safe for the > duration of your scope is becomes an invalid assumption in a > multithreaded environment. Doesn't that ruin one of the main > advantages of autoreleasing, namely the fact that you don't have to > retain and release each and every single object you work with, if > you're just using them temporarily? Or is there some internal > magic going on with the autorelease pools that keeps this from > happening? Yes and no. There is a seperate autorelease pool stack for each thread, so any objects that are dealt with only by a single thread is going to follow the same rules as in the single threaded case. When you actually pass an object between threads you do need to be more careful. If you use Distributed Objects to invoke methods in another thread, the DO mechanics will solve this for you. If you are passing objects using some other method you will need to make sure that the objects are retained appropriately. For example, in the DO case, if you have something like: argument = [NSString stringWithCString:@"argument"]; [aProxyToAnotherThread someMethod:argument]; Note that the argument object is autoreleased in the main thread, so you might worry that it will go away before the other thread is done with it (assume that someMethod: is oneway so the sender thread isn't hanging around waiting for it to complete). What will actually happen is the proxy creates an NSInvocation to give to the other thread, and the invocation retains your argument. The invocation is autoreleased in the _other_ thread. The retain counting goes: 1) created: retain count 1, autoreleased in sender thread 2) retained by invocation: retain count 2, still autoreleased in sender 3) other thread done: invocation releases the argument so retain count goes down to 1, still autoreleased in sender 4) sender thread leaves scope: retain count 0, object dealloced Note how even if steps 3 and 4 are reversed (the sender thread leaves scope before the other thread is done working with the argument) you are still okay. (Because the invocation is still retaining the argument.) Now if you are accessing objects from more than one thread some other way (say by a class method) and those objects might go away, you do need to do some extra work to make sure to retain them in each thread where they are used. The straightforward way to do this is with code like: + someGlobalObject { return [[object retain] autorelease]; } No matter what thread +someGlobalObject is called from, the object is autoreleased _in that thread_ so the rules are exactly the same as in the single threaded case. Hope this helps, -Greg --------------------- Greg Titus Omni Development Inc. greg@omnigroup.com
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6u62u6$if$3522@artemis.backbone.ou.edu> Control: cancel <6u62u6$if$3522@artemis.backbone.ou.edu> Date: 22 Sep 1998 01:51:25 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6u62u6$if$3522@artemis.backbone.ou.edu> Sender: <myemail@any.where.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: nurban@crib.corepower.com (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Autorelease not threadsafe? Date: 21 Sep 1998 15:11:03 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6u68c7$lpu$1@crib.corepower.com> NNTP-Posting-Date: 21 Sep 1998 19:11:09 GMT Suppose an object is created, autoreleased, in the main thread and then returned to a method in a potentially (but not necessarily) different thread. If it gets returned to a different thread, then it's possible for the AppKit's runloop in the main thread to end and clean up everything in the autorelease pool. Thus, the assumption that if you're handed an autoreleased object, it's safe for the duration of your scope is becomes an invalid assumption in a multithreaded environment. Doesn't that ruin one of the main advantages of autoreleasing, namely the fact that you don't have to retain and release each and every single object you work with, if you're just using them temporarily? Or is there some internal magic going on with the autorelease pools that keeps this from happening?
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <8cdc.3275.28a@my_system> Control: cancel <8cdc.3275.28a@my_system> Date: 22 Sep 1998 07:33:01 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.8cdc.3275.28a@my_system> Sender: MarkW <ntec@mustbuy.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: Autorelease not threadsafe? Date: 21 Sep 98 14:20:38 Organization: Is a sign of weakness Message-ID: <SCOTT.98Sep21142038@slave.doubleu.com> References: <6u68c7$lpu$1@crib.corepower.com> In-reply-to: nurban@crib.corepower.com's message of 21 Sep 1998 15:11:03 -0400 In article <6u68c7$lpu$1@crib.corepower.com>, nurban@crib.corepower.com (Nathan Urban) writes: Suppose an object is created, autoreleased, in the main thread and then returned to a method in a potentially (but not necessarily) different thread. If it gets returned to a different thread, then it's possible for the AppKit's runloop in the main thread to end and clean up everything in the autorelease pool. Thus, the assumption that if you're handed an autoreleased object, it's safe for the duration of your scope is becomes an invalid assumption in a multithreaded environment. Not necessarily - how did the object get returned to another thread? Autoreleasing should only address how things work within specific threads. If you're transferring objects between threads, you should re-autorelease them in the new thread. It _could_ be a problem if you're gatewaying the objects between threads using an asynchronous method. For instance, if you sent the object as part of a Mach message send, and the thread continues executing after the message was sent, it could release the object before the message is received. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: <webmaster@ebay.com> Newsgroups: comp.sys.next.programmer Subject: Titanic on C Date: 22 Sep 1998 11:50:24 GMT Organization: The University of Oklahoma (USA) Message-ID: <6u82u0$d13$3522@artemis.backbone.ou.edu>
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6u4gf9$ctl$1847@hyperion.nitco.com> Control: cancel <6u4gf9$ctl$1847@hyperion.nitco.com> Date: 21 Sep 1998 06:00:56 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6u4gf9$ctl$1847@hyperion.nitco.com> Sender: qxichsqf@somethingfunny.net Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
Newsgroups: comp.sys.next.programmer From: maury@remove_this.istar.ca (Maury Markowitz) Subject: Setting fonts on a number of text views? Content-Type: text/plain; charset=us-ascii Message-ID: <Ezr0rB.CM5@T-FCN.Net> Sender: news@T-FCN.Net Content-Transfer-Encoding: 7bit Organization: none Mime-Version: 1.0 Date: Wed, 23 Sep 1998 18:01:57 GMT As is common in any number of other programs, I'm using a single shared instance of a text field in order to display various text labels and such. Many of our beta testers have asked for a way to set the font of a selection of items, so I started work on this a few days ago. I got a lot of other stuff done too, but this one still doesn't work. Basically each shape holds an attributed string. When it's time to edit it or display it, we move the text view to that spot and do whatever. The problem is that when you do this and insert the text from that shape, the font panel gets updated. So if you're trying to set the font of a selection of shapes, every time you load up the string to work on it the font panel is changed to that font and so much for that. I thought the "easy" solution was to tell the text view to ignore the font panel via setUsesFontPanel, but this simply doesn't work. I turn it off, load up the text and select it, and it still changes the panel. I know I can scan the string and find the font runs and such, but the intricacies of this is rather unclear and not terribly well documented. However this is what Andy did, so maybe I have to as well. The odd thing is that Draw seems to do what I do, but it doesn't need to use the setUsesFontPanel at all - it loads the string in question into some other NSText object entirely and uses that. Are these the only solutions? Maury
From: scro0249@sable.ox.ac.uk (Paolo Zuliani) Newsgroups: comp.sys.next.programmer Subject: Accessing sound Date: 24 Sep 1998 15:33:13 GMT Organization: Oxford University, England Message-ID: <6udonp$p3u$1@news.ox.ac.uk> NNTP-Posting-Date: 24 Sep 1998 15:33:13 GMT Hello, does NeXTStep use the /dev/audio to access the audio device? Is it full duplex? Many thanks. Best wishes, Paolo
From: dfevans@bbcr.uwaterloo.ca (David Evans) Newsgroups: comp.sys.next.programmer Subject: Re: Accessing sound Date: 24 Sep 1998 15:55:08 GMT Organization: University of Waterloo Message-ID: <906652508.629082@watserv4.uwaterloo.ca> References: <6udonp$p3u$1@news.ox.ac.uk> Cache-Post-Path: watserv4.uwaterloo.ca!unknown@bcr11.uwaterloo.ca In article <6udonp$p3u$1@news.ox.ac.uk>, Paolo Zuliani <scro0249@sable.ox.ac.uk> wrote: >Hello, > does NeXTStep use the /dev/audio to access the audio device? No. >Is it full duplex? NeXTSTEP can do full-duplex sound, hardware and drivers permitting. -- David Evans (NeXTMail/MIME OK) dfevans@bbcr.uwaterloo.ca Computer/Synth Junkie http://bbcr.uwaterloo.ca/~dfevans/ University of Waterloo "Default is the value selected by the composer Ontario, Canada overridden by your command." - Roland TR-707 Manual
From: scro0249@sable.ox.ac.uk (Paolo Zuliani) Newsgroups: comp.sys.next.programmer Subject: Re: Accessing sound Date: 24 Sep 1998 16:49:23 GMT Organization: Oxford University, England Message-ID: <6udt6j$706$1@news.ox.ac.uk> References: <6udonp$p3u$1@news.ox.ac.uk> <906652508.629082@watserv4.uwaterloo.ca> NNTP-Posting-Date: 24 Sep 1998 16:49:23 GMT David Evans (dfevans@bbcr.uwaterloo.ca) wrote: : In article <6udonp$p3u$1@news.ox.ac.uk>, : Paolo Zuliani <scro0249@sable.ox.ac.uk> wrote: : >Hello, : > does NeXTStep use the /dev/audio to access the audio device? : No. : >Is it full duplex? : NeXTSTEP can do full-duplex sound, hardware and drivers permitting. I think I have not stated the question in the right way, sorry about that! Does NeXTStep provide /dev/audio to access the audio device? Thanks. Best regards, Paolo Zuliani
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.next.programmer Subject: Re: Accessing sound Date: 24 Sep 1998 17:34:05 GMT Organization: MiscKit Development Message-ID: <6udvqd$5tu$1@news.xmission.com> References: <6udonp$p3u$1@news.ox.ac.uk> <906652508.629082@watserv4.uwaterloo.ca> <6udt6j$706$1@news.ox.ac.uk> NNTP-Posting-Date: 24 Sep 1998 17:34:05 GMT scro0249@sable.ox.ac.uk (Paolo Zuliani) wrote: > David Evans (dfevans@bbcr.uwaterloo.ca) wrote: > : In article <6udonp$p3u$1@news.ox.ac.uk>, > : Paolo Zuliani <scro0249@sable.ox.ac.uk> wrote: > : >Hello, > : > does NeXTStep use the /dev/audio to access the audio device? > > : No. > > : >Is it full duplex? > > : NeXTSTEP can do full-duplex sound, hardware and drivers permitting. > > I think I have not stated the question in the right way, sorry about that! > Does NeXTStep provide /dev/audio to access the audio device? > Thanks. For both NEXTSTEP and OPENSTEP there is a /dev/sound but I don't think you can just cat it to a file to record from it nor can you cat stuff into it for playback. You should typically use the "sndrecord" and "sndplay" commands or the SoundKit APIs instead... -- Later, -Don Yacktman don@misckit.com <a href="http://www.yacktman.org/don/index.html">My home page</a>
From: dfevans@bbcr.uwaterloo.ca (David Evans) Newsgroups: comp.sys.next.programmer Subject: Re: Accessing sound Date: 24 Sep 1998 16:57:58 GMT Organization: University of Waterloo Message-ID: <906656278.378443@watserv4.uwaterloo.ca> References: <6udonp$p3u$1@news.ox.ac.uk> <906652508.629082@watserv4.uwaterloo.ca> <6udt6j$706$1@news.ox.ac.uk> Cache-Post-Path: watserv4.uwaterloo.ca!unknown@bcr11.uwaterloo.ca In article <6udt6j$706$1@news.ox.ac.uk>, Paolo Zuliani <scro0249@sable.ox.ac.uk> wrote: > >I think I have not stated the question in the right way, sorry about that! >Does NeXTStep provide /dev/audio to access the audio device? Sorry--the answer's still no. ;-) -- David Evans (NeXTMail/MIME OK) dfevans@bbcr.uwaterloo.ca Computer/Synth Junkie http://bbcr.uwaterloo.ca/~dfevans/ University of Waterloo "Default is the value selected by the composer Ontario, Canada overridden by your command." - Roland TR-707 Manual
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Mixing Objective-C and C++ Date: 25 Sep 1998 04:22:20 GMT Organization: Idiom Communications Message-ID: <6uf5ps$8od$4@news.idiom.com> References: <360b03cf.313699317@news.charm.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: discord@charm.net Karl Hsu may or may not have said: -> According to at Apple's Documentation, "Objective-C can also be used -> as an extension to C++", incorporating C++ (and C and, now, Java) -> directly into Objective-C code. How far does this ability go? Would it -> be possible to use C++'s ability to do Multiple Inheritance to do -> Multiple Inheritance with Objective-C? Or does the ability to mix -> constructs only extend as far language semantics, as opposed to -> language constructs? This isn't a short-answer question, but briefly: The GCC supplied with NEXTSTEP (and Mac OS X ) is capable of compiling a language that comprises all of the grammar of C, C++, and Objective-C. In a typical application that combines C++ and Objective-C code, you will find C++ objects and Objective-C objects invoking each other's methods, but you will not find any objects which are simultaneously members of a C++ class and an Objective-C class. (Note: This *can* be done, but it shouldn't be done.) If you need to combine C++ and Objective-C, write your UI in Objective-C. If you're porting a C++ app, throw away your UI code and write it in Obj-C. It will get *much* smaller, simpler, and easier to maintain. If you're writing a new app, write it entirely in Obj-C. As for multiple inheritance: it's a bug, not a feature. If you think you need multiple inheritance, revise your design, because it's *broken.* -jcr -- What the Bard *should* have said: "All the World's a Stage, and all the Men and Women rather poorly rehearsed."
From: dfevans@bbcr.uwaterloo.ca (David Evans) Newsgroups: comp.sys.next.programmer Subject: Re: Accessing sound Date: 24 Sep 1998 19:00:21 GMT Organization: University of Waterloo Message-ID: <906663621.197829@watserv4.uwaterloo.ca> References: <6udonp$p3u$1@news.ox.ac.uk> <906652508.629082@watserv4.uwaterloo.ca> <6udt6j$706$1@news.ox.ac.uk> <6udvqd$5tu$1@news.xmission.com> Cache-Post-Path: watserv4.uwaterloo.ca!unknown@bcr11.uwaterloo.ca In article <6udvqd$5tu$1@news.xmission.com>, Don Yacktman <don@misckit.com> wrote: > >For both NEXTSTEP and OPENSTEP there is a /dev/sound but I don't think you >can just cat it to a file to record from it nor can you cat stuff into it for >playback. No, you can't: gallifrey:/Users/dfevans# dd if=/etc/hostconfig of=/dev/sound write: No such device 1+0 records in 1+0 records out -- David Evans (NeXTMail/MIME OK) dfevans@bbcr.uwaterloo.ca Computer/Synth Junkie http://bbcr.uwaterloo.ca/~dfevans/ University of Waterloo "Default is the value selected by the composer Ontario, Canada overridden by your command." - Roland TR-707 Manual
From: discord@charm.net (Karl Hsu) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Mixing Objective-C and C++ Date: Fri, 25 Sep 1998 05:37:18 GMT Organization: Charm.Net Baltimore Internet Access, Hon (410) 558-3900 Message-ID: <360b29d5.323434583@news.charm.net> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit jcr.remove@this.phrase.idiom.com (John C. Randolph) wrote: > Karl Hsu may or may not have said: >-> According to at Apple's Documentation, "Objective-C can also be used >-> as an extension to C++", incorporating C++ (and C and, now, Java) >-> directly into Objective-C code. How far does this ability go? Would it >-> be possible to use C++'s ability to do Multiple Inheritance to do >-> Multiple Inheritance with Objective-C? Or does the ability to mix >-> constructs only extend as far language semantics, as opposed to >-> language constructs? >The GCC supplied with NEXTSTEP (and Mac OS X ) is capable of compiling a >language that comprises all of the grammar of C, C++, and Objective-C. > >In a typical application that combines C++ and Objective-C code, you will >find C++ objects and Objective-C objects invoking each other's methods, but >you will not find any objects which are simultaneously members of a C++ class >and an Objective-C class. (Note: This *can* be done, but it shouldn't be >done.) > >As for multiple inheritance: it's a bug, not a feature. If you think you >need multiple inheritance, revise your design, because it's *broken.* I have to ask: why is it broken? There are only a few problems with Objective-C that I have right now: 1. Lack of GC (this is a personal thing, especially since I'm having problems with the whole reference counting thing right now). 2. Lack of MI. I agree that a great deal of the time, its difficult to implement, but I spend some time poking at Eiffel and their implementation seemed reasonable. Primarily I feel that while protocols allow for abstract MI, I should be able to inheirit working code. Objects in the real world are different things to different people - why not in a program? In all fairness, I have yet to extensively use the concept of a proxy object to replace MI. On the other hand, if the proxy concept can replace MI, why couldn't it replace Inheritance in general? If instead of inheritance, you had new user objects simply be proxies to the objects they wanted to "inherit" and then adhere to the same protocols, would that work? Example: class A <protocolA> {...} class B <protocolB> {...} class C <protocolA> {A a} <- Single Inheritance example class D <protocolA, ProtocolB> {A a, B b, ...} <- MI example Is there anyplace that actual inheritance would be better? Karl Hsu
From: discord@charm.net (Karl Hsu) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Mixing Objective-C and C++ Date: Fri, 25 Sep 1998 02:47:22 GMT Organization: Charm.Net Baltimore Internet Access, Hon (410) 558-3900 Message-ID: <360b03cf.313699317@news.charm.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit According to at Apple's Documentation, "Objective-C can also be used as an extension to C++", incorporating C++ (and C and, now, Java) directly into Objective-C code. How far does this ability go? Would it be possible to use C++'s ability to do Multiple Inheritance to do Multiple Inheritance with Objective-C? Or does the ability to mix constructs only extend as far language semantics, as opposed to language constructs? thanks, Kar Hsu
From: discord@charm.net (Karl Hsu) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Mixing Objective-C and C++ Date: Fri, 25 Sep 1998 11:50:15 GMT Organization: Charm.Net Baltimore Internet Access, Hon (410) 558-3900 Message-ID: <360d81cd.345958292@news.charm.net> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434583@news.charm.net> <6ufrvn$1sl$1@news.idiom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit jcr.remove@this.phrase.idiom.com (John C. Randolph) wrote: > Karl Hsu may or may not have said: >-> >As for multiple inheritance: it's a bug, not a feature. If you think you >-> >need multiple inheritance, revise your design, because it's *broken.* >-> >-> I have to ask: why is it broken? > >Because it introduces complexity without any benefit. I agree that it introduces a lot of complexity (mostly for the compiler writer, some for the programmer), but it seems that it gives some immense benifits: I acutally reuse an implementation that (potentially) someone else has written, instead of just the interface. >When a class inherits from multiple superclasses, there is ambiguity in >invoking inherited methods. Resolving this ambiguity requires that the >subclass state which of its inherited methods it's overriding or invoking, >which necessarily means that the subclass must be aware of where inherited >methods are implemented. To be honest, I haven't used MI that much. Most of programming has been done in C and Java, neither of which allow MI (for different reasons :). Still, when I inherit from a class, don't I do so explicitly because of it's behavior? In this case, that should be because of its methods. Thus, when I inherit from multiple classes, I should definitely know where inherited methods are implemented - I should just never know _how_ inherited methods are implemented. And if I rename multiply inherited methods during class implementation, I never do. >The kinds of things that C++ coders do with MI, Obj-C coders do with >delegation, protocols, and/or forwarding to "owned" instances of other >classes. > One could argue that requiring 3 language constructs to replace one construct is not necessarily simplifying. OTOH, splitting the MI functionality into 3 distinct parts may make it more useful (tho possibly less powerful) in the long run. Karl Hsu
From: gene@laa.com Newsgroups: comp.sys.next.programmer Subject: Server threads will not go away Date: 25 Sep 1998 15:32:51 GMT Organization: PSINet Message-ID: <6ugd33$s4n$1@client3.news.psi.net> We have implemented client-server programs based on the simple multithreaded example provided by Next. I have appended it to the end of this missive. Each client gets its own thread. Our problem is that when the client exits, the thread remains. Over time, as clients connect and exit, the server accumulates useless threads. How do I get rid of them ? =================================================================== #import <objc/Object.h> #import <remote/NXConnection.h> #import <remote/NXProxy.h> #import <mach/cthreads.h> #import <machkit/NXPort.h> #import <stdio.h> /* * A simple, multi-threaded example. * Each new client gets a dedicated thread. */ @protocol ServerProtocol - newThread; - (int) doWork; @end @interface MTS : Object <ServerProtocol,NXSenderIsInvalid> @end @implementation MTS // when a new client shows up, build a private connection to it - connection: (NXConnection *)oldConn didConnect:(NXConnection *)newConn { id outPort = [newConn outPort]; NXProxy *bogus = [NXConnection connectToPort:outPort]; NXConnection *newnewConn = [bogus connectionForProxy]; [newConn free]; // don't need new connection on old inPort [newnewConn runInNewThread]; // service new things in new thread [newnewConn setRoot:self]; // give new conn a root to serve! [newnewConn registerForInvalidationNotification:self]; return newnewConn; // service this first request in main thread } - newThread { return self; // this will point to newnewConn (and new thread) } - (int) doWork { return (int) cthread_self(); } - senderIsInvalid:sender { [sender free]; return self; } @end void main() { MTS *mts; NXConnection *server; id proxy1, proxy2; // [NXConnection debug:"MTS "]; mts = [MTS new]; server = [NXConnection registerRoot:mts withName:"MTS example"]; if (server) { // we're it! // start up a new thread that will deliver client death notifications [NXPort worryAboutPortInvalidation]; // tell initial connection to tell us about new incoming connections [server setDelegate:mts]; // serve requests [server run]; } else { [mts free]; proxy1 = [NXConnection connectToName:"MTS example"]; if (!proxy1) { printf("couldn't get to server\n"); exit(1); } [proxy1 setProtocolForProxy:@protocol(ServerProtocol)]; proxy2 = [proxy1 newThread]; if (!proxy2) { printf("couldn't get new thread\n"); exit(2); } [proxy2 setProtocolForProxy:@protocol(ServerProtocol)]; printf("new thread is %x\n", [proxy2 doWork]); } exit(0); }
From: "G. Angus McCollum" <mccollum@staff.juno.com> Newsgroups: comp.sys.next.programmer Subject: Re: Openstep NT: dll, lib, exp files and all that Date: Thu, 24 Sep 1998 16:50:59 -0400 Organization: D. E. Shaw & Co, L.P. Message-ID: <6uebp2$8n9$1@news1.deshaw.com> References: <3602551B.F87EC751@energotec.de> I the Windows world users do not usually program, so the .dll file is the only one needed to run the program. The lib file is only usefull for developers, and is normally not distributed with an application. - Angus Constantin Szallies wrote in message <3602551B.F87EC751@energotec.de>... >Hi everybody, >I'm just porting an Openstep/Mach application to Openstep/NT. No big >modifications in the sourcecode are needed, just the build process gives >me a headache! > >I have a dynamic library which I developed with Openstep/Mach. If I >compile this library with NT, make creates a dll, exp and lib file. > >1) As far as I understand I need these files for the following purpose: > - The exp file is needed by libtool to create the library > - The lib file is needed to link an application that uses my library > - The dll file is needed to start the application that was linked > against my lib file. >Is this true? I didn't find any detailed information about all these >things in the Openstep NT documentation. > >2) Why doesn't the makefile install the lib file? It only installs the >dll files? > >3) I have a palette. Under Openstep/Mach, I modified the makefiles to >create a dylib for the classes included in the NIB file. Unter NT, again >only a dll is installed, no lib. > >Thanx for any replies! > >Greetings >cs
From: holger@object-factory.REMOVETHIS.com (Holger Hoffstaette) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Mixing Objective-C and C++ Date: 25 Sep 1998 08:31:53 GMT Organization: the secret circle of the NSRC Message-ID: <6ufkdp$9l6$1@leonie.object-factory.com> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434583@news.charm.net> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit I am not going to address the issue of MI (since it too closesly depends on personal preference), but: > There are only a few problems with Objective-C that I have right now: > 1. Lack of GC (this is a personal thing, especially since I'm having > problems with the whole reference counting thing right now). I can tell you from my experience with both GC- and non-GC languages that if you 'have problems with refcounting', GC will *NOT* solve your problems; it will only lure you into thinking that the problems went away. I recommend you try to think really long and really hard about why refcounting (especially the way it's done in Foundation, re: autorelease) is *easier* to use than traditional alloc/dealloc (or new/delete in C++ parlor). Holger
From: David Stes <stes@mundivia.es> Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Mixing Objective-C and C++ Date: Sun, 27 Sep 1998 08:55:05 +0000 Organization: Mundivia, Santander Message-ID: <360DFD69.41C67EA6@mundivia.es> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434583@news.charm.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Karl Hsu wrote: > > There are only a few problems with Objective-C that I have right now: > 1. Lack of GC (this is a personal thing, especially since I'm having > problems with the whole reference counting thing right now). This truly isn't an Objective-C problem. You could get for example the book by Pinson and Wiener about Objective-C from http://www.amazon.com, the title is Objective-C: Object Oriented Programming Techniques, (it's a good book anyway) and you will see that the autorelease instance methods for reference counting are not part of the Objective-C language itself.
From: jik- <fract@sprintmail.com> Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Mixing Objective-C and C++ Date: Sun, 27 Sep 1998 13:27:30 -0700 Organization: EarthLink Network, Inc. Message-ID: <360E9FB2.6D718960@sprintmail.com> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434583@news.charm.net> <360DFD69.41C67EA6@mundivia.es> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit David Stes wrote: > Karl Hsu wrote: > > > > There are only a few problems with Objective-C that I have right now: > > 1. Lack of GC (this is a personal thing, especially since I'm having > > problems with the whole reference counting thing right now). > > This truly isn't an Objective-C problem. You could get for example the > book by Pinson and Wiener about Objective-C from http://www.amazon.com, > the title is Objective-C: Object Oriented Programming Techniques, (it's > a good book anyway) and you will see that the autorelease instance > methods for reference counting are not part of the Objective-C language > itself. I believe GC is being integrated into gcc, and that it is already in the egcs snapshot. Besides that, isn't the same GC used for C++ the one in the Objective-C faqs?? Objective-C has no standard, so some Obj-C compilers will have GC while others will not.... THAT is the only problem I see with Objective-C.
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Mixing Objective-C and C++ Date: 25 Sep 1998 13:24:55 GMT Organization: Idiom Communications Message-ID: <6ug5j7$1sl$3@news.idiom.com> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434583@news.charm.net> <6ufrvn$1sl$1@news.idiom.com> <360d81cd.345958292@news.charm.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: discord@charm.net Karl Hsu may or may not have said: [snip] -> To be honest, I haven't used MI that much. Good! It's a habit you should avoid, for the sake of anyone who has to maintain your code (including you.) -> Most of programming has been done in C and Java, neither of which allow MI -> (for different reasons :). Well, Java doesn't allow it, because Gosling knows better. C of course, has no concept of inheritance, so there's no *multiple* inheritance. -> Still, when I inherit from a class, don't I do so explicitly because -> of it's behavior? In this case, that should be because of its methods. This is a tautology... -> Thus, when I inherit from multiple classes, I should definitely know -> where inherited methods are implemented - I should just never know -> _how_ inherited methods are implemented. No, You should only need to know *that* you have an inherited method, not how far back in the tree it was implemented, nor which ancestor it was defined in (in the case of MI.) Making your code dependent on finding a particular implementation of a method somewhere in your inheritance tree breaks encapsulation and makes modification more difficult. ->And if I rename multiply -> inherited methods during class implementation, I never do. -> -> >The kinds of things that C++ coders do with MI, Obj-C coders do with -> >delegation, protocols, and/or forwarding to "owned" instances of other -> >classes. -> > -> One could argue that requiring 3 language constructs to replace one -> construct is not necessarily simplifying. It's not really a matter of *replacing* MI with these constructs. MI is something you *don't* need. Delegation, protocols, and forwarding are devices that address the kinds of tasks that C++ coders tend to do with MI, because they don't know any better. I'm going to have to defer this explanation until you've written some Obj-C code. Obj-C is a *much* simpler language than C++. -jcr -- What the Bard *should* have said: "All the World's a Stage, and all the Men and Women rather poorly rehearsed."
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Mixing Objective-C and C++ Date: 25 Sep 1998 13:30:01 GMT Organization: Idiom Communications Message-ID: <6ug5sp$1sl$4@news.idiom.com> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b7edd.345205632@news.charm.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: discord@charm.net Karl Hsu may or may not have said: -> How about this scenario: You have 2 objective-c classes that you want -> to inherit from. First, you create a c++ class that inherits from -> both. Well, that's poor design for a start. ->Then you create an Objective-C class that inherits from the c++ -> class. Sorry, you *can't* make an Objective-C class inherit from a C++ class. ->Will that work? No. ->What will happen? Look at it this way: The entire binding of a structure in memory to an Objective-C class is via the "isa" pointer, which is the first instance variable in any Objective-C instance. This is how Obj-C does runtime binding. So, for an object to receive Objective-C method invocations, it must begin with a pointer to an objective-C class. For an object to be the target of a C++ method invocation, it has to be declared or blessed as a member of the C++ class. If you want the same piece of memory to be *both*, you have to satisfy both conditions. You can't make an Obj-C class inherit from a C++ class, because the C++ class can't respond to Obj-C method invocations. (Note that I'm talking about the *classes* here, not the instances.) If you want to understand the Obj-C runtime code in depth, go read the GNU Obj-C runtime source. NeXT's implementation isn't coded *quite* the same way, but both runtimes do the same things. -jcr -- What the Bard *should* have said: "All the World's a Stage, and all the Men and Women rather poorly rehearsed."
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Mixing Objective-C and C++ Date: 25 Sep 1998 13:40:16 GMT Organization: Idiom Communications Message-ID: <6ug6g0$1sl$5@news.idiom.com> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b7edd.345205632@news.charm.net> <6ug5sp$1sl$4@news.idiom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: jcr.remove@this.phrase.idiom.com John C. Randolph may or may not have said: -> Karl Hsu may or may not have said: -> -> -> How about this scenario: You have 2 objective-c classes that you want -> -> to inherit from. First, you create a c++ class that inherits from -> -> both. Whoa! I didn't read this closely the first time. You can't derive a C++ class from an Objective-C class. The type of the Objective-C class is a "Class" which is typdef'd as typedef struct objc_class *Class; <- note this is a struct, not a "class" And a C++ class can't inherit from a struct. -jcr -- What the Bard *should* have said: "All the World's a Stage, and all the Men and Women rather poorly rehearsed."
From: discord@charm.net (Karl Hsu) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Mixing Objective-C and C++ Date: Fri, 25 Sep 1998 14:07:35 GMT Organization: Charm.Net Baltimore Internet Access, Hon (410) 558-3900 Message-ID: <360ba21d.354231470@news.charm.net> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434583@news.charm.net> <6ufrvn$1sl$1@news.idiom.com> <360d81cd.345958292@news.charm.net> <6ug5j7$1sl$3@news.idiom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit jcr.remove@this.phrase.idiom.com (John C. Randolph) wrote: >-> Still, when I inherit from a class, don't I do so explicitly because >-> of it's behavior? In this case, that should be because of its methods. >-> Thus, when I inherit from multiple classes, I should definitely know >-> where inherited methods are implemented - I should just never know >-> _how_ inherited methods are implemented. > >No, You should only need to know *that* you have an inherited method, not >how far back in the tree it was implemented, nor which ancestor it was >defined in (in the case of MI.) Making your code dependent on finding a >particular implementation of a method somewhere in your inheritance tree >breaks encapsulation and makes modification more difficult. I think I'm lost here. Why do I need to know how far back in the inheritance tree a method is implemented? I define renaming as follows: given n methods that have the same signature that are inherited from my parents, where n > 1, i must explicity declare in the class implementation which method I am "choosing" to inherit. So if I use renaming to do MI, I shouldn't have to pay any attention to anyone but my direct parents, right? I mean, the same thing applies when I do single inheritance - I should still know which methods I have are inherited, and of those which ones I'm overridding. In fact, thats _all_ I should know about the parent class(es). Or am I missing something? BTW - I understand about the C++ / Objective-C mixed inheritance thing now. I forgot that Objective-C was implemented at (more or less) a C level. Karl
From: nurban@crib.corepower.com (Nathan Urban) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Mixing Objective-C and C++ Date: 25 Sep 1998 11:38:07 -0400 Organization: Data Systems Consulting, Inc. Message-ID: <6ugdcv$80$1@crib.corepower.com> References: <360b03cf.313699317@news.charm.net> <6ufkdp$9l6$1@leonie.object-factory.com> <360f8387.346399652@news.charm.net> <6ugarg$nfo$1@leonie.object-factory.com> NNTP-Posting-Date: 25 Sep 1998 15:37:42 GMT In article <6ugarg$nfo$1@leonie.object-factory.com>, holger@object-factory.REMOVETHIS.com (Holger Hoffstaette) wrote: > The concept of basic refcounting is simple, as you said. The concept of > Foundation's refcounting is slightly more difficult to understand *if > you only compare it to traditional refcounting*. Having only had experience with Foundation's reference counting, how does it differ from traditional/naive refcounting?
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Mixing Objective-C and C++ Date: 25 Sep 1998 16:19:45 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn70ngl6.9gd.rog@talisker.ohm.york.ac.uk> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <6ugarg$nfo$1@leonie.object-factory.com> On 25 Sep 1998 14:54:40 GMT, Holger Hoffstaette <holger@object-factory.REMOVETHIS.com> wrote: > Normal refcounting is not really helpful in a truly dynamic > environment; the mechanism in Foundation definitely is, *if you > understand its rationale*. That means, among other things, reading the > documentation and thinking hard for at least 5 minutes. normal refcounting *can* be very helpful in a truly dynamic environment, if it's done well, and has language support. for an example of how it can be done well, see the language "limbo" in the Inferno environment which has language support for refcounting. most data structures in that language are refcounted, and the language ensures that no cyclic references are possible. if you *need* a potentially self-referential structure, you can do it - such structures are cleared up by a more conventional GC. this model allows the small memory footprint that refcounting allows without being prone to the errors that manual refcounting in the openstep style is. under openstep i find it's very easy indeed to get unintentional cycles, often because system objects retain their arguments when i don't expect them to. for instance the NSNotification object retains the object that's sending it - this means that it's not possible to create an NSNotification once and send it multiple times (avoiding its creation overhead) because it immediately creates a circular reference when you retain it. when declaration, initialisation, finalisation and use are all in separate methods or files, it is incredibly easy to make a small mistake, forgetting to release the object at the right time. i find it's more straightforward with standard malloc/free. also, allocating and freeing an autorelease pool involves appreciable overhead, so there are circumstances (recursive functions for example) where it's quite difficult to avoid a large buildup of unfreed objects without significant runtime penalty. all that said, the scheme does work, and it is just about the only such scheme that could work without altering the language runtime support in the slightest... cheers, rog. PS limbo docs can be found at: http://www.lucent-inferno.com/Pages/Developers/Documentation/Limbo_Ref20/index.html PPS if anyone's interested, i've just knocked up a simple program that tracks references in a running openstep program and can at any point print out the addresses of current instances of particular classes. email me for the source. i've found it very useful for tracking down exactly *which* objects have been leaked, something which neither ObjectAlloc.app or AnalyseAllocation seem to be able to do...
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Mixing Objective-C and C++ Date: 25 Sep 1998 16:25:46 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn70nh0f.9gd.rog@talisker.ohm.york.ac.uk> References: <360b03cf.313699317@news.charm.net> <6ufkdp$9l6$1@leonie.object-factory.com> <360f <6ugdcv$80$1@crib.corepower.com> On 25 Sep 1998 11:38:07 -0400, Nathan Urban <nurban@crib.corepower.com> wrote: > Having only had experience with Foundation's reference counting, how > does it differ from traditional/naive refcounting? traditional refcounting doesn't involve the use of autorelease pools, so it's not possible to return an object that you don't own because as soon as you don't own it, it goes away. autoreleasepools are a hack that enables the objects to be "owned" while they're in limbo between callee and caller. the tradeoff is that you lose one of the principal advantages of refcounting; the objects don't go away immediately they're not owned. cheers, rog.
From: waggonem@hartmanis.name mathcs.carleton.edu (Mike Waggoner) Newsgroups: comp.sys.next.programmer Subject: c++ and fork, pipe, execvp Date: 25 Sep 98 20:28:03 GMT Organization: [poster's organization not specified] Message-ID: <360bfcd3.0@news.carleton.edu> I want to use fork, pipe, execvp and all the other functions related to those with c++ and they aren't being found. any suggestions on where to point the compilor/how to get these to work.. thanks -mike mikew@iea.com hartmanis> g++ -o rip rip.cpp -lgcc rip.cpp: In function `int makepipe(int *, int *)': rip.cpp:13: warning: implicit declaration of function `int pipe(...)' rip.cpp: In function `int givecmd(char *)': rip.cpp:26: warning: implicit declaration of function `int write(...)' rip.cpp: In function `int fetchresult()': rip.cpp:41: warning: implicit declaration of function `int read(...)' rip.cpp: In function `int main()': rip.cpp:62: warning: implicit declaration of function `int fork(...)' rip.cpp:73: warning: implicit declaration of function `int close(...)' rip.cpp:88: warning: implicit declaration of function `int dup2(...)' rip.cpp:94: warning: implicit declaration of function `int execvp(...)'
From: uli@tallowcross.uni-frankfurt.de (Uli Zappe) Newsgroups: comp.sys.next.programmer Subject: 3.3 objects still of any interest? Date: 25 Sep 1998 21:53:38 GMT Organization: NEXTTOYOU Message-ID: <6uh3d2$67d$1@tallowcross.uni-frankfurt.de> NNTP-Posting-Date: 25 Sep 1998 21:53:38 GMT Hi, as a side effect of a project that I still had to do in 3.3, I have written a browser delegate object that allows to easily read all kinds of data into an NXBrowser object. The thing is very efficient, easy to use and well documented. It's built as a palette that even allows a fully functional NXBrowser in IB test mode. BUT - it's still 3.3. (And converting to OPENSTEP is non-trivial not only because of NSBrowser's different API, but because the object makes heavy use of the Storage class.) Is there still any interest for such a beast? Then I could put it on a server. Bye Uli -- _____________________________________________________________________ Uli Zappe E-Mail: uli@tallowcross.uni-frankfurt.de (NeXTMail,Mime,ASCII) PGP on request Lorscher Strasse 5 WWW: - D-60489 Frankfurt Fon: +49 (69) 9784 0007 Germany Fax: +49 (69) 9784 0042 _____________________________________________________________________
From: Michael Simpson <t_msimps@qualcomm.com> Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Mixing Objective-C and C++ Date: Fri, 25 Sep 1998 14:58:47 -0700 Organization: Qualcomm Message-ID: <360C1217.B1F8ABC1@qualcomm.com> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b7edd.345205632@news.charm.net> <6ug5sp$1sl$4@news.idiom.com> <6ug6g0$1sl$5@news.idiom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit John C. Randolph wrote: > > John C. Randolph may or may not have said: > -> Karl Hsu may or may not have said: > -> > -> -> How about this scenario: You have 2 objective-c classes that you want > -> -> to inherit from. First, you create a c++ class that inherits from > -> -> both. > > Whoa! I didn't read this closely the first time. > > You can't derive a C++ class from an Objective-C class. The type of the > Objective-C class is a "Class" which is typdef'd as > > typedef struct objc_class *Class; <- note this is a struct, not a "class" > > And a C++ class can't inherit from a struct. Sure it can! A class is a struct, but the default access control is private. A struct on the other hand is public. In C++ this is handy. I've inherited from structs to add functionality to a 3rd party defined structure. I've been following this thread, and in general, your criticism of MI is justified. Developers in general use MI to solve poor designs. However, having said that, I've seen some truly magnificent implementations of MI. MetroWerks PowerPlant is an example of this. Michael > -jcr > > -- > > What the Bard *should* have said: "All the World's a Stage, and > all the Men and Women rather poorly rehearsed."
From: ahoesch@on-luebeck.de (Andreas Hoeschler) Newsgroups: comp.sys.next.programmer Subject: ProjectBuilder - alphabetically sorted classses Date: 26 Sep 1998 19:47:14 GMT Organization: Offenes Netz Luebeck e.V. Message-ID: <6ujgc2$mf@merkur.smartsoft.de> Hi, I know that one can cause PB to show method names alphabetically. Is this also possible for class names? I've projects with 40 and more classes. They appear in the creation date ordering what makes it really hard to find a specific one. Any hints appreciated, Andreas
From: frank@REMOVE_SPAMLOCK.wizards.de (Frank M. Siegert) Newsgroups: comp.sys.next.programmer Subject: Re: 3.3 objects still of any interest? Date: 26 Sep 1998 10:30:42 GMT Organization: Frank's Area 51 Message-ID: <6uifoi$50k$2@news.seicom.net> References: <6uh3d2$67d$1@tallowcross.uni-frankfurt.de> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Cc: uli@tallowcross.uni-frankfurt.de In <6uh3d2$67d$1@tallowcross.uni-frankfurt.de> Uli Zappe wrote: > as a side effect of a project that I still had to do in 3.3, I have written a > browser delegate object that allows to easily read all kinds of data into an > NXBrowser object. > > The thing is very efficient, easy to use and well documented. It's built as a > palette that even allows a fully functional NXBrowser in IB test mode. BUT - > it's still 3.3. (And converting to OPENSTEP is non-trivial not only because > of NSBrowser's different API, but because the object makes heavy use of the > Storage class.) > > Is there still any interest for such a beast? Then I could put it on a > server. Definitely, please put it on the web or ftp! [I should point out that I really like to see more people doing this, I believe some really nice - but now outdated - NS stuff is flowing around on some backup tapes. People, if this is the case and you do not mind please put it on some public ftp archive. As a kind of going forth in good intentions full source for my CAPer.app and the full CAP server and afpmount sources for NS3.3 will appear shortly on this.net for everyone to fetch and recompile] -- * Frank M. Siegert [frank@wizards.de] - Home http://www.wizards.de * NeXTSTEP, IRIX, Solaris, Linux, BeOS, PDF & PostScript Wizard * Note: [frank@this.net] is still a valid option to send me eMail * "The answer is vi, what was your question...?"
From: cejensen@winternet.com (Christian Jensen) Newsgroups: comp.sys.next.programmer Subject: Re: 3.3 objects still of any interest? Date: 26 Sep 1998 18:10:58 GMT Organization: StarNet Communications, Inc. Message-ID: <6ujani$gr4$1@blackice.winternet.com> References: <6uh3d2$67d$1@tallowcross.uni-frankfurt.de> NNTP-Posting-Date: 26 Sep 1998 18:10:58 GMT You (Uli Zappe <uli@tallowcross.uni-frankfurt.de>) wrote in newsgroup comp.sys.next.programmer, on 25 Sep 1998 21:53:38 GMT: > Is there still any interest for such a beast? Then I could put it on a > server. Yes, please post it. There is interest! NEXTSTEP is not dead yet! There are people out there happily running operating systems a whole lot older and less "up to date" than NS. With the continuing support of a faithful user/developer base, NS can continue to be a useable OS for a long time to come. BTW, Anyone up for a Mozilla/NEXTSTEP project? ;-) --Chris ************************** Chris Jensen cejensen@winternet.com MIME, Sun, NeXTMail OK "Sacred cows make the best hamburger." --Mark Twain
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Mixing Objective-C and C++ Date: 28 Sep 1998 08:24:46 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn70uhug.beo.rog@talisker.ohm.york.ac.uk> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434 <1dfxhgw.1f8hacy1y1yvaoN@dialup113-2-7.swipnet.se> On Fri, 25 Sep 1998 21:09:47 +0200, Lars Farm <lars.farm@ite.mh.se> wrote: > Holger Hoffstaette <holger@object-factory.REMOVETHIS.com> wrote: > http://reality.sgi.com/employees/boehm_mti/gc.html > > Unfortunatelly the NeXT > derived frameworks plays some nasty tricks on pointers to allocated > memory and hides pointers from the collector. So it can't replace or > even be used with the YB refcounting scheme. This is a flaw in YB, not > the collector. there are things that a conforming ANSI C program can do that will fool any garbage collector without language support; the real problem is the C language is not truly garbage-collectable. rog.
From: giesen@informatik.uni-koblenz.de (Heinrich Giesen) Newsgroups: comp.sys.next.programmer Subject: Re: ProjectBuilder - alphabetically sorted classses Date: 28 Sep 1998 08:41:17 GMT Organization: University Koblenz / CC Message-ID: <6uni3d$al$1@newshost> References: <6ujgc2$mf@merkur.smartsoft.de> In article <6ujgc2$mf@merkur.smartsoft.de> ahoesch@on-luebeck.de (Andreas Hoeschler) writes: > Hi, > > I know that one can cause PB to show method names alphabetically. Is this > also possible for class names? I've projects with 40 and more classes. They > appear in the creation date ordering what makes it really hard to find a > specific one. > > Any hints appreciated, > > Andreas If you create a new class with 'File -> New in Project n' PB does not sort. If you insert an existing file with 'Project -> Add Files... A' PB does. Reorder the file names by dragging the browser cells (press the Control key, then click and drag with the mouse.) -- Heinrich Giesen | giesen@infko.uni-koblenz.de Universitaet Koblenz, | MIME / NeXTmail ok Institut fuer Informatik | Voice: +49 261 9119-416 Rheinau 1, D-56075 Koblenz, Germany | Fax: +49 261 9119-497
From: David Stes <stes@mundivia.es> Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Mixing Objective-C and C++ Date: Mon, 28 Sep 1998 14:57:27 +0000 Organization: Mundivia, Santander Message-ID: <360FA3D7.41C67EA6@mundivia.es> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434583@news.charm.net> <360DFD69.41C67EA6@mundivia.es> <360E9FB2.6D718960@sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit jik- wrote: > > > Karl Hsu wrote: > > > > > > There are only a few problems with Objective-C that I have right now: > > > 1. Lack of GC (this is a personal thing, especially since I'm having > > > problems with the whole reference counting thing right now). > Objective-C has no standard > THAT is the only problem I see with Objective-C. Read his email. He's not saying his having problems with no standard. He's saying he's having problems with the reference counting thing.
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: getting window frame under NT Date: 28 Sep 1998 15:14:22 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn70v9uj.e2o.rog@talisker.ohm.york.ac.uk> i need to line up one window with another (no overlap) and it seems that the [NSWindow frame] method does *not* give the area taken up by the window on the screen. i.e. the following code snippet does not work. // place top left of win2 at the top right of win1 void align(NSWindow *win1, NSWindow *win2) { NSRect fr; NSPoint topright; fr = [win1 frame]; topright.x = fr.origin.x + fr.size.width; topright.y = fr.origin.y + fr.size.height; [win2 setFrameTopLeftPoint:topright]; } it looks like the window frame is ignoring the title bar and the per-window menus. surely there must be a way around this!? cheers, rog. PS. this is under openstep for NT (actually WebObjects)
From: Tom Gugger <ehutch@norden1.com> Newsgroups: comp.sys.next.misc,comp.sys.next.programmer Subject: NEXTSTEP/ Contract-Long Term/Va Date: 28 Sep 1998 17:38:26 GMT Message-ID: <6uohii$sr4$0@192.153.35.30> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: "ehutch@norden1.com" <ehutch@norden1.com> Developer NEXTSTEP---------2yrs commercial experience Objective C----------2yrs experience Contract--------------long term Area------------------Virginia Start Date-------------Oct 1998 Must Have------------Good Communication Skills Must Be---------------US Citizen,US Greencard, or Canadian Citizen
From: Chris Penrose <penrose@cmlabfs.sfc.keio.ac.jp> Newsgroups: comp.sys.next.programmer Subject: where does tcgetpgrp() live? Date: 29 Sep 1998 02:13:22 +0900 Organization: Keio University Shonan Fujisawa Campus, Fujisawa Japan Message-ID: <wzk92oqhxp.fsf@cmlabfs.sfc.keio.ac.jp> Hi folks! I am trying to utilize the tcgetpgrp() function under Openstep 4.1. When I compile my small program, linking with the usually happy enough sys_s: cc -o forebye forebye.c -lsys_s /bin/ld: Undefined symbols: _tcgetpgrp tcgetpgrp() is MIA. I am sorry if I am too impatient to forage through all of the new painfully undersupported dynamic libraries to find where this function lives (being able to just use ar to look at a TOC was so much nicer -- the golden days). I hope instead for some kind intrepid spirit to share wisdom with me. I have included unistd.h in this program; this header file promises that tcgetpgrp() lives on the system. superego> pwd /NextLibrary/Frameworks/System.framework/Headers/bsd superego> grep tcgetpgrp * unistd.h: extern pid_t tcgetpgrp(int fildes); unistd.h: extern pid_t tcgetpgrp(); Is unistd.h lying to us? Has recent U.S. testing of nuclear weapons at the Nevada test site selectively bombarded my dynamic libraries of their process-id functionality or do I need to specify a library that I am ignorant of? Inquiring minds care to know. Chris Penrose penrose@sfc.keio.ac.jp
From: Tom Gugger <ehutch@norden1.com> Newsgroups: comp.sys.next.programmer,ab.jobs,calgary.jobs,hamilton.jobs Subject: [Fwd: NEXTSTEP/ Contract-Long Term/Va] Date: 28 Sep 1998 17:48:16 GMT Message-ID: <6uoi50$sr4$5@192.153.35.30> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="------------619E14D0DC2BB7FDFB0DDC0D" This is a multi-part message in MIME format. --------------619E14D0DC2BB7FDFB0DDC0D Content-Type: text/plain; charset=us-ascii Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Transfer-Encoding: 7bit --------------619E14D0DC2BB7FDFB0DDC0D Content-Type: message/rfc822 Content-Transfer-Encoding: 7bit Content-Disposition: inline Received: from ehutch.norden1.com (ehutch.norden1.com [192.153.35.30]) by norden1.com (8.9.1a/8.9.1) with ESMTP id NAA17222 for <ehutch@norden1.com>; Mon, 28 Sep 1998 13:50:52 -0400 (EDT) Message-ID: <360FC9AD.4AF1CD37@norden1.com> Date: Mon, 28 Sep 1998 13:38:54 -0400 From: Tom Gugger <ehutch@norden1.com> X-Mailer: Mozilla 4.0 [en] (Win95; I) MIME-Version: 1.0 Newsgroups: comp.sys.next.misc,comp.sys.next.programmer To: "ehutch@norden1.com" <ehutch@norden1.com> Subject: NEXTSTEP/ Contract-Long Term/Va X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Developer NEXTSTEP---------2yrs commercial experience Objective C----------2yrs experience Contract--------------long term Area------------------Virginia Start Date-------------Oct 1998 Must Have------------Good Communication Skills Must Be---------------US Citizen,US Greencard, or Canadian Citizen --------------619E14D0DC2BB7FDFB0DDC0D--
From: Melissa O'Neill <NoOnSePiAlMl@cs.sfu.ca> Newsgroups: comp.sys.next.programmer Subject: GNU Source for 3.3 and 4.2 developer tools? Date: 27 Sep 1998 23:31:11 GMT Organization: School of Computing Science, Simon Fraser University, BC, Canada Message-ID: <6umhrv$a2v$1@morgoth.sfu.ca> Originator: oneill@cs.sfu.ca (Melissa O'Neill) Perhaps I've been looking in the wrong place, but I don't see the GNU Source for the 3.3 and 4.2 developer tools on any of the NeXT archive sites -- the best I've found is 3.2. As I understand it, all someone has to do is copy a file off the CD-ROM (maybe tar it up, if it's still called GNUSource.pkg) and put it on an archive site. *PLEASE* would someone do this. Melissa. --- To send me e-mail, remove `N O S P A M' from my address.
From: rdieter@math.unl.edu (Rex Dieter) Newsgroups: comp.sys.next.programmer Subject: Re: where does tcgetpgrp() live? Date: 28 Sep 1998 18:34:04 GMT Organization: University of Nebraska-Lincoln Message-ID: <6uokqs$p7r$1@unlnews.unl.edu> References: <wzk92oqhxp.fsf@cmlabfs.sfc.keio.ac.jp> In article <wzk92oqhxp.fsf@cmlabfs.sfc.keio.ac.jp> Chris Penrose <penrose@cmlabfs.sfc.keio.ac.jp> writes: > I am trying to utilize the tcgetpgrp() function under Openstep 4.1. > cc -o forebye forebye.c -lsys_s > /bin/ld: Undefined symbols: > _tcgetpgrp > tcgetpgrp() is MIA. It's one of those POSIX only functions. In short, this function is no longer available in Openstep. -- Rex A. Dieter rdieter@math.unl.edu (NeXT/MIME OK) Computer System Manager http://www.math.unl.edu/~rdieter/ Mathematics and Statistics University of Nebraska-Lincoln
From: heiko.panther@mni.fh-giessen.de (Heiko Panther) Newsgroups: comp.sys.next.programmer Subject: Installing Rhapsody DR2 on a PM4400? Date: Mon, 28 Sep 1998 21:20:32 +0200 Organization: University of Giessen, Germany Message-ID: <heiko.panther-2809982120320001@asc-p07.mni.fh-giessen.de> Hi, is there any way I can get the installer to do this? (It just complains that DR2 won't run on this machine) Will it work on that machine if I get it to install? Thank you, Heiko
From: Chris Penrose <penrose@cmlabfs.sfc.keio.ac.jp> Newsgroups: comp.sys.next.programmer Subject: Re: Accessing sound Date: 29 Sep 1998 01:57:08 +0900 Organization: Keio University Shonan Fujisawa Campus, Fujisawa Japan Message-ID: <wzlnn4qior.fsf@cmlabfs.sfc.keio.ac.jp> References: <6udonp$p3u$1@news.ox.ac.uk> <906652508.629082@watserv4.uwaterloo.ca> dfevans@bbcr.uwaterloo.ca (David Evans) >Paolo Zuliani <scro0249@sable.ox.ac.uk> wrote: >>Is it (NeXTSTEP audio) full duplex? > NeXTSTEP can do full-duplex sound, hardware and drivers permitting. I don't mean to insult David, but there are no known drivers that support full duplex audio for NeXTstep/Openstep, intel or motorola. We are half-duplex until drivers appear. Also, the Openstep software architecture precludes support for 24-bit audio. So even if you are willing and able to write a 24-bit happy driver, NeXTstep/Openstep/RhapsodyDR2 won't let you do it because of NeXT's (and now Apple's) lack of foresight and concern for things sonic. My Adb card is both full-duplex and 24-bit but I can't yet (if ever) access its potentials under Openstep. The 24-bit issue is something that I could fix in an hour if Apple would release the audio driver. But alas, there are more important things to whine about: like how the United States just hypocritically recently resumed nuclear testing at the Nevada test site, despite imposing economic sanctions upon India and Pakistan for doing the same, and how the U.S. press has neglected to cover the event and its fallout despite significant worldwide protests. And you thought the U.S. had a free and uncensored news media? Chris Penrose penrose@sfc.keio.ac.jp
From: "Justin Morgan" <jmorgan@objectronics.com> Newsgroups: comp.sys.next.programmer References: <6udonp$p3u$1@news.ox.ac.uk><906652508.629082@watserv4.uwaterloo.ca> <wzlnn4qior.fsf@cmlabfs.sfc.keio.ac.jp> Subject: Re: Accessing sound Date: Mon, 28 Sep 1998 22:45:10 -0700 Organization: Objectronics Consulting Message-ID: <3610733f.0@blushng.jps.net> Chris Penrose <penrose@sfc.keio.ac.jp> held his breath as long as he could, pondered world events, and then frantically posted the following to c.s.n.p while turning a deep shade of purple... >[snip] >But alas, there are more important things to whine about: like how the >United States just hypocritically recently resumed nuclear testing at >the Nevada test site, despite imposing economic sanctions upon India >and Pakistan for doing the same, and how the U.S. press has neglected >to cover the event and its fallout despite significant worldwide >protests. And you thought the U.S. had a free and uncensored news >media? >[snip] [*LOL*] Actually, we've been testing underground multi-zillion megaton warheads in secret every single minute for years and years. Surprised it took this long for someone to notice. You imply the US press is censored. You bet it's censored! In fact, we're so sensor-happy here that we've somehow managed to sensor foreign sources, such as ITN world news (from England, http://www.itn.co.uk/). In fact, we've censored all the other world news agencies, so they can't report on our secret nuclear tests either. There is NOTHING available on those world news sites about our super-secret tests, is there? And that's why you couldn't give any reliable, verifiable sources, right? But you DO have them, right? (Better not give them away, or those sources will get censored too!!) Naturally, all those multi-zillion megaton warheads that we've been secretly testing would have sent every seismic sensor on the planet right off the scale. You are, naturally, an expert in geology and seismology and are aware that these things can't be easily hidden. But of course, that's it!...All of the seismic sensors in the world are censored by the American government! That story on CNN about how our government spent millions of dollars developing nuclear warhead simulation software--it's all a cover-up! Just so we could go and defy everybody else in the world, in secret. And don't you believe this nuclear detonation story is just a conspiracy theory! Conspiracy theories are a plot by the intelligencia to keep you from learning the truth! Of course, you realize...now that you've let the secret out, the CIA is sending somebody "for you". Watch your back, buddy! [*ROTFL*] Justin Morgan jmorgan@objectronics.com
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Mixing Objective-C and C++ Date: 28 Sep 98 11:16:13 Organization: Is a sign of weakness Message-ID: <SCOTT.98Sep28111613@slave.doubleu.com> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434583@news.charm.net> <bwebster-2509982219450001@swb3-asyn-27.rice.edu> In-reply-to: bwebster@rice.edu's message of Fri, 25 Sep 1998 22:19:44 -0500 In article <bwebster-2509982219450001@swb3-asyn-27.rice.edu>, bwebster@rice.edu (Brian Webster) writes: Well, making multiple protocols and having one class conform to them would work in the that the class would then be able to respond to messages in either protocol, but since protocols are just a list of functions, there is no code to be inherited, which is one of the main reasons that MI is used at all, to reuse code. The first use, interface inheritence, is what people generally intend to use MI for. The second use, implementation inheritence, is where people who use MI inappropriately tend to get into trouble. Implementation inheritence is a subtle siren call. It sounds like it will save you a lot of work - but the unfortunate fact is, an "Employee" is _not_ a "SubscriptionService", even though both may require periodic payments, and it may seem convenient to inherit the periodic payment implementation from "SubscriptionService". Subclassing from the wrong class is bad, bad, bad - mixing in implementation details from the wrong class using MI is much worse. Implementation inheritence is a perfectly good reason to use MI. So long as your implementations are well focussed on accomplishing very specific things. Most implementation are not that clean, and thus many times MI would be better left alone. Just like use of "goto", many people have seen the problem enough times, and have found through experience that almost invariably there is an alternative which works just as well but doesn't use the offending construct. Is forbidding use of MI or "goto" the right way to go? Well, in nearly a decade I've used "goto" exactly once in work I was being paid for, and that time it made an otherwise unwieldy routine significantly easier (though not "easy") to understand. I expect I could _justify_ MI more often than that, but how often will use of MI really make things clearer? Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: "Michelle L. Buck" <buck.erik@mcleod.net> Newsgroups: comp.sys.next.programmer Subject: Re: ProjectBuilder - alphabetically sorted classses Date: Mon, 28 Sep 1998 21:00:10 -0500 Organization: McleodUSA - http://www.mcleodusa.net Message-ID: <6upf2k$mos1@odie.mcleod.net> References: <6ujgc2$mf@merkur.smartsoft.de> You can Control Drag the classes to order them in any way you want. This is sometimes needed to control build order. Andreas Hoeschler wrote in message <6ujgc2$mf@merkur.smartsoft.de>... >Hi, > >I know that one can cause PB to show method names alphabetically. Is this >also possible for class names? I've projects with 40 and more classes. They >appear in the creation date ordering what makes it really hard to find a >specific one. > >Any hints appreciated, > >Andreas
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: ProjectBuilder - alphabetically sorted classses Date: 28 Sep 98 11:19:01 Organization: Is a sign of weakness Message-ID: <SCOTT.98Sep28111901@slave.doubleu.com> References: <6ujgc2$mf@merkur.smartsoft.de> <6uni3d$al$1@newshost> In-reply-to: giesen@informatik.uni-koblenz.de's message of 28 Sep 1998 08:41:17 GMT In article <6uni3d$al$1@newshost>, giesen@informatik.uni-koblenz.de (Heinrich Giesen) writes: In article <6ujgc2$mf@merkur.smartsoft.de>, ahoesch@on-luebeck.de (Andreas Hoeschler) writes: > I know that one can cause PB to show method names alphabetically. > Is this also possible for class names? I've projects with 40 and > more classes. They appear in the creation date ordering what > makes it really hard to find a specific one. If you create a new class with 'File -> New in Project n' PB does not sort. If you insert an existing file with 'Project -> Add Files... A' PB does. Reorder the file names by dragging the browser cells (press the Control key, then click and drag with the mouse.) Ick! Just remove all the files from the project area in question (don't delete when the panel asks!), then use Add Files... to put them all back in in order. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: getting window frame under NT Date: 28 Sep 98 11:21:18 Organization: Is a sign of weakness Message-ID: <SCOTT.98Sep28112118@slave.doubleu.com> References: <slrn70v9uj.e2o.rog@talisker.ohm.york.ac.uk> In-reply-to: rog@ohm.york.ac.uk's message of 28 Sep 1998 15:14:22 GMT In article <slrn70v9uj.e2o.rog@talisker.ohm.york.ac.uk>, rog@ohm.york.ac.uk (Roger Peppe) writes: i need to line up one window with another (no overlap) and it seems that the [NSWindow frame] method does *not* give the area taken up by the window on the screen. <...> it looks like the window frame is ignoring the title bar and the per-window menus. This is documented operation (though I don't recall where :-). Under NT, OpenStep doesn't own the titlebar or per-window menus or resize strip, so OpenStep doesn't report about them. I think if you want to find out such information, you have to Descend To The Windows API Level... and there I am not willing to go... -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: getting window frame under NT Date: 29 Sep 1998 10:39:43 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn711e7i.g06.rog@talisker.ohm.york.ac.uk> References: <slrn70v9uj.e2o.rog@talisker.ohm.york.ac.uk> <SCOTT.98Sep28112118@slave.doubleu.com> On 28 Sep 98 11:21:18, Scott Hess <scott@nospam.doubleu.com> wrote: > In article <slrn70v9uj.e2o.rog@talisker.ohm.york.ac.uk>, > it looks like the window frame is ignoring the title bar and the > per-window menus. > > This is documented operation (though I don't recall where :-). Under > NT, OpenStep doesn't own the titlebar or per-window menus or resize > strip, so OpenStep doesn't report about them. oh. perhaps i was being naive in believing the following (from NSWindow docs): : An NSWindow is defined by a _frame rectangle_ that encloses the entire : window, including its title bar, border, and other peripheral elements > I think if you want to find out such information, you have to Descend > To The Windows API Level... and there I am not willing to go... me neither. the only problem is that it's now impossible to do decent window placement management under NT. this buggers up my "designed in from the beginning" window management a treat. back to square one. cheers, rog. PS if anyone *does* know of a hack for obtaining the correct window frame under NT, i'd love to hear of it. if it's not possible, how does the "cascadeTopLeftFromPoint:" method manage to work properly? (maybe it doesn't, i haven't tried)
From: dossr@ecs.ecs.csus.edu (Robert C. Doss Jr.) Newsgroups: comp.sys.next.programmer Subject: OPENSTEP Enterprise on Windows 98? Date: 29 Sep 1998 16:04:27 GMT Organization: California State University, Sacramento Message-ID: <6ur0eb$t19$1@csusac.ecs.csus.edu> I was running OPENSTEP Enterprise 4.2 just fine on a Windows 95 box, but when I put it on a box with Windows 95 relase B, it no longer worked. It looks like I need to start with a fresh copy of Windows and I figure if I had to do that, I would rather use Windows 98 instead of downgrading to the original Windows 95 release. I was wondering if anyone out there is running OPENSTEP Enterprise 4.2 on Windows 98? I would like to know if it is possible before I take the plunge. :-) -- Robert C. Doss, Jr. California State University, Sacramento <dossr@csus.edu> <http://gaia.ecs.csus.edu/~dossr> Facsimile: (707) 253-3063
From: dfevans@bbcr.uwaterloo.ca (David Evans) Newsgroups: comp.sys.next.programmer Subject: Re: Accessing sound Date: 29 Sep 1998 15:45:10 GMT Organization: University of Waterloo Message-ID: <907083910.141945@watserv4.uwaterloo.ca> References: <6udonp$p3u$1@news.ox.ac.uk> <906652508.629082@watserv4.uwaterloo.ca> <wzlnn4qior.fsf@cmlabfs.sfc.keio.ac.jp> Cache-Post-Path: watserv4.uwaterloo.ca!unknown@bcr11.uwaterloo.ca In article <wzlnn4qior.fsf@cmlabfs.sfc.keio.ac.jp>, Chris Penrose <penrose@cmlabfs.sfc.keio.ac.jp> wrote: > >dfevans@bbcr.uwaterloo.ca (David Evans) >>Paolo Zuliani <scro0249@sable.ox.ac.uk> wrote: >>>Is it (NeXTSTEP audio) full duplex? > >> NeXTSTEP can do full-duplex sound, hardware and drivers permitting. > >I don't mean to insult David, but there are no known drivers that >support full duplex audio for NeXTstep/Openstep, intel or motorola. Are you sure about that? I recall doing some full-duplex work on my cube back in 1994 but perhaps the memory has simply warped with time. >We are half-duplex until drivers appear. So I suppose that my statement does actually stand, although it's not very useful. ;-) -- David Evans (NeXTMail/MIME OK) dfevans@bbcr.uwaterloo.ca Computer/Synth Junkie http://bbcr.uwaterloo.ca/~dfevans/ University of Waterloo "Default is the value selected by the composer Ontario, Canada overridden by your command." - Roland TR-707 Manual
From: toon@omnigroup.com (Greg Titus) Newsgroups: comp.sys.next.programmer Subject: Re: Plutonium Tests (was Re: Accessing sound) Date: 29 Sep 1998 17:39:14 GMT Organization: Omni Development, Inc. Message-ID: <6ur602$2tk$1@gaea.omnigroup.com> References: <3610733f.0@blushng.jps.net> "Justin Morgan" <jmorgan@objectronics.com> writes > Chris Penrose <penrose@sfc.keio.ac.jp> held his breath as long as he > could, pondered world events, and then frantically posted the > following to c.s.n.p while turning a deep shade of purple... > >[snip about nuclear testing] > > [*LOL*] > > Actually, we've been testing underground multi-zillion megaton > warheads in secret every single minute for years and years. > Surprised it took this long for someone to notice. > > You imply the US press is censored. You bet it's censored! In fact, > we're so sensor-happy here that we've somehow managed to sensor > foreign sources, such as ITN world news (from England, > http://www.itn.co.uk/). In fact, we've censored all the other world > news agencies, so they can't report on our secret nuclear tests > either. There is NOTHING available on those world > news sites about our super-secret tests, is there? And that's why > you couldn't give any reliable, verifiable sources, right? But you DO > have them, right? How about the Las Vegas Sun: A year old article that describes the type of tests in detail: http://www.lasvegassun.com/sunbin/stories/archives/1997/may/30/50594727 5.html A mention of the latest test on Saturday and the protests in Japan: http://www.lasvegassun.com/sunbin/stories/archives/1998/sep/27/09270054 1.html Technically these are subcritical plutonium explosive tests, and so it might have been a bit misleading for Chris to have said "nuclear test", but they certainly do exist, and there was surprisingly little mention in the U.S. media even though it is a very grey area as far as the test ban treaty goes. Even though there are a lot of wackos out there who post paranoid B.S. that sounds similar to Chris' comment, you should have given him the benefit of the doubt in this case. --Greg ---------------------- Greg Titus Omni Development Inc. greg@omnigroup.com
From: toon@omnigroup.com (Greg Titus) Newsgroups: comp.sys.next.programmer Subject: Re: Plutonium Tests (was Re: Accessing sound) Date: 29 Sep 1998 17:39:07 GMT Organization: Omni Development, Inc. Message-ID: <6ur5vr$2ti$1@gaea.omnigroup.com> References: <3610733f.0@blushng.jps.net> "Justin Morgan" <jmorgan@objectronics.com> writes > Chris Penrose <penrose@sfc.keio.ac.jp> held his breath as long as he > could, pondered world events, and then frantically posted the > following to c.s.n.p while turning a deep shade of purple... > >[snip about nuclear testing] > > [*LOL*] > > Actually, we've been testing underground multi-zillion megaton > warheads in secret every single minute for years and years. > Surprised it took this long for someone to notice. > > You imply the US press is censored. You bet it's censored! In fact, > we're so sensor-happy here that we've somehow managed to sensor > foreign sources, such as ITN world news (from England, > http://www.itn.co.uk/). In fact, we've censored all the other world > news agencies, so they can't report on our secret nuclear tests > either. There is NOTHING available on those world > news sites about our super-secret tests, is there? And that's why > you couldn't give any reliable, verifiable sources, right? But you DO > have them, right? How about the Las Vegas Sun: A year old article that describes the type of tests in detail: http://www.lasvegassun.com/sunbin/stories/archives/1997/may/30/50594727 5.html A mention of the latest test on Saturday and the protests in Japan: http://www.lasvegassun.com/sunbin/stories/archives/1998/sep/27/09270054 1.html Technically these are subcritical plutonium explosive tests, and so it might have been a bit misleading for Chris to have said "nuclear test", but they certainly do exist, and there was surprisingly little mention in the U.S. media even though it is a very grey area as far as the test ban treaty goes. Even though there are a lot of wackos out there who post paranoid B.S. that sounds similar to Chris' comment, you should have given him the benefit of the doubt in this case. --Greg ---------------------- Greg Titus Omni Development Inc. greg@omnigroup.com
From: Tom Gugger <ehutch@norden1.com> Newsgroups: comp.sys.next.programmer,us.jobs.offered,us.jobs.contract Subject: [Fwd: NEXTSTEP/ Contract-Long Term/Va] Date: 29 Sep 1998 18:24:25 GMT Message-ID: <6ur8kp$8hg$2@192.153.35.30> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="------------92AD62F54FFF969079468A22" This is a multi-part message in MIME format. --------------92AD62F54FFF969079468A22 Content-Type: text/plain; charset=us-ascii Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Transfer-Encoding: 7bit --------------92AD62F54FFF969079468A22 Content-Type: message/rfc822 Content-Transfer-Encoding: 7bit Content-Disposition: inline Received: from ehutch.norden1.com (ehutch.norden1.com [192.153.35.30]) by norden1.com (8.9.1a/8.9.1) with ESMTP id NAA17222 for <ehutch@norden1.com>; Mon, 28 Sep 1998 13:50:52 -0400 (EDT) Message-ID: <360FC9AD.4AF1CD37@norden1.com> Date: Mon, 28 Sep 1998 13:38:54 -0400 From: Tom Gugger <ehutch@norden1.com> X-Mailer: Mozilla 4.0 [en] (Win95; I) MIME-Version: 1.0 Newsgroups: comp.sys.next.misc,comp.sys.next.programmer To: "ehutch@norden1.com" <ehutch@norden1.com> Subject: NEXTSTEP/ Contract-Long Term/Va X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Developer NEXTSTEP---------2yrs commercial experience Objective C----------2yrs experience Contract--------------long term Area------------------Virginia Start Date-------------Oct 1998 Must Have------------Good Communication Skills Must Be---------------US Citizen,US Greencard, or Canadian Citizen --------------92AD62F54FFF969079468A22--
Newsgroups: comp.sys.next.programmer From: maury@remove_this.istar.ca (Maury Markowitz) Subject: Re: Plutonium Tests (was Re: Accessing sound) Content-Type: text/plain; charset=us-ascii Message-ID: <F028xr.I7q@T-FCN.Net> Sender: news@T-FCN.Net Content-Transfer-Encoding: 7bit Cc: toon@omnigroup.com Organization: none References: <3610733f.0@blushng.jps.net> <6ur5vr$2ti$1@gaea.omnigroup.com> Mime-Version: 1.0 Date: Tue, 29 Sep 1998 19:32:14 GMT In <6ur5vr$2ti$1@gaea.omnigroup.com> Greg Titus wrote: > How about the Las Vegas Sun: > > A year old article that describes the type of tests in detail: > http://www.lasvegassun.com/sunbin/stories/archives/1997/may/30/50594727 > 5.html > > A mention of the latest test on Saturday and the protests in Japan: > http://www.lasvegassun.com/sunbin/stories/archives/1998/sep/27/09270054 > 1.html Neither link works for me. Maybe they were censored? :-) > Technically these are subcritical plutonium explosive test Do you know the purpose of the tests? Maury
From: scro0249@sable.ox.ac.uk (Paolo Zuliani) Newsgroups: comp.sys.next.programmer Subject: Re: Plutonium Tests (was Re: Accessing sound) Date: 29 Sep 1998 20:47:29 GMT Organization: Oxford University, England Message-ID: <6urh11$bl5$1@news.ox.ac.uk> References: <3610733f.0@blushng.jps.net> <6ur5vr$2ti$1@gaea.omnigroup.com> NNTP-Posting-Date: 29 Sep 1998 20:47:29 GMT : Technically these are subcritical plutonium explosive tests, and so it : might have been a bit misleading for Chris to have said "nuclear test", : but they certainly do exist, and there was surprisingly little mention : in the U.S. media even though it is a very grey area as far as the test : ban treaty goes. Even the International Herald Tribune has said almost nothing about the tests. I remember a few months ago when India and Pakistan did their tests: the IHT came out with the yellow banner in the first page (their sign for a big news) and a lot of articles inside. Now they have put just one short article in 2nd page... Regards, Paolo PS Thanks for the articles of the Las Vegas Sun.
From: scro0249@sable.ox.ac.uk (Paolo Zuliani) Newsgroups: comp.sys.next.programmer Subject: Re: Plutonium Tests (was Re: Accessing sound) Date: 29 Sep 1998 20:49:03 GMT Organization: Oxford University, England Message-ID: <6urh3v$bl5$2@news.ox.ac.uk> References: <3610733f.0@blushng.jps.net> <6ur5vr$2ti$1@gaea.omnigroup.com> <F028xr.I7q@T-FCN.Net> NNTP-Posting-Date: 29 Sep 1998 20:49:03 GMT Maury Markowitz (maury@remove_this.istar.ca) wrote: : In <6ur5vr$2ti$1@gaea.omnigroup.com> Greg Titus wrote: : > : > A year old article that describes the type of tests in detail: : > http://www.lasvegassun.com/sunbin/stories/archives/1997/may/30/50594727 : > 5.html : > : > A mention of the latest test on Saturday and the protests in Japan: : > http://www.lasvegassun.com/sunbin/stories/archives/1998/sep/27/09270054 : > 1.html : Neither link works for me. Maybe they were censored? :-) Worked ok for me. Regards, Paolo
Newsgroups: comp.sys.next.programmer From: maury@remove_this.istar.ca (Maury Markowitz) Subject: Re: Plutonium Tests (was Re: Accessing sound) Content-Type: text/plain; charset=us-ascii Message-ID: <F02H3I.MpC@T-FCN.Net> Sender: news@T-FCN.Net Content-Transfer-Encoding: 7bit Cc: scro0249@sable.ox.ac.uk Organization: none References: <3610733f.0@blushng.jps.net> <6ur5vr$2ti$1@gaea.omnigroup.com> <F028xr.I7q@T-FCN.Net> <6urh3v$bl5$2@news.ox.ac.uk> Mime-Version: 1.0 Date: Tue, 29 Sep 1998 22:28:27 GMT In <6urh3v$bl5$2@news.ox.ac.uk> Paolo Zuliani wrote: > : > A year old article that describes the type of tests in detail: > : > http://www.lasvegassun.com/sunbin/stories/archives/1997/may/30/50594727 > : > 5.html > : > > : > A mention of the latest test on Saturday and the protests in Japan: > : > http://www.lasvegassun.com/sunbin/stories/archives/1998/sep/27/09270054 > : > 1.html > > : Neither link works for me. Maybe they were censored? :-) > > Worked ok for me. It was the line wrapping actually. OmniWeb ate the chars. Maury
From: "Michelle L. Buck" <buck.erik@mcleod.net> Newsgroups: comp.sys.next.programmer Subject: Re: getting window frame under NT Date: Tue, 29 Sep 1998 21:09:00 -0500 Organization: McleodUSA - http://www.mcleodusa.net Message-ID: <6us40h$13c63@odie.mcleod.net> References: <slrn70v9uj.e2o.rog@talisker.ohm.york.ac.uk> <SCOTT.98Sep28112118@slave.doubleu.com> <slrn711e7i.g06.rog@talisker.ohm.york.ac.uk> Roger Peppe wrote in message ... >oh. >perhaps i was being naive in believing the following (from NSWindow docs): > >: An NSWindow is defined by a _frame rectangle_ that encloses the entire >: window, including its title bar, border, and other peripheral elements > >> I think if you want to find out such information, you have to Descend >> To The Windows API Level... and there I am not willing to go... > >me neither. >the only problem is that it's now impossible to do decent window placement >management under NT. this buggers up my "designed in from the beginning" >window management a treat. > >back to square one. > > cheers, > rog. > >PS if anyone *does* know of a hack for obtaining the correct window >frame under NT, i'd love to hear of it. if it's not possible, how does >the "cascadeTopLeftFromPoint:" method manage to work properly? (maybe >it doesn't, i haven't tried) > -cascadeTopLeftFromPoint: can work entirely with relative coordinates. As long as the title bar and menus take the same amount of vertical space in each window, their actual dimensions can be ignored. For the specific solution to your problem, it is possible to get the hWnd parameter that Windows uses to reference the window. Their is a simple Win32 API to get the window frame from the hWnd. You will need one of the rare and ugly ifdef _WIN32 macros.
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: Plutonium Tests (was Re: Accessing sound) Date: 30 Sep 1998 14:06:30 GMT Organization: Idiom Communications Message-ID: <6utdt6$cok$2@news.idiom.com> References: <3610733f.0@blushng.jps.net> <6ur5vr$2ti$1@gaea.omnigroup.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: toon@omnigroup.com How about taking this to a talk.politics group? It's *way* off-topic here. -jcr -- What the Bard *should* have said: "All the World's a Stage, and all the Men and Women rather poorly rehearsed."
From: "Charles W. Swiger" <chuck@codefab.com> Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: automatic GC in ObjC (was: Mixing Objective-C and C++) Date: 1 Oct 1998 18:28:04 GMT Organization: Spacelab.net Internet Access Message-ID: <6v0hjk$rt5$1@news.spacelab.net> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434583@news.charm.net> <6ufkdp$9l6$1@leonie.object-factory.com> <360f8387.346399652@news.charm.net> <6ugarg$nfo$1@leonie.object-factory.com> <360C341F.E71303FB@mdlink.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Helge Hess <hh@mdlink.de> wrote: > I just want to clear up with the idea that GC doesn't integrate with > ObjC because ObjC is based on C. This is plain wrong, it does integrate > in a very nice way, which was shown with the GC aware implementation of > Ovidiu Predescu and me. Like these debates always seem to, people tend to talk past each others' points. In particular, I am not aware of any garbage collection scheme which can always correctly determine whether memory regions are in use if the language permits arbitrary pointer arithmetic, which C and Obj-C do. Consider a program which internally implements a language (one example out of many is Emacs with its embedded ELISP interpreter) with tagged pointers, for example. [ ... ] >> Right! - when you were finished with them. Most programmers (especially >> OO newbies dabbling in Java) just don't know when they are finished with >> an object. You (or, more generally speaking, anybody who understands >> about the 'GC makes everything easy' pitfall) seem to be an exception. >.. > A GC used in a wrong way may make a program less efficient but it nearly > never makes a program buggy. I'll agree with this, assuming the collector itself isn't buggy and assuming the [often deliberate] limitations of the language permit strong GC, such as Java. > Every manual memory management, RC as well, is very difficult to track if > you have complex memory structures. Somewhat. It helps a great deal to have a well-understood convention for your reference counting so that everyone knows what to do. > Especially in conjunction with autorelease pools you really quick get > into some of the most nasty bugs in development. RC as in the > FoundationKit makes memory management 'look' easier, which IMHO is the > reason why newbies tend to have problems with it. Foundation's memory management makes memory management much easier than explicit malloc()/free(). It is a middle ground between the convenience and expense of full GC and the inconvenience and (potentially) higher efficiency of explicit memory management. > Don't get me wrong, RC in the Foundation Kit is much better than the > simple alloc/free scheme before, but real GC is _far_ better. If > complex/dynamic structures are not a good enough example for you, think > about multi-threading. It's almost impossible to use ref-counting in a > efficient and bugfree way here, which is why the Foundation Docu > recommends to use DO for inter-thread communication. That's an orthogonal issue. DO is recommended for inter-thread communication more in order to take advantage of the properties of Mach messaging and less in order to avoid these supposed problems with MT reference counting. The Foundation's GC works just fine in a multithreaded enviroment, unless you have some major performance penalties you can demonstrate for us. [ ... ] > Ever heard about conservative GC's ? C is of course a good thing, but > conservative GC's support C and C++. Sure. They require one to restrict the ways in which one does pointer arithmetic so that the GC doesn't mistakenly free a region of memory that is still in use. Most of the time, that restriction is no problem...but not always. -Chuck Charles Swiger | chuck@codefab.com | Yeah, yeah-- disclaim away. ----------------+-------------------+---------------------------- You have come to the end of your journey. Survival is everything.
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: automatic GC in ObjC (was: Mixing Objective-C and C++) Date: 5 Oct 1998 12:07:14 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn71hdjl.5o2.rog@talisker.ohm.york.ac.uk> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434 <1dge0ka.8jmb84v8jlocN@dialup120-2-14.swipnet.se> On Sun, 4 Oct 1998 22:07:22 +0200, Lars Farm <lars.farm@ite.mh.se> wrote: > Roger Peppe <rog@ohm.york.ac.uk> wrote: > > BTW, how do conservative GCs work when you've got large quantities of > > non-pointer data around - e.g. large bitmaps? does it have to scan all > > that data? i would imagine that would incurr a fairly large performance > > overhead, particularly when you're using 20MB images. (my current app, > > for example) > > As always, it depends :-) Since there is no standard for these things > I'll just use Boehms collector as an example. Use GC_malloc_atomic() > which returns a block that you have promised will NOT carry any pointer > data. That block will not be scanned and will do fine for image data, > text and such. that technique wouldn't do much good when using third-party code which does its own allocation; the AppKit, for example :-) still, it's quite amazing that C can be made garbage-collectable at all, so i shouldn't gripe! cheers, rog.
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: stripping yellowbox .EXE files? Date: 5 Oct 1998 12:19:36 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn71hear.5o2.rog@talisker.ohm.york.ac.uk> is it possible to strip the symbolic information from executables produced under the YellowBox? it would be great if possible, because currently we've got the following file sizes: Configuration Size (bytes) stripped (openstep for Intel) 523872 unstripped (openstep for Intel) 3957460 .exe (yellowbox) 9090560 so currently we have to ship an NT executable that's more than 17 times as big as the openstep one! cheers, rog.
From: Valentino Kyriakides <vkyr@lavielle.com> Newsgroups: comp.sys.next.programmer Subject: Re: WP Generating the wrong code Date: Mon, 05 Oct 1998 14:46:51 +0200 Organization: Lavielle Message-ID: <3618BFBA.95755D46@lavielle.com> References: <6v3k31$cmf@obi-wan.fdt.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit TjL wrote: > NOTE: FOLLOWUPS TO COMP.SYS.NEXT.PROGRAMMER > > Something that has always bothered me about WordPerfect for NeXT that I just > figured out today. > > WP has a feature called "Flush Right" which allows you to put some text flush > against the right margin, ie: > > John Q. Smith 2 October 1998 > > where the "1998" is right against the right margin. > > Unfortunately this never worked in WP for NeXT. Today I discovered why > (although it wasn't difficult to figure out). The menu item labelled "Flush > Right" actually generates the code for "Right Justify" instead, so the entire > line is right justified: > > John Q. Smith 2 October 1998 > > I was wondering if there was some neat NeXTStep hackish trick that could > change the code associated with the menu item? > > Did you tried to open the corresponding nib-file (the main one which contains the menu) in NS3.3 IB and figure out the connection settings? - Maybe there is a connectable method "flushRight" and you can reconnect the menu item "Flush Right" to this method? That's the only idea I have so far related to your problem. Greetings Valentino -- Valentino Kyriakides Lavielle EDV Systemberatung GmbH & Co. Tel.: +49(0)40 / 65 80 8 - 997 Lotharstrasse 2b, D-22041 Hamburg, Germany Fax.: +49(0)40 / 65 808-202 http://www.lavielle.com/ mailto: vkyr@lavielle.com
From: holger@object-factory.REMOVETHIS.com (Holger Hoffstaette) Newsgroups: comp.sys.next.programmer Subject: Re: Compile time performance: OS/Mach vs OE/NT Date: 5 Oct 1998 13:32:10 GMT Organization: the secret circle of the NSRC Message-ID: <6vahoq$59$1@leonie.object-factory.com> References: <6v9rmh$deu$1@oxygen.technet.net> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Constantin Szallies wrote: > Here is my result from a performance test I did in your firm. I posted this > because it might be interesting for other people as well and maybe I get a > useful hint or surgestion 8-) Yeah, building with OS-E/NT is a real yawner. The biggest hit is caused by the compiler's inability to use precompiled headers, so if you're recompiling many small files which #import EOInterface.h instead of only those classes that they really need, each little file has to slurp in all of AppKit and EOF. Some tips: - make use of the @class directive - avoid AppKit/EOInterface imports when possible (I've seen a _large_ EO framework with dozens of EO classes, and each of them #imported EOInterface.h; turning this into EOControl.h cut the build time in half!) - avoid having both source and object files on the same disk: keep the source on the net, but build into a local directory - when you have #imports on the network (e.g. custom frameworks), put $(NEXT_ROOT)/NextLibrary/Frameworks to the TOP of the framework search path; otherwise, each AppKit/Foundation/EO #import will hit the net *first*, resulting in a lot of unnecessary network traffic - try to use the install_debug target: this should skip the totally braindead copy-everything-for-stripping-and-relinking, resulting in faster framework turnaround time I *think* YB/NT finally uses precompiled headers, but I'm not really sure (can anybody confirm this?) Holger
From: holger@object-factory.REMOVETHIS.com (Holger Hoffstaette) Newsgroups: comp.sys.next.programmer Subject: Re: stripping yellowbox .EXE files? Date: 5 Oct 1998 13:43:36 GMT Organization: the secret circle of the NSRC Message-ID: <6vaie8$59$2@leonie.object-factory.com> References: <slrn71hear.5o2.rog@talisker.ohm.york.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Roger Peppe wrote: > is it possible to strip the symbolic information from executables > produced under the YellowBox? Sure. Have you tried the 'install' build target? Works for me, and produces amazingly small executables. :-) Holger
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: stripping yellowbox .EXE files? Date: 5 Oct 98 08:07:48 Organization: Is a sign of weakness Message-ID: <SCOTT.98Oct5080748@slave.doubleu.com> References: <slrn71hear.5o2.rog@talisker.ohm.york.ac.uk> In-reply-to: rog@ohm.york.ac.uk's message of 5 Oct 1998 12:19:36 GMT In article <slrn71hear.5o2.rog@talisker.ohm.york.ac.uk>, rog@ohm.york.ac.uk (Roger Peppe) writes: is it possible to strip the symbolic information from executables produced under the YellowBox? Use the "install" target in ProjectBuilder, it'll leave the stripped version wherever you said to install it in the Build Attributes of Project Inspector. There is no strip(1) type command under NT. so currently we have to ship an NT executable that's more than 17 times as big as the openstep one! Ours come out about the same size... Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: nospam@nospam.com Newsgroups: comp.sys.next.programmer Subject: Re: Yellow box under Windows CE anyone ? Date: 5 Oct 1998 19:26:52 GMT Organization: Internet Specialties West, Inc. Message-ID: <6vb6hs$r5h$1@news.iswest.net> References: <6v87m3$28k$1@pump1.york.ac.uk> In-Reply-To: <6v87m3$28k$1@pump1.york.ac.uk> On 10/04/98, -bat. wrote: >Has anybody tried this ? I've not actualy seen CE myself, but I have to >deploy some code on it. Not necessarily an App, but I do need to use the >foundation kit at least - can I just drop in a YellowBox / OpenStep DLL >and it will work fine or am I being overly optimistic ? > >[and if so which DLL and how do I do it ?] I think CE typically runs on non-Intel processors (MIPS, for example). So YB won't work unless it's recompiled and released by Apple for the CE system you're interested in. (unlikely) PS. This is why we need a VM based language for a real multiplatform development solution. And I'm thinking of NewtonScript, not Java. Steve
From: Steve Watt <steve@watt.com> Organization: USENET spam abatement Sender: pywiagjh@bigfoot.com Date: 5 Oct 98 19:51:37 GMT Message-ID: <cancel.130943492386605824@bigfoot.com> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <130943492386605824@bigfoot.com> ignore Control: cancel <130943492386605824@bigfoot.com> I have cancelled this article which had a BI of more than 20. Selected original headers: }From: pywiagjh@bigfoot.com }Subject: Owning Your Own Adult Interent Business Is Easy }Path: ...!howland.erols.net!newsfeed.cwix.com!192.232.20.2!malgudi.oar.net!plonk.apk.net!news.apk.net!news.micro-net.net!not-for-mail }NNTP-Posting-Host: ip173.harvey.la.pub-ip.psi.net }Lines: 11
From: luomat@peak.org.this.all.must.be.removed (TjL) Newsgroups: comp.sys.next.programmer Subject: Re: WP Generating the wrong code Date: 5 Oct 1998 22:23:38 GMT Organization: Florida Digital Turnpike Message-ID: <6vbgta$bvs@obi-wan.fdt.net> References: <6v3k31$cmf@obi-wan.fdt.net> <6v4aln$80n$1@ash.prod.itd.earthlink.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: cdvorak@news.earthlink.net In <6v4aln$80n$1@ash.prod.itd.earthlink.net> cdvorak@news.earthlink.net wrote: > In typesetting terminology, Right Justify and Flush Right mean precisely > the same thing. (I'm a typesetter for 11+ years.) Your example of: > > John Q. Smith 2 October 1998 > > is actually a Center Justify meaning that two groups of words are pushed > out to the left and right margins. Well, I know nothing about typesetting terminology. All I was speaking of is WordPerfect's terminology, where they are two different things. Flush Right is "FLSH RT" and Right Justify is "RIGHT JUST". In WP's usage, they are two very different things. Flush Right will automatically position and re-position the date (using the above example) whereas Right Justify will collapse the tabs between the name and the date. "Flush Right" is a WP feature that has always worked (well, at least in WP5.1 for DOS which is what I used) differently than Right Justify. The benefit is not having to manually set the spacing, especially if I change the text (ie if I want to use "2 Oct 98" instead of "2 October 1998" it will still be flush against the right margin and the name will stay against the left side). I might be using the wrong typesetter terminology, because in the context of WP "Flush Right", "Right Justify" and "Center Justify" are all separate things. This was a programmer boo-boo on their part. All I want to do is see if there is way to work around it. TjL -- Spam-altered address in effect, remove obvious portion if replying by email.
From: luomat@peak.org.this.all.must.be.removed (TjL) Newsgroups: comp.sys.next.programmer Subject: Re: WP Generating the wrong code Date: 5 Oct 1998 22:31:28 GMT Organization: Florida Digital Turnpike Message-ID: <6vbhc0$c36@obi-wan.fdt.net> References: <6v3k31$cmf@obi-wan.fdt.net> <907465808.485937@watserv4.uwaterloo.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: dfevans@bbcr.uwaterloo.ca In <907465808.485937@watserv4.uwaterloo.ca> David Evans wrote: > you can always open (a copy of!) WP's main .nib file in InterfaceBuilder > and have at the object connections. I've used IB to set things like NXCommandKeys in apps where using the dwrite isn't enough.... but I don't know enough about it to do something like this. Do you have WP and if so do you know how I would do it? I found that it was connected, apparently, to something called "WPMacroInstance" in "/LocalApps/WordPerfect.app/English.lproj/WordPerfect.nib" but that was as far as I could get in determining where I might go to fiddle with things. I'm afraid I've exceeded my grasp TjL -- Spam-altered address in effect, remove obvious portion if replying by email.
From: default <default@qualcomm.com> Newsgroups: comp.sys.next.programmer Subject: Global Menu??? Date: Mon, 05 Oct 1998 16:27:47 -0700 Organization: QUALCOMM, Incorporated; San Diego, CA, USA Message-ID: <361955F3.3427457C@qualcomm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I just created a multi document project under the Windows for Yellow Box, DR2. I built the application and the resulting app brings up a window with a menu in it. The file menu has a new which will create an additional window. Under windows, this paradym is wrong as an MDI frame should be created. The MDI frames menu acts like the global menu on the Mac. I was hoping that Openstep would be smart enough to create the MDI frame. Am I right or wrong? If I'm wrong, then how do I get the MDI frame up under Windows. Thanks, Michael
From: default <default@qualcomm.com> Newsgroups: comp.sys.next.programmer Subject: Global Menu??? Date: Mon, 05 Oct 1998 16:28:40 -0700 Organization: QUALCOMM, Incorporated; San Diego, CA, USA Message-ID: <36195628.8C977801@qualcomm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I just created a multi document project under the Windows for Yellow Box, DR2. I built the application and the resulting app brings up a window with a menu in it. The file menu has a new which will create an additional window. Under windows, this paradym is wrong as an MDI frame should be created. The MDI frames menu acts like the global menu on the Mac. I was hoping that Openstep would be smart enough to create the MDI frame. Am I right or wrong? If I'm wrong, then how do I get the MDI frame up under Windows. Thanks, Michael
From: "Alex Molochnikov" <alex@gestalt.com> Newsgroups: comp.sys.next.programmer Subject: Re: stripping yellowbox .EXE files? Date: Mon, 5 Oct 1998 17:51:55 -0600 Organization: Canada Connect Corp. Message-ID: <6vbm1o$frh$1@cleavage.canuck.com> References: <slrn71hear.5o2.rog@talisker.ohm.york.ac.uk> Roger Peppe wrote in message ... >is it possible to strip the symbolic information from executables >produced under the YellowBox? > You will have to edit /Apple/Developer/Makefiles/pb_makefiles/flags.make file. Change the default setting of "DEBUG_SYMBOLS_CFLAG = -g" to "DEBUG_SYMBOLS_CFLAG =". This way the default debug mode will be OFF when you build your projects. If you want to turn it ON for a selected project, add the following flag to your Makefile.postamble file: DEBUG_SYMBOLS_CFLAG = -g (note that the name of the flag in the comments is plural: DEBUG_SYMBOLS_CFLAGS. Do not use this form, it is just a bug). Alex Molochnikov Phoenix Data Trend alex@gestalt.com
From: dfevans@bbcr.uwaterloo.ca (David Evans) Newsgroups: comp.sys.next.programmer Subject: Re: WP Generating the wrong code Date: 5 Oct 1998 23:53:54 GMT Organization: University of Waterloo Message-ID: <907631634.744153@watserv4.uwaterloo.ca> References: <6v3k31$cmf@obi-wan.fdt.net> <907465808.485937@watserv4.uwaterloo.ca> <6vbhc0$c36@obi-wan.fdt.net> Cache-Post-Path: watserv4.uwaterloo.ca!unknown@bcr11.uwaterloo.ca In article <6vbhc0$c36@obi-wan.fdt.net>, TjL <luomat@peak.org.this.all.must.be.removed> wrote: > >Do you have WP and if so do you know how I would do it? > I looked at it and, like you, saw the connection to WPMacroInstance. The menu cell has a tag that's some large five-digit number (I'm not in front of my NS machine right now), so that's likely how the program determines which menu item was selected (ugh...) So the "trick" is to change that tag number to the "correct" one. I'm not sure how you'd proceed from here either, really. -- David Evans (NeXTMail/MIME OK) dfevans@bbcr.uwaterloo.ca Computer/Synth Junkie http://bbcr.uwaterloo.ca/~dfevans/ University of Waterloo "Default is the value selected by the composer Ontario, Canada overridden by your command." - Roland TR-707 Manual
Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: automatic GC in ObjC (was: Mixing Objective-C and C++) From: lars.farm@ite.mh.se (Lars Farm) Message-ID: <1dg9v7o.1wzca4uura4gN@dialup149-1-21.swipnet.se> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434583@news.charm.net> <6ufkdp$9l6$1@leonie.object-factory.com> <360f8387.346399652@news.charm.net> <6ugarg$nfo$1@leonie.object-factory.com> <360C341F.E71303FB@mdlink.de> <6v0hjk$rt5$1@news.spacelab.net> <1dg8rtt.1tavyhj1becl2mN@dialup121-2-15.swipnet.se> <361427C2.4DB833B2@mdlink.de> Organization: pv Date: Fri, 2 Oct 1998 18:40:26 +0200 NNTP-Posting-Date: Fri, 02 Oct 1998 18:36:43 MET DST Helge Hess <hh@mdlink.de> wrote: > char *ptr = GC_malloc_atomic(256); > strcpy(ptr, "hi you"); > while (*ptr != ' ') ptr++; > > This is perfectly valid C and is often used code. It doesn't point to > the beginning of the object though, therefore the malloc'ed block will > be considered dead by the GC. You can configure it to ignore interior pointers but that would be foolish for C. A pointer is valid only if it points at an object or one after, in the case above the object is a character within a vector, not the vector, so it is still valid. If the pointer was set to array - 5 (which is sometimes done) then the results would be undefined. Same if it was to point more than one after the last element of the array. A pointer that points inside an array is valid and no problem for a GC (I can't see any measurable performance problem there in Boehms GC). Pointer arithmetic is defined only for valid pointers. Disclaimer. The C++ standard borrows these rules from C and I know them only from the C++ standard. Perhaps ANSI C is more permitting? > I think he meant that a conservative GC cannot reliably collect all > garbage since it has no ability to distiguish between eg a pointer and > an int on the stack. When using untyped memory this also applies to the > heap, but with ObjC you have typing information for all classes at > runtime, so no compiler support is needed. The GC has to be conservative > only about the root set. It is true that a conservative collector is conservative. It collects only what it can prove is a block of memory that can not be reached from the root set (global data, stack, registers and data accessible from there). Fortunately experience shows that very little is retained that could have been collected and then for very short periods. The probability for bit patterns that can be interpreted as pointers to allocated data is low and when they do occur in registers or on the stack they are often very short lived. There are also techniques to deal with some of these (implemented in Boehms collector). Even if not every byte that could be collected is collected at the first possible moment the worst that happens is that it is collected the next time or the time after that. In practice this really is no problem. A more common problem with GC (any GC) is a real and valid pointer that is stored in a global or longlived datastructure when there logically is no longer any use for it. No automatic GC can predict what the program will do (or not do) with valid data in the future. Even that is a small problem compared to manual garbage collection (delete p or RC) because the very same problem occurs with manual garbage collection. Finally, I agree with what was said earlier that GC doesn't solve every memory related problem imaginable. Compare it to a car with manual or automatic gear. Automatic works just fine almost always. Or perhaps an automatic washer compared to hand washing. A machine will not give special attention to individual spots, but the end result is still more than acceptable most of the time and one can still pay special attention to the very few spots that the machine cant handle if one wants to. Or perhaps assembly language. Who in their right mind writes the entire program in assembly these days? and if he does will it really perform better? contain fewer bugs? Perhaps it is better to first identify the hot spots and then consider assembly for those (if it really does make a difference)? > > No, they require one to restrict the ways in which one does pointer > > arithmetic so that the program is legal C. > > Hm. What do you consider legal C ? See above. There is little, if any defined behaviour that can fool a conservative collector. OTOH what is or isn't defined is not always obvious since the standard often allows a compiler to do what it likes by stating that the effects of this or that condition is undefined. "undefined behaviour" combined with "No diagnostic required". The reason is that these things, although illegal, are in general impossible to detect and report during compilation. So it is up to the programmer to know what the language allows and stay within those bounds. That's C and the C derivatives. Not the prettiest language there is, but it is popular - C is the DOS of programming languages. As long as one writes portable and legal C or C++ it is quite safe to use a conservative collector. Ah well, four decades of development of GC is perhaps not enough. Perhaps it takes another four decades before people sees the light? In any event C is not the obstacle that prevents the use of automatic GC. > For example what is illegal in > X-oring a pointer in C ? It may not be good style, but it's legal. The result is no longer a valid pointer and can not be used as a pointer unless the exact original bitpattern is restored. The results of the casts are implementation defined. Implementation defined is not portable, not even between versions of the same compiler. This of course makes the "feature" useless. This still doesn't fool a conservative collector. You have to manipulate the from a pointer converted integer such that the bitpattern no longer can be interpreted as a valid pointer to or into the block. But you can do it and it will fool a conservative collector as well as any programmer with just a tiny bit of common sense. One really has to be a fool to do such things and think that one can justify it. > Anyway, the posting was regarding ObjC and not C. With full C the > problem is much bigger (see the char example above) than with only > garbage collecting objects which the posting was about. The problem gets > even less because the libFoundation library already handles many cases > where special note would be needed (eg the above char-ptr examples) by > providing objects that wrap common C data types (eg NSString). > In practice I usually use it this way. I manage C structures/libraries > explicitly using malloc and free and ObjC objects with the collector. > This works quite fine for me, especially in cases like EOF where cycles > actually appear. I find this surprising because I asked the old time NeXTers on rhapsody-dev last autumn about using Boehms GC (that I had ported to Rhapsody PPC) with the NeXT libraries and the reply (from an Apple engineer) was that it wont work. NeXT does hide pointers from the collector using the above dubious techniques. Very disappointing. This is (still IMHO) a flaw in YB for reasons that I've tried to explain above. - Lars -- Lars Farm; lars.farm@ite.mh.se - Limt/Channelmatic: lars.farm@limt.se
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: automatic GC in ObjC (was: Mixing Objective-C and C++) Date: 2 Oct 1998 18:31:35 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn71a70c.229.rog@talisker.ohm.york.ac.uk> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434 <1dg9v7o.1wzca4uura4gN@dialup149-1-21.swipnet.se> On Fri, 2 Oct 1998 18:40:26 +0200, Lars Farm <lars.farm@ite.mh.se> wrote: > There is little, if any defined behaviour that can fool a > conservative collector. the following code is legal, and will fool any conservative collector i know of: void *foo(void) { static char ptstore[50]; void *mem; if (ptstore[0]) { mem = malloc(5000000); sprintf(ptstore, "%p", mem); } else { sscanf(ptstore, "%p", &mem); } return mem; } cheers, rog.
From: "Charles W. Swiger" <chuck@codefab.com> Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: automatic GC in ObjC (was: Mixing Objective-C and C++) Date: 2 Oct 1998 19:11:28 GMT Organization: Spacelab.net Internet Access Message-ID: <6v38h0$5bg$1@news.spacelab.net> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434583@news.charm.net> <6ufkdp$9l6$1@leonie.object-factory.com> <360f8387.346399652@news.charm.net> <6ugarg$nfo$1@leonie.object-factory.com> <360C341F.E71303FB@mdlink.de> <6v0hjk$rt5$1@news.spacelab.net> <1dg8rtt.1tavyhj1becl2mN@dialup121-2-15.swipnet.se> <361427C2.4DB833B2@mdlink.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Helge Hess <hh@mdlink.de> wrote: >Lars Farm wrote: >> Charles W. Swiger <chuck@codefab.com> wrote: >>> Like these debates always seem to, people tend to talk past each >>> others' points. In particular, I am not aware of any garbage >>> collection scheme which can always correctly determine whether memory >>> regions are in use if the language permits arbitrary pointer >>> arithmetic, which C and Obj-C do. >> >> But C doesn't allow arbitrary address arithmetic. K&R C does. ANSI C, as I recall from earlier discussions, states that the results are "undefined". However, on any system I know of where a pointer can be represented by casting to some integral type, complicated pointer arithmetic like shifting bits, subtracting and later adding to a base register (for a base/bounds setup) does not lose representation. >> A valid pointer must >> point at the object or one past. All other pointer values are invalid >> and can not be used (other than == and =), not even for arithmetic >> (undefined behaviour). > >This is not true. Consider this easy example which isn't tracked by >conservative GCs (Boehm GC can be configured to track it, but it results >in a rather large performance penalty because you need to track all >interior pointers): > > char *ptr = GC_malloc_atomic(256); > strcpy(ptr, "hi you"); > while (*ptr != ' ') ptr++; > >This is perfectly valid C and is often used code. Exactly, although a good GC must pay the performance penalty of tracking interior pointers to work with C-derived code at all. My point was that even this isn't good enough to work with all C programs, since some do more complicated manipulation of pointers than simple addition or subtraction of an integral value within an array. >This isn't so bad for ObjC, since you usually do not modify references >to objects. Actually I cannot think of an example why one would want to >do this. I've already given examples, like typing pointers or implementing a language interpreter with it's own heap (and base/bounds memory setup). >> > > A GC used in a wrong way may make a program less efficient but it >> > > nearly never makes a program buggy. >> > >> > I'll agree with this, assuming the collector itself isn't buggy and >> > assuming the [often deliberate] limitations of the language permit >> > strong GC, such as Java. >> >> How is this different from a complier with a code generation bug? and >> why is that an argument against GC? My former comment is no different from a code generation bug. My latter comment is in reference to what this current discussion about arbitrary pointer arithmetic is saying. >I think he meant that a conservative GC cannot reliably collect all >garbage since it has no ability to distiguish between eg a pointer and >an int on the stack. That you can live with, since a conservative GC will not free memory it thinks is in use, even if the memory is actually no longer in use and an integer happens to be aliasing to a valid memory address. >>>> Ever heard about conservative GC's ? C is of course a good thing, >>>> but conservative GC's support C and C++. >>> >>> Sure. They require one to restrict the ways in which one does pointer >>> arithmetic so that the GC doesn't mistakenly free a region of memory >>> that is still in use. Most of the time, that restriction is no >>> problem...but not always. >> >> No, they require one to restrict the ways in which one does pointer >> arithmetic so that the program is legal C. False. Undefined behavior in ANSI C does not mean the code is not legal. In particular, implementations are free to define what a particular operation will do so that the behavior becomes well-defined. -Chuck Charles Swiger | chuck@codefab.com | Yeah, yeah-- disclaim away. ----------------+-------------------+---------------------------- You have come to the end of your journey. Survival is everything.
Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: automatic GC in ObjC (was: Mixing Objective-C and C++) From: lars.farm@ite.mh.se (Lars Farm) Message-ID: <1dg8rtt.1tavyhj1becl2mN@dialup121-2-15.swipnet.se> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434583@news.charm.net> <6ufkdp$9l6$1@leonie.object-factory.com> <360f8387.346399652@news.charm.net> <6ugarg$nfo$1@leonie.object-factory.com> <360C341F.E71303FB@mdlink.de> <6v0hjk$rt5$1@news.spacelab.net> Organization: pv Date: Thu, 1 Oct 1998 23:02:46 +0200 NNTP-Posting-Date: Thu, 01 Oct 1998 22:59:03 MET DST Charles W. Swiger <chuck@codefab.com> wrote: > Like these debates always seem to, people tend to talk past each others' > points. In particular, I am not aware of any garbage collection scheme which > can always correctly determine whether memory regions are in use if the > language permits arbitrary pointer arithmetic, which C and Obj-C do. But C doesn't allow arbitrary address arithmetic. A valid pointer must point at the object or one past. All other pointer values are invalid and can not be used (other than == and =), not even for arithmetic (undefined behaviour). There are some "implementation defined" casts between integer types and pointers and those are the problems. One can also use a uninon with pointer/integer, but behaviour is only defined as long as you use it as one or the other, not both. Any program that uses them are non portable, or much more commonly illegal C programs. In general the use of such techniques must be considered good reason for an employer to fire the programmer that uses them. For the (very) few and nonportable C programs where one can justify it and knowingly bends the rules a conservative GC can not be used. In all the other GC can be used if a user wants to. > Helge Hess <hh@mdlink.de> wrote: > > A GC used in a wrong way may make a program less efficient but it nearly > > never makes a program buggy. > > I'll agree with this, assuming the collector itself isn't buggy and assuming > the [often deliberate] limitations of the language permit strong GC, such as > Java. How is this different from a complier with a code generation bug? and why is that an argument against GC? > > Ever heard about conservative GC's ? C is of course a good thing, but > > conservative GC's support C and C++. > > Sure. They require one to restrict the ways in which one does pointer > arithmetic so that the GC doesn't mistakenly free a region of memory that is > still in use. Most of the time, that restriction is no problem...but not > always. No, they require one to restrict the ways in which one does pointer arithmetic so that the program is legal C. -- Lars Farm; lars.farm@ite.mh.se - Limt/Channelmatic: lars.farm@limt.se
From: Helge Hess <hh@mdlink.de> Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: automatic GC in ObjC (was: Mixing Objective-C and C++) Date: Fri, 02 Oct 1998 02:09:22 +0100 Organization: MDlink online service center Message-ID: <361427C2.4DB833B2@mdlink.de> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434583@news.charm.net> <6ufkdp$9l6$1@leonie.object-factory.com> <360f8387.346399652@news.charm.net> <6ugarg$nfo$1@leonie.object-factory.com> <360C341F.E71303FB@mdlink.de> <6v0hjk$rt5$1@news.spacelab.net> <1dg8rtt.1tavyhj1becl2mN@dialup121-2-15.swipnet.se> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lars Farm wrote: > Charles W. Swiger <chuck@codefab.com> wrote: > > Like these debates always seem to, people tend to talk past each others' > > points. In particular, I am not aware of any garbage collection scheme which > > can always correctly determine whether memory regions are in use if the > > language permits arbitrary pointer arithmetic, which C and Obj-C do. > > But C doesn't allow arbitrary address arithmetic. A valid pointer must > point at the object or one past. All other pointer values are invalid > and can not be used (other than == and =), not even for arithmetic > (undefined behaviour). This is not true. Consider this easy example which isn't tracked by conservative GCs (Boehm GC can be configured to track it, but it results in a rather large performance penalty because you need to track all interior pointers): char *ptr = GC_malloc_atomic(256); strcpy(ptr, "hi you"); while (*ptr != ' ') ptr++; This is perfectly valid C and is often used code. It doesn't point to the beginning of the object though, therefore the malloc'ed block will be considered dead by the GC. This isn't so bad for ObjC, since you usually do not modify references to objects. Actually I cannot think of an example why one would want to do this. > > > A GC used in a wrong way may make a program less efficient but it nearly > > > never makes a program buggy. > > > > I'll agree with this, assuming the collector itself isn't buggy and assuming > > the [often deliberate] limitations of the language permit strong GC, such as > > Java. > > How is this different from a complier with a code generation bug? and > why is that an argument against GC? I think he meant that a conservative GC cannot reliably collect all garbage since it has no ability to distiguish between eg a pointer and an int on the stack. When using untyped memory this also applies to the heap, but with ObjC you have typing information for all classes at runtime, so no compiler support is needed. The GC has to be conservative only about the root set. > > > Ever heard about conservative GC's ? C is of course a good thing, but > > > conservative GC's support C and C++. > > > > Sure. They require one to restrict the ways in which one does pointer > > arithmetic so that the GC doesn't mistakenly free a region of memory that is > > still in use. Most of the time, that restriction is no problem...but not > > always. > > No, they require one to restrict the ways in which one does pointer > arithmetic so that the program is legal C. Hm. What do you consider legal C ? For example what is illegal in X-oring a pointer in C ? It may not be good style, but it's legal. Anyway, the posting was regarding ObjC and not C. With full C the problem is much bigger (see the char example above) than with only garbage collecting objects which the posting was about. The problem gets even less because the libFoundation library already handles many cases where special note would be needed (eg the above char-ptr examples) by providing objects that wrap common C data types (eg NSString). In practice I usually use it this way. I manage C structures/libraries explicitly using malloc and free and ObjC objects with the collector. This works quite fine for me, especially in cases like EOF where cycles actually appear. Greetings Helge
From: Helge Hess <hh@mdlink.de> Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: automatic GC in ObjC (was: Mixing Objective-C and C++) Date: Fri, 02 Oct 1998 02:46:55 +0100 Organization: MDlink online service center Message-ID: <3614308F.7D8B98B9@mdlink.de> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434583@news.charm.net> <6ufkdp$9l6$1@leonie.object-factory.com> <360f8387.346399652@news.charm.net> <6ugarg$nfo$1@leonie.object-factory.com> <360C341F.E71303FB@mdlink.de> <6v0hjk$rt5$1@news.spacelab.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Charles W. Swiger wrote: > Helge Hess <hh@mdlink.de> wrote: > Like these debates always seem to, people tend to talk past each others' > points. :-) IMHO many people from the ObjC edge just didn't tried a real GC yet. Most are really happy with the advantage RC brought over the old alloc/free approach. Actually I noticed the advantages when I started with Java. If you don't do nasty programming in ObjC you can have a nearly as good GC support as in Java. > In particular, I am not aware of any garbage collection scheme which > can always correctly determine whether memory regions are in use if the > language permits arbitrary pointer arithmetic, which C and Obj-C do. This isn't possible I think. But why would you want to apply arithmetic on pointers to objects ? I always considered id 'pointers' being object references. Object references do not support operations on them. Making this assumption has no drawbacks as far as I can see but makes conservative GC possible. > Consider a program which internally implements a language (one example out of > many is Emacs with its embedded ELISP interpreter) with tagged pointers, for > example. This doesn't count IMHO. The interpreter is outside the scope of the ObjC environment and must be considered external. So it need to be tracked by other means. If you take for example the guile interpreter, here you have to setup a SMOB structure which manages foreign structures. The same applies to TCL-ObjC gateways where ObjC pointers are usually stored in TCL strings (and aren't trackable because of that). You will explictly have to manage RC in this case as well. The example doesn't support GC, but it also does not deny it, it just needs special care. > > A GC used in a wrong way may make a program less efficient but it nearly > > never makes a program buggy. > > I'll agree with this, assuming the collector itself isn't buggy and assuming > the [often deliberate] limitations of the language permit strong GC, such as > Java. ObjC supports strong GC over heap allocated objects since the runtime provides all necessary information to allocate typed memory. The only point where the GC needs to be conservative is the root-set. This seems very acceptable to me, especially considering the fact that the heap is the difficult thing to track (because of cyclic refs), not the stack. > > Every manual memory management, RC as well, is very difficult to track if > > you have complex memory structures. > > Somewhat. It helps a great deal to have a well-understood convention for > your reference counting so that everyone knows what to do. This doesn't help you much with dynamic structures that involve user interaction. The user can connect/disconnect objects in ways you cannot foresee. RC is only applicable in static structures (eg hierachies, like the NSView tree), otherwise you have to live with the possibility of leaks. In EOF applications you usually have retain-cycles, these are resolved by reconverting objects to faults from time to time. Anyway, RC makes it much easier than with alloc/free, so it's a good thing as well :) RC helps you with object ownership and GC additionally helps you with cyclic references and automatic maintenance of ownership. > Foundation's memory management makes memory management much easier than > explicit malloc()/free(). It is a middle ground between the convenience and > expense of full GC and the inconvenience and (potentially) higher efficiency > of explicit memory management. Note that with Boehm GC you actually can explicitly free objects. In fact it would be possible to use Boehm GC together with RC, letting the GC resolve retain counts but making the object free itself explicitly when it reaches a retain count of zero. But in practice this doesn't make much sense since RC in Foundation has a lot of messaging overhead that makes RC slower than any collection of the GC .. However, you can configure Boehm GC as a leak detector ! This is quite cool, since you can discover retain cycles this way. In this mode the GC will report you any structures which are not reachable from the root set .. nice feature IMHO. > DO is recommended for inter-thread communication more in order to take > advantage of the properties of Mach messaging and less in order to avoid > these supposed problems with MT reference counting. Hm, I don't understand this, could you explain ? What advantages has Mach messaging over directly passing a pointer to the other thread ? Actually the shared memory is one of the major benefits of multiple threads above multiple processes. > The Foundation's GC > works just fine in a multithreaded enviroment, unless you have some major > performance penalties you can demonstrate for us. The Foundations RC itself is thread-safe as far as I know. However it doesn't solve the problems which arise with the parallel deallocation of objects in thread-shared data structures. MT adds time complexity (locking) to the already existing complexity of space (cyclic references) of GC. I'm not sure about the performance. I don't think one of the two approaches has particular advantages/disadvantages in speed in a multithreaded environment. Note that only parts of the GC can run in parallel, so for a full collection the GC needs safe points (which means stopping all user threads). There is a paper out there in the net called 'mostly parallel garbage collection' which discusses this topic. > [ ... ] > > Ever heard about conservative GC's ? C is of course a good thing, but > > conservative GC's support C and C++. > > Sure. They require one to restrict the ways in which one does pointer > arithmetic so that the GC doesn't mistakenly free a region of memory that is > still in use. Most of the time, that restriction is no problem...but not > always. Right. I like having GC most of the time. If there is a problem I encapsulate the problem into a C library using usual memory management and access it as an external resource. Only very rare/special tasks require special handling of garbage collected objects in ObjC. Greetings Helge
From: wori0011@rz03.FH-Karlsruhe.DE (Richard Woeber) Newsgroups: comp.sys.next.programmer Subject: PSWrap - Make error Date: 6 Oct 1998 08:55:07 GMT Organization: Fachhochschule of Karlsruhe, Germany Message-ID: <6vcltb$bfn$1@nz12.rz.uni-karlsruhe.de> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Every time I try to compile a project (Project Builder NSI 3.3) which contains an *.psw file, make complains about it 'Don't know how to compile xxx.psw' It used to work in former projects. Whats wrong. (I did install DeveloperPackeges a new) -- ============================================================================== Richard Woeber | login: Bill G Student of Cartography @ FH Karlsruhe | password: 97816 Lohr am Main /Germany | $msword | ed is the standard UNIX Text editor rich@qtvr.com NeXTMail ok | Your disk quota has been set to 50k http://www.qtvr.com/rich | $_ ==============================================================================
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: automatic GC in ObjC (was: Mixing Objective-C and C++) Date: 2 Oct 1998 11:44:18 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn719f4l.140.rog@talisker.ohm.york.ac.uk> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434 <361427C2.4DB833B2@mdlink.de> On Fri, 02 Oct 1998 02:09:22 +0100, Helge Hess <hh@mdlink.de> wrote: > Hm. What do you consider legal C ? For example what is illegal in > X-oring a pointer in C ? It may not be good style, but it's legal. it's legal to xor a pointer that has been converted to an integer type, but it is not legal to use it as a pointer again after you've done so. pointer arithmetic *is* restricted by the C language definition; however most compilers or C runtime environments are not so picky. rog.
From: luomat@peak.org.this.all.must.be.removed (TjL) Newsgroups: comp.sys.next.programmer,comp.sys.next.software Subject: WP Generating the wrong code Followup-To: comp.sys.next.programmer Date: 2 Oct 1998 22:28:49 GMT Organization: Florida Digital Turnpike Message-ID: <6v3k31$cmf@obi-wan.fdt.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NOTE: FOLLOWUPS TO COMP.SYS.NEXT.PROGRAMMER Something that has always bothered me about WordPerfect for NeXT that I just figured out today. WP has a feature called "Flush Right" which allows you to put some text flush against the right margin, ie: John Q. Smith 2 October 1998 where the "1998" is right against the right margin. Unfortunately this never worked in WP for NeXT. Today I discovered why (although it wasn't difficult to figure out). The menu item labelled "Flush Right" actually generates the code for "Right Justify" instead, so the entire line is right justified: John Q. Smith 2 October 1998 I was wondering if there was some neat NeXTStep hackish trick that could change the code associated with the menu item? Hope-fully, TjL -- Spam-altered address in effect, remove obvious portion if replying by email.
From: dfevans@bbcr.uwaterloo.ca (David Evans) Newsgroups: comp.sys.next.programmer Subject: Re: WP Generating the wrong code Date: 4 Oct 1998 01:50:08 GMT Organization: University of Waterloo Message-ID: <907465808.485937@watserv4.uwaterloo.ca> References: <6v3k31$cmf@obi-wan.fdt.net> Cache-Post-Path: watserv4.uwaterloo.ca!unknown@bcr11.uwaterloo.ca In article <6v3k31$cmf@obi-wan.fdt.net>, TjL <luomat@peak.org.this.all.must.be.removed> wrote: > >I was wondering if there was some neat NeXTStep hackish trick that could >change the code associated with the menu item? > Although Charlie's response makes it sound as though WP won't do what you want you can always open (a copy of!) WP's main .nib file in InterfaceBuilder and have at the object connections. This is a fun way to customise all your apps, although I think there's some good taste law against its excessive use. -- David Evans (NeXTMail/MIME OK) dfevans@bbcr.uwaterloo.ca Computer/Synth Junkie http://bbcr.uwaterloo.ca/~dfevans/ University of Waterloo "Default is the value selected by the composer Ontario, Canada overridden by your command." - Roland TR-707 Manual
From: Chris Hurlbutt-XIOtech <chrish@xiotech.com> Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.palmtops,comp.sys.pen,comp.sys.prime,comp.sys.proteon,comp.sys.pyramid,comp.sys.ridge,comp.sys.sequent,comp.sys.sgi.admin,comp Subject: Re: "$$$U NEED CASH $$$?READ THIS$$$$!!! Date: Tue, 06 Oct 1998 15:12:11 -0400 Organization: XIOtech Message-ID: <361A6B8B.E0505B96@xiotech.com> References: <6uqh5p$nt6@news1.emarites.net.ae> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="------------B4785114D607764882106CB6" To: Asma <rebel@emirates.net.ae> This is a multi-part message in MIME format. --------------B4785114D607764882106CB6 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit SGI salesman looking for revenue angle and a fast buck while NT product is in development... Asma wrote: > You Gotta Try This!!!!!!!! > > Turn $6 into $60,000 in just a matter of weeks!!! > This is simple, safe, and it really works! For $6 (U.S), 6 stamps, and > as much of your time as you want to devote, you could earn a year's salary > in a > month. > Sounds to good to be true, but just imagine. WHAT IF? > A little while back, I was browsing these newsgroups, just like > you are now, and came across an article similar to this that said you > could make thousands of dollars within weeks with only an initial > investment of $6.00! > So I thought, "Yeah, right, this must be a scam!" but like most of us, I > was curious. Like most of us, I kept reading. Anyway, it said that if > you send $1.00 to each of the 6 names and addresses stated in the > article, you could make thousands in a very short period of time. You then > place your own name and address at the bottom of the list at #6, and post > the > article to at least 200 newsgroups. (There are about 22,000.) or e-mail > them to friends, or e-mailing lists... > No catch, that was it. > Even though the investment was a measly $6, I had three questions that > needed to be answered before I could get involved in this sort of thing. > 1. IS THIS REALLY LEGAL? > I called a lawyer first. The lawyer was a little skeptical that I would > actually make any money but he said it WAS LEGAL if I wanted to try it. > I told him it sounded a lot like a chain letter but the details of the > system (SEE BELOW) actually made it a legitimate legal business. > 2. Would the Post Office be ok with this...? > I called them: 1-800-725-2161 and they confirmed THIS IS ABSOLUTELY > LEGAL! > (See Title 18,h sections 1302 NS 1341 of Postal Lottery Laws). This > clarifies the program of collecting names and addresses for a mailing > list. > 3. Is this moral? > Well, everyone who sends me a buck has a good chance of getting A LOT > of money ... a much better chance than buying a lottery ticket! > So, having these questions answered, I invested EXACTLY $7.92 ... six > $1.00 bills and six 32 cent postage stamps ... and boy am I glad I did! > Within 7 days, I started getting money in the mail! I was shocked! I > still figured it would end soon, and didn't give it another thought. But the > money just continued coming in. In the first week I made a few dollars > in the second week I made about twenty and now 5 weeks into it I've made > $500 > and > it's still coming in..... To make lots of money at this you need to be > persistent and > continue to post, even if you only spend 20 minutes a day doing it the > payoff > is > worth the effort. > It's certainly worth $6.00 and 6 stamps! > So now I'm reposting this so I can make even more money! The *ONLY* > thing stopping *ANYONE* from enriching their own bank account is pure > laziness! > It took me all of 5 MINUTES to print this out, follow the directions, > and begin posting to newsgroups. It took me a mere hour and a half to post > to over 500 newsgroups/bulletin boards. And for this GRAND TOTAL investment > of $ 7.92 (US) > and under TWO HOURS of my time, I have reaped an incredible amount of > money > -- > like nothing I've ever even heard of anywhere before! > 'Nuff said! > Let me tell you how this works, and most importantly, why it works. > Also, make sure you print a copy of this article now, so you can get the > information off of it when you need it. The process is very simple and > consists of THREE easy steps. > ============ > HOW IT WORKS > ============ > STEP 1: > ------ > Get 6 separate pieces of paper and write the following on each piece of > paper: > PLEASE ADD ME TO YOUR MAILING LIST. > $1 US DOLLAR PROCESSING FEE IS ENCLOSED. > (THIS IS KEY AS THIS IS WHAT MAKES IT LEGAL SINCE YOU ARE > PAYING FOR AND > LATER OFFERING A SERVICE). > Now get 6 $1.00 bills and place ONE inside EACH of the 6 pieces of paper > so the bill will not be seen through the envelope to prevent theft/robbery. > Then, place one paper in each of the 6 envelopes and seal them. You > should now have 6 sealed envelopes, each with a piece of paper stating > the above phrase and an U.S. $1.00 bill. > Mail the 6 envelopes to the following addresses: > > #1- Carsten Madsen > Hovabgsvej 59 > 9500 Hobro > Dk, Denmark > #2- Marley Hanson > 402 East Eagle Road > Havertown,pa 19083 > USA > #3- P. Lavin > 510 Walnut Street > Suite 1000 > Philadelphia,Pa 19106 > USA > #4- N. Russell > 7518 N. 20th. Street > Philadelphia,Pa 19138 > USA > #5- Ms. S. Buchanan > 706 23rd Avenue South > Seattle, WA 98144 > USA > #6- Asma Imtiaz > P.O Box 1493 > Sharjah > U.A.E > STEP 2: Now take the #1 name off the list that you see above, move the > other names up (6 becomes 5, 5 becomes 4, etc.) and add YOUR Name as > number 6 on the list. (If you want to remain anonymous put a nickname, > but the address MUST be correct. It, of course, MUST contain your > country, state/district/area, zip code, etc! You wouldn't want your money to > fly > away, would you?). > STEP 3: Now post your amended article to at least 200 newsgroups. > Remember that 200 postings are just a guideline. The more you post, the > more money you make! > Don't know HOW to post in the news groups? Well do exactly the > following: > ------------------------------------------------------------------------ > HOW TO POST TO NEWSGROUPS FAST WITH YOUR WEB > BROWSER: > The fastest way to post a newsletter: Highlight and COPY > (Ctrl-C) the text of this posted message and PASTE (Ctrl-V) it into a > plain text editor (as Wordpad) and save it. After you have made the > necessary changes that are stated above, simply COPY (Ctrl-C) and PASTE > (Ctrl-V) > the text into the message composition window, after selecting a > newsgroup, and post it! (Or you can attach the file, without writing > anything to > the message window.) > ------------------------ > ------------------------------------------------------------------------ > If you have Netscape Navigator 3.0 do the following: > 1. Click on any newsgroup like normal, then click on 'TO NEWS'. This > will bring up a box to type a message in. > 2. Leave the newsgroup box like it is, change the subject box to > something flashy, something to catch the eye, as "$$$ NEED CASH $$$? > READ HERE! $! $! $" Or "$$$! MAKE FAST CASH, YOU CAN'T LOSE! $$$". Or you > can use my subject title. > 3. Now click on 'ATTACHMENTS'. Then click on 'ATTACH FILE'. Find your > file on your Hard Disk (the one you saved from the text editor). Once > you find it, click on it and then click 'OPEN' and 'OK'. You should now see > your file name in the attachments box. > 4. Now click on 'SEND'/'POST'. You see? Now you just have 199 to go! > (Don't worry, it's easy and quick once you get used to it.) NOTE: All > the versions of Netscape Navigator's are similar to each other, so you'll > have no problem to do this if you don't have Netscape Navigator 3.0. > ------------------------------------------------------------------------ > ! QUICK TIP! > (For Netscape Navigator 3.x and above) > You can post this message to many newsgroups at a time, by simply > selecting a newsgroup near the top of the screen, hold > down the SHIFT, and then select a newsgroup near the bottom of the > screen. All of the newsgroups in/between will be selected. After that, you > follow/do the basic steps, stated below at this letter, except for step #1. > You can go to the page stated below in this letter and click on a > newsgroup to open up the newsgroups window. Once you've done this, in the > same window go to 'OPTIONS', and then mark 'SHOW ALL NEWSGROUPS' and 'SHOW > ALL MESSAGES'. > Now you can see all the newsgroups and you can apply easier the above > tip. > ------------------------------------------------------------------------ > If you have MS Internet Explorer do the following: > 1. Go to the newsgroups and press 'POST AN ARTICLE'. To the new > window type your headline in the subject area and then click in the > large window below. > There either PASTE your letter (which it's been copied from the text > editor), or attach the file which contains it. > 2. Then click on 'SEND' or 'OK'. NOTE: All versions of MS Internet > Explorer are similar to each other, so you won't have any problem doing > this. > GENERAL NOTES ON POSTING: A nice page where you'll find all the > newsgroups if you want help is http://www.liszt.com/ (When you go to the > home page, click on the link 'Newsgroup Directory'). But I don't think > you'll have any problem posting because it's very easy once you've > found the newsgroups. All these web browsers are similar. It doesn't > matter which one you have. (But it makes it very easy if you have > Netscape Navigator 3.0 or later. You may download it from the Internet if > you > don't have it.) You just have to remember the basic steps, stated below. > BASIC STEPS FOR POSTING: > 1. Find a newsgroup and you click on it. > 2. You click on 'POST AN/NEW ARTICLE' or 'TO NEWS' or anything else > similar to these. > 3. You type your flashy headline in the subject box. > 4. Now, either you attach the file containing your amended letter, or > you PASTE the letter. (You have to COPY it from the text editor, of > course, from before.) > 5. Finally, you click on 'SEND' or 'POST' or 'OK', whatever is there. > ------------------------------------------------------------------------ > **REMEMBER, THE MORE NEWSGROUPS YOU POST IN, THE MORE > MONEY YOU WILL MAKE! BUT YOU HAVE TO POST A MINIMUM OF 200** > That's it! You will begin receiving money from around the world within > days! If you don't see immediate results don't give up just post to more > newsgroups, > e-mail your friends or whatever and I guarentee you'll see BIG results! > You may eventually want to rent a P.O.Box due to the large amount > of mail you receive. If you wish to stay anonymous, you can invent a > name to use, as long as the postman will deliver it. > **JUST MAKE SURE ALL THE ADDRESSES ARE CORRECT. ** > ------------------------------------------------------------------------- > ================= > Now the WHY part: > ================= > Out of 200 postings; say I receive only 5 replies (a very low example). > So then I made $5.00 with my name at #6 on the letter. Now, each of the > 5 persons who just sent me $1.00 make the MINIMUM 200 postings, each > with my name at #5 and only 5 persons respond to each of > the original 5, that is another $25.00 for me, now those 25 each make > 200 MINIMUM posts with my name at #4 and only 5 replies each, I will > bring in an additional $125.00! Now, those 125 persons turn around and post > the MINIMUM 200 with my name at #3 and only receive 5 replies > each, I will make an additional $625.00! OK, now here is the fun part, > each of those 625 persons post a MINIMUM 200 letters with my name at #2 > and they each only receive 5 replies, that just made me $3,125.00! Those > 3,125 persons will all deliver this message to 200 newsgroups with my name > at #1 and if still 5 persons per 200 newsgroups react I will receive > $15,625,00! MAKE SURE YOU RECIEVE AT LEAST 5 RESPONSES WITH IN THE > FIRST 2 WEEKS AS THIS WILL INSURE THE ABOVE SCENARIO WORKS!!!!! > With an original investment of only $6.00! AMAZING! And as I said 5 > responses is actually VERY LOW! Average are probable 20 to 30! So lets > put those figures at just 15 responses per person. Here is what you will > make: > at #6 $15.00 > at #5 $225.00 > at #4 $3,375.00 > at #3 $50,625.00 > at #2 $759,375.00 > at #1 $11,390,625.00 > When your name is no longer on the list, you just take the latest > posting in the newsgroups, and send out another $6.00 to names on the > list, putting your name at number 6 again. And start posting again. The > thing to remember is, do you realize that thousands of people all over the > world > are joining the internet and reading these articles everyday, JUST LIKE YOU > are now! > So can you afford $6.00 and see if it really works? I think so... People > have said, "what if the plan is played out and no one sends you the > money? So what! What are the chances of that happening when there are tons > of new honest users and new honest people who are joining the internet and > newsgroups everyday and are willing to give it a try? Estimates are at > 20,000 to 50,000 new users, every day, with thousands of those joining > the actual Internet. Remember, play FAIRLY and HONESTLY and this will > work. You just have to be honest. > By the way, if you try to deceive people by posting the messages with > your name in the list and not sending the money to the rest of the > people already on the list, you will NOT get as much. Someone I talked to > knew someone who did that and he only made about $150.00, and that's after > eight weeks! Then he sent the 6 $1.00 bills, people added him to their > lists, > and in 4-5 weeks he had over $10k. This is the fairest and > most honest way I have ever seen to share the wealth of the world without > costing anything but our time! You also may want to buy mailing and > e-mail lists for future dollars. Make sure you print this article out > RIGHT NOW, also. Try to keep a list of everyone that sends you money and > always keep an eye on the newsgroups to make sure everyone is playing > fairly. > Remember that HONESTY IS THE BEST POLICY. You don't need to cheat the > basic idea to make the money! > GOOD LUCK to all and please play fairly and reap the huge rewards from > this, which is tons of extra CASH. Please remember to declare your extra > income. Thanks once again... > =========================================================== > ========== > LEGAL? ? ? (Comments from Bob Novak who started this new version.) > "People have asked me if this is really legal. Well, it is! You are > using the Internet to advertise you business. What is that business? You are > assembling a mailing list of people who are interested in home based > computer and online business and methods of generating income at home. > Remember that people send you a small fee to be added to your mailing > list. > It is legal. What will you do with your list of thousands of names? > Compile all of them into a database and sell them as "Mailing Lists" > on the internet in a similar manner, if you wish, and make more money. > How do you think you get all the junk mail that you do? Credit card > companies, mail order, Utilities, anyone you deal with through the mail > can sell your name and address on a mailing list, unless you ask them > not to, in addition to there regular business, So, why not do the same > with the list you collect. You can find more info about "Mailing Lists" > on the internet using any search engine. --------------B4785114D607764882106CB6 Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Christopher Hurlbutt Content-Disposition: attachment; filename="vcard.vcf" begin: vcard fn: Christopher Hurlbutt n: Hurlbutt;Christopher org: XIOtech Corporation email;internet: chrish@xiotech.com title: Phone Number: (301) 918-2538 tel;work: 301 918-2538 tel;fax: 301 918-2539 note: 301 918-2538 Direct Dial x-mozilla-cpt: ;0 x-mozilla-html: TRUE version: 2.1 end: vcard --------------B4785114D607764882106CB6--
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: automatic GC in ObjC (was: Mixing Objective-C and C++) Date: 4 Oct 1998 12:03:27 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn71ep0i.20a.rog@talisker.ohm.york.ac.uk> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434 <36157A24.CBBAB7B4@adaptive.de> On Sat, 03 Oct 1998 01:13:08 +0000, Sinisa Susnjar <sini@adaptive.de> wrote: > Roger Peppe wrote: > > it's legal to xor a pointer that has been converted to an integer > > type, but it is not legal to use it as a pointer again after > > you've done so. > > pointer arithmetic *is* restricted by the C language definition; > > however most compilers or C runtime environments are not so picky. > > I fully agree with you that one cannot xor a pointer directly and that > pointer arithmetic is restricted by the C language itself, but I do > disagree that the use of such a xored pointer is "illegal" - it is > admitedly bad style. In my opinion xoring a pointer (after a cast) is > not > more "illegal" than assigning an int to a pointer. The point is, that as > far as C-programming is concerned, nothing is "illegal" when you know > what you're doing. for one release of one compiler one one platform, perhaps. but most of us try to write code that is as portable as is feasible, given the requirements (well, i do anyway!). writing code that invokes behaviour that is undefined within the C standard is almost always unnecessary. > The following is completely "legal": > #include <stdio.h> > int main(void) > { > char *p1 = "100", *p2 = "0"; > int i1 = (int)p1, i2 = (int)p2; /* assumes sizeof(int) == > sizeof(char*) > i1 ^= i2 ^= i1 ^= i2; > printf("%s%% tricky, %s%% bad style\n", (char*)i1, (char*)i2); > return 0; > } it might work on some machines. it will fail on many others. the assignment of pointers to ints uses implementation-defined behaviour; on many platforms it will work, but many machines have 2-byte ints and 4-byte pointers. that xor-assignment trick invokes undefined behaviour, because the language standard does not define in what order the assignments will take place, it's just as nasty as: i = i++; and just as unnecessary (and usually slower than {int t=i1; i1=i2; i2=t}) the point is that if you make assumptions like the above, then you are writing C that is very specific to a particular platform, and you have no guarantees that it'll work at all under any other circumstances, for instance in the presence of a conservative GC, or with a different release of the compiler. there is a big difference between "implementation defined" (e.g. the assignment of pointers to ints) and "undefined" (the compiler can do what it damn well pleases). > BTW, what do you mean by "C runtime environments" ? the code environment in which the compiled C program is running. often this is minimal, consisting only of support for a stack, but under unix, for example, access to dodgy areas of memory is trapped; some C runtime environments manage to trap accesses beyond array boundaries (a neat trick). this is what i meant by the runtime environment being "picky". > Sorry for being off-topic, i don't think you're that far off topic - it's important that people writing Objective-C realise they're writing C too, and understand the constraints of the language. cheers, rog.
Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: automatic GC in ObjC (was: Mixing Objective-C and C++) From: lars.farm@ite.mh.se (Lars Farm) Message-ID: <1dge0ka.8jmb84v8jlocN@dialup120-2-14.swipnet.se> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434583@news.charm.net> <6ufkdp$9l6$1@leonie.object-factory.com> <360f8387.346399652@news.charm.net> <6ugarg$nfo$1@leonie.object-factory.com> <360C341F.E71303FB@mdlink.de> <6v0hjk$rt5$1@news.spacelab.net> <1dg8rtt.1tavyhj1becl2mN@dialup121-2-15.swipnet.se> <361427C2.4DB833B2@mdlink.de> <1dg9v7o.1wzca4uura4gN@dialup149-1-21.swipnet.se> <slrn71a70c.229.rog@talisker.ohm.york.ac.uk> <1dgamnc.1uxxnsa1xsfimoN@dialup149-1-21.swipnet.se> <slrn71ennh.20a.rog@talisker.ohm.york.ac.uk> Organization: pv Date: Sun, 4 Oct 1998 22:07:22 +0200 NNTP-Posting-Date: Sun, 04 Oct 1998 22:03:33 MET DST Roger Peppe <rog@ohm.york.ac.uk> wrote: > On Fri, 2 Oct 1998 23:23:41 +0200, Lars Farm <lars.farm@ite.mh.se> wrote: > i was just pointing out that you can never be *certain* that a > conforming C program will behave correctly with a conservative GC. i This is true. There are some obscure things to avoid as you've shown. The rule is simple: Don't hide the last pointer to a block from the collector. > BTW, how do conservative GCs work when you've got large quantities of > non-pointer data around - e.g. large bitmaps? does it have to scan all > that data? i would imagine that would incurr a fairly large performance > overhead, particularly when you're using 20MB images. (my current app, > for example) As always, it depends :-) Since there is no standard for these things I'll just use Boehms collector as an example. Use GC_malloc_atomic() which returns a block that you have promised will NOT carry any pointer data. That block will not be scanned and will do fine for image data, text and such. - Lars -- Lars Farm; lars.farm@ite.mh.se - Limt/Channelmatic: lars.farm@limt.se
From: szallies@energotec.de (Constantin Szallies) Newsgroups: comp.sys.next.programmer Subject: Compile time performance: OS/Mach vs OE/NT Date: 5 Oct 1998 07:15:29 GMT Organization: TechNet GmbH Message-ID: <6v9rmh$deu$1@oxygen.technet.net> NNTP-Posting-Date: 5 Oct 1998 07:15:29 GMT Here is my result from a performance test I did in your firm. I posted this because it might be interesting for other people as well and maybe I get a useful hint or surgestion 8-) --------- Compile time testrun OS4.2/Mach vs OS Enterprise WinNT Client in both cases: Intel P200, 64 MB, SCSI Fileserver: Intel P200, 128 MB, SCSI, OS4.2/Mach Test results for "make all" om a library aggregate project. "make clean" was executed before the test. The indexer was not running. NT Client - m and o files on fileserver, access with NFS Hummingbird -> 15 min - m files on fileserver, o files local, access with NFS Hummingbird -> 8 min - m and o files on fileserver, access with samba on the server-side -> 16 min - m files on fileserver, o files local, access with samba on the server-side -> 8 min - m files and o files local -> 6 min OS Client - m and o files on fileserver, access with NFS -> 6 min - m files on fileserver, o files local, access with NFS -> 4 min - m files and o files local -> 3 min Result: - On NT performance doesn't benefit from using NFS - On Mach you win only 1 minute by placing the m files locally. - On NT you win 8 minutes by placing the o files locally. But you win only 2 minutes more by placing also the m files on the local harddisk. - Compiling on NT takes at least two times longer than on Mach -- # Constantin Szallies, Energotec GmbH # szallies@energotec.de # http://www.energotec.de/~szallies/ # 49211-9144018
From: "Ethan" <ethan@enteract.com> Newsgroups: comp.sys.next.programmer Subject: Linux Compatibility Library? Date: Tue, 6 Oct 1998 22:48:15 -0500 Organization: EnterAct L.L.C. Turbo-Elite News Server Message-ID: <6venrv$jr1$1@eve.enteract.com> Since BeOS R4 is going to be ELF, has anyone started programming a Linux System Library for BeOS... Maybe libc or glibc. It would be theoretically possible, and not too hard since a lot of BeOS's API's are somewhat similar to some Linux programming interfaces. FreeBSD is able to run Linux apps, and I think it would be a great strength for BeOS if someone programmed some Linux compatibility libraries... -Ethan ---------------------- ethan at enteract dot com http://www.enteract.com/~ethan
From: Alex Blakemore <alex@genoa.com> Newsgroups: comp.sys.next.programmer Subject: Re: Global Menu??? MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: default@qualcomm.com References: <361955F3.3427457C@qualcomm.com> Message-ID: <361aed09.0@nntp.mediasoft.net> Date: 7 Oct 98 04:24:41 GMT In <361955F3.3427457C@qualcomm.com> default wrote: > I just created a multi document project under the Windows for Yellow Box, DR2. > I built the application and the resulting app brings up a window with a menu in > it. The file menu has a new which will create an additional window. Under > windows, this paradym is wrong as an MDI frame should be created. The MDI > frames menu acts like the global menu on the Mac. I was hoping that Openstep > would be smart enough to create the MDI frame. Am I right or wrong? If I'm > wrong, then how do I get the MDI frame up under Windows. You're wrong, but Microsoft is even more wrong. MDI is just awful UI design. Even Microsoft is finally moving away from it (try outlook for instance). Given the absence of a global menu under Windows, and the need to have at least one window visible to hold a menu, you have these choices: a. create an empty document at launch (ala Draw example) b. create some utility or browser or open document list window at launch (ala Project Builder's open projects window) MDI really is not a (practical) option with OpenStep, and that's a good thing. -- Alex Blakemore alex@genoa.com NeXT, MIME and ASCII mail accepted
Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.palmtops,comp.sys.pen,comp.sys.prime,comp.sys.proteon,comp.sys.pyramid,comp.sys.ridge,comp.sys.sequent,comp.sys.sgi.admin,comp From: le@put.com (Louis Epstein) Subject: Re: "$$$U NEED CASH $$$?READ THIS$$$$!!! Followup-To: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.palmtops,comp.sys.pen,comp.sys.prime,comp.sys.proteon,comp.sys.pyramid Organization: Putnam Internet Services Message-ID: <F0FwLC.80o@news2.new-york.net> References: <6uqh5p$nt6@news1.emarites.net.ae> <361A6B8B.E0505B96@xiotech.com> Date: Wed, 7 Oct 1998 04:32:00 GMT Chris Hurlbutt-XIOtech (chrish@xiotech.com) wrote: : : SGI salesman looking for revenue angle and a fast buck while NT product is in : development... Well,as a comp.sys.northstar reader I can't wait to see how they're going to make NT run on an 8-bit CPU!
From: Christian Neuss <neuss.@informatik.th-darmstadt.de.nos-pam> Newsgroups: comp.sys.next.programmer,comp.sys.next.software Subject: Re: WO vs. ASP Date: 7 Oct 1998 10:35:37 GMT Organization: Technische Universitaet Darmstadt Message-ID: <6vfg5p$8d4$1@sun27.hrz.tu-darmstadt.de> References: <361A7394.49687B8@indiana.no_spam.edu> Allan Streib <astreib@indiana.no_spam.edu> wrote: >Could someone please give me impressions/pros/cons of developing Web >applications with WO compared to Microsoft's Active Server Pages? I >know the Dell web site used to be WO but now looks like ASP. I have not >used either but may have the opportunity here to push for one or the >other. Thanks for any info or references. WO scales well. Hence, it does not require that you move application state to the client side when traffic gets heavy :-) Also, would you rather program in Java and ObjectiveC or in some Fortran dialect? And you're looking at only $99 anyway (at least you post from an edu site), so hey, what more do you want. I've not even started to talk about EOF. Best regards, Chris -- // Christian Neuss "static typing? how quaint.." // http://www.nexttoyou.de/~neuss/ // fax: (+49) 6151 16 5472
Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.palmtops,comp.sys.pen,comp.sys.prime,comp.sys.proteon,comp.sys.pyramid From: mrm@lerami.lerctr.org (Ray Mullins) Subject: Re: "$$$U NEED CASH $$$?READ THIS$$$$!!! Sender: mrm@lerami.lerctr.org (Ray Mullins) Message-ID: <F0GtpH.3uB@lerami.lerctr.org> Date: Wed, 7 Oct 1998 16:27:17 GMT References: <6uqh5p$nt6@news1.emarites.net.ae> <361A6B8B.E0505B96@xiotech.com> <F0FwLC.80o@news2.new-york.net> Organization: Larry Rosenman's personal UnixWare 2.1.3 system In article <F0FwLC.80o@news2.new-york.net>, Louis Epstein <le@put.com> wrote: >Chris Hurlbutt-XIOtech (chrish@xiotech.com) wrote: >: >: SGI salesman looking for revenue angle and a fast buck while NT product is in >: development... > > >Well,as a comp.sys.northstar reader I can't wait to see how they're >going to make NT run on an 8-bit CPU! Ringing in from comp.sys.prime, I hadn't heard that NT was going to be ported over to the 50-series architecture... Later, Ray -- M. Ray Mullins (http://www.lerctr.org/~mrm/) from Roseville, California California Transit Publications - your one stop shop for transit marketing, publications, planning and web services at http://www.catransit.com/ TIPs: http://socaltip.lerctr.org http://norcaltip.lerctr.org http://cencaltip.lerctr.org
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: hangup backtabbing out of scrollview: any workarounds? Date: 7 Oct 1998 18:09:32 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn71nbiq.83.rog@talisker.ohm.york.ac.uk> this is a problem with the appkit under openstep 4.2 under intel and m68k; also under the yellowbox under NT. to see the problem: create some NSControl objects in interface builder (e.g. some text fields and some buttons) and group some of them into a scrollview; then test the interface (command-r) and try tabbing through the fields. tabbing forward works fine but tabbing backwards out of the first object in the scrollview gives an infinite loop within NSView's previousValidKeyView method. you'll have to kill IB. the same problem exhibits itself in the running application. does anyone have a workaround for this? (preferably that doesn't involve subclassing the first object in any scrollview!) cheers, rog.
From: tsfXXX@slip.net Newsgroups: comp.sys.next.programmer Subject: hard drive for color slab Date: Wed, 07 Oct 1998 20:21:20 GMT Organization: Slip.Net (http://www.slip.net) Message-ID: <361bcd29.16433934@news.slip.net> I have a friend with a non turbo color slab 68040 running OpenStep 4.2. It has both the original internal hard drive and an external hard drive. The external drive appears to be dieing and he would like to upgrade it. Does this hardware support normal SCSI, fast SCSI, or ultra-SCSI (this seems impossible due to the date it was made). Is there a problem on this type of system using drives > 2 gig. I remember that there was such a problem on my Intel system running OS 4.2 . Is there some way to partition a drive >2 gig to use the space? Are there any 2+ gig drives (manefacturer and/or models) with known problems on this system? I have checked NeXT answers ,in particular 2645, to no avail. If someone could help us out or point us to a faq on this subject, we would appreciate it very much. Thanks in advance. Tom Fisher
From: tsfXXX@slip.net Newsgroups: comp.sys.next.programmer Subject: Re: hard drive for color slab Date: Wed, 07 Oct 1998 20:23:31 GMT Organization: Slip.Net (http://www.slip.net) Message-ID: <361bcd8c.16532527@news.slip.net> References: <361bcd29.16433934@news.slip.net> My apologies, this was supposed to be posted on comp.sys.next.hardware. I have posted it there as well. On Wed, 07 Oct 1998 20:21:20 GMT, tsfXXX@slip.net wrote: > >I have a friend with a non turbo color slab 68040 running OpenStep >4.2. It has both the original internal hard drive and an external >hard drive. The external drive appears to be dieing and he would like >to upgrade it. > >Does this hardware support normal SCSI, fast SCSI, or ultra-SCSI (this >seems impossible due to the date it was made). > >Is there a problem on this type of system using drives > 2 gig. I >remember that there was such a problem on my Intel system running OS >4.2 . > >Is there some way to partition a drive >2 gig to use the space? > >Are there any 2+ gig drives (manefacturer and/or models) with known >problems on this system? > >I have checked NeXT answers ,in particular 2645, to no avail. If >someone could help us out or point us to a faq on this subject, we >would appreciate it very much. > >Thanks in advance. > >Tom Fisher > > >
From: Allan Streib <astreib@indiana.no_spam.edu> Newsgroups: comp.sys.next.programmer,comp.sys.next.software Subject: WO vs. ASP Date: Tue, 06 Oct 1998 14:46:28 -0500 Organization: Indiana University, Bloomington Message-ID: <361A7394.49687B8@indiana.no_spam.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Could someone please give me impressions/pros/cons of developing Web applications with WO compared to Microsoft's Active Server Pages? I know the Dell web site used to be WO but now looks like ASP. I have not used either but may have the opportunity here to push for one or the other. Thanks for any info or references. Allan Streib (remove no_spam from email address to reply)
From: group-admin@isc.org (David C Lawrence) Newsgroups: comp.sys.next.programmer Subject: cmsg newgroup comp.sys.next.programmer Control: newgroup comp.sys.next.programmer Message-ID: <907808748.2530@isc.org> Date: Thu, 08 Oct 1998 01:05:48 -0000 ftp://ftp.isc.org/pub/pgpcontrol/README iQCVAwUBNhwP7MJdOtO4janBAQFzAAP8DYy8tDFC3FR16/qZI4Y5Zkc1+2qA6vXt 7CFFx8QJdzKLXXGcmtiNXqlisY4OkSf6qRff9fmQCPetIwA6D87OXvhOb4yIilyW dVZGoUlLvwbZLIy3Q37Dsxf5ZhnsrjTxT+S5pWbZFC1iuDVZB6zCEkckmy/y86wp WnEsQByNG4o= =4R6D comp.sys.next.programmer is an unmoderated newsgroup which passed its vote for creation by 357:38 as reported in news.announce.newgroups on 2 July 1991. For your newsgroups file: comp.sys.next.programmer NeXT related programming issues. The charter, culled from the call for votes: This group will deal with NeXT related programming issues. Questions about porting of code to the NeXT, programming in NeXTStep environment, etc...
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <6vghn0$oc5823@pegasus.hkstar.com> Control: cancel <6vghn0$oc5823@pegasus.hkstar.com> Date: 07 Oct 1998 20:09:52 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6vghn0$oc5823@pegasus.hkstar.com> Sender: <hk_simon@internet-club.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: brian@dashboards-sw.com Newsgroups: comp.soft-sys.nextstep,comp.sys.next.programmer,uk.jobs.offered,misc.jobs.offered,us.jobs.offered Subject: U.S/U.K - NeXT/OpenStep/Yellowbox/WebObjects - International Development Project - London/Frankfurt/Philly/West Coast Date: Wed, 07 Oct 1998 14:20:41 GMT Organization: Dashboards Message-ID: <361b7745.55301811@news.btinternet.com> We are currently staffing for a multinational, multi-site OO project for a large international financial services company. The company is about to embark on a complete re-work of all front-end and middle office systems. As a member of one of the four major development teams you will be involved in the design and coding of an n-tier distributed system with Internet/Intranet front ends, that will be utilised by 300 trading personal and their associated support staff in four international locations. Current requirements exist for candidates with a good Project/Team Leader or Senior Developer/Developer background. These roles are technically demanding and require strong skills in the areas of OO analysis, design and development. Technically, candidates must have a strong OO background with solid NeXT/OpenStep/Yellowbox skills and at least two of the following: Java MacOS X Objective-C Objective Everything Objective Python Oak (Orb) Corba Carbon API WebObjects Positions are available in the following four locations: Philadelphia London Frankfurt San Francisco This is an opportunity to work on a brand new, high profile, international software development project using the latest in distributed OO techniques. Salaries are designed to attract the best: $75k to $120k + Bonus in the U.S £45k to £75k + Bonus + BB's in the U.K € Open + Bonus on European Mainland This is an exclusive assignment through this agency only. In the first instance, please e-mail your details (Resume/CV) to the address below. CONTACT: Brian Mitchell Dashboards Limited - London e-mail - brian@dashboards-sw.com
From: "Dirk P. Fromhein" <Dirk.Fromhein@watershed.com> Newsgroups: comp.sys.next.programmer,comp.sys.next.software Subject: Re: WO vs. ASP Date: Wed, 07 Oct 1998 10:38:22 -0400 Organization: Watershed Technologies, Inc. Message-ID: <361B7CDE.CC4D7053@watershed.com> References: <361A7394.49687B8@indiana.no_spam.edu> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="------------2F6DE851F3B3D18089ADC3C2" This is a multi-part message in MIME format. --------------2F6DE851F3B3D18089ADC3C2 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit There is yet another alternative Java Server Pages combined with something like our Relational Object Framework. JSP compiles to become a servlet (in our case) and then backs into our object to relational management system. It is then fully portable across all platforms and fully thread safe. We have it running on AIX, HP/UX, Solaris, and WinNT. Given the end result is a servlet it can also be run in any servlet capable web server. Good luck, Dirk Fromhein http://www.watershed.com Allan Streib wrote: > Could someone please give me impressions/pros/cons of developing Web > applications with WO compared to Microsoft's Active Server Pages? I > know the Dell web site used to be WO but now looks like ASP. I have not > used either but may have the opportunity here to push for one or the > other. Thanks for any info or references. > > Allan Streib > (remove no_spam from email address to reply) --------------2F6DE851F3B3D18089ADC3C2 Content-Type: text/x-vcard; charset=us-ascii; name="Dirk.Fromhein.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Dirk P. Fromhein Content-Disposition: attachment; filename="Dirk.Fromhein.vcf" begin:vcard n:Fromhein;Dirk P. x-mozilla-html:FALSE org:Watershed Technologies, Inc. adr:;;;Framingham;MA;01701;USA version:2.1 email;internet:Dirk.Fromhein@watershed.com title:CTO/President note:http://www.watershed.com info@watershed.com x-mozilla-cpt:;0 fn:Dirk P. Fromhein end:vcard --------------2F6DE851F3B3D18089ADC3C2--
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: stripping yellowbox .EXE files? Date: 8 Oct 1998 12:16:31 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn71pb92.12k.rog@talisker.ohm.york.ac.uk> References: <slrn71hear.5o2.rog@talisker.ohm.york.ac.uk> <SCOTT.98Oct5080748@slave.doubleu.com> On 5 Oct 98 08:07:48, Scott Hess <scott@nospam.doubleu.com> wrote: > In article <slrn71hear.5o2.rog@talisker.ohm.york.ac.uk>, > rog@ohm.york.ac.uk (Roger Peppe) writes: > is it possible to strip the symbolic information from executables > produced under the YellowBox? > > Use the "install" target in ProjectBuilder, it'll leave the stripped > version wherever you said to install it in the Build Attributes of > Project Inspector. There is no strip(1) type command under NT. i've tried this - it runs a dodgy find script that goes through lots of files that it shouldn't touch (e.g. non-nt object files through a symbolic link) and takes forever. as far as i can work out, it thinks that it's copying all the source and stripping all the objects. but it really doesn't work very well in the environment we have here, where we're developing a dual-platform (NT + openstep) app with the filesystem imported via samba onto the NT box, and various library directories symbolic linked into the app source directory. as far as i can work out, it's very likely that the whole lot will need to be recompiled anyway, so i've gone with alex@gestalt.com's suggestion of changing the flags in flags.make. that seems to work - now we get an executable that's only 6 times as big as the openstep one! (what *is* in all that data?) cheers, rog.
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: automatic GC in ObjC (was: Mixing Objective-C and C++) Date: 8 Oct 1998 09:45:25 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn71p2do.ss.rog@talisker.ohm.york.ac.uk> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434 <6vhu9n$7ib$1@kohl.informatik.uni-bremen.de> On Thu, 8 Oct 1998 10:50:25 +0200, "Sascha Bohnenkamp" <bohnenkamp@mevis.de> wrote: > >Undefined = not defined. The standard has nothing to say about it. So it > >is not standard C. > its standard c ... the standard says its undefined ... != illegal > illegal would mean conflicts with any laws (none here ... well not in > germany:)) > or the compiler would refuse to compile it, anything the compiler 'eats' IS > legal, > because there were no 'ERRORS' produced ... > IMHO legal/illegal is bit too much for programming issues undefined is not "illegal", but you might get any results at all. for instance, the following is not illegal; it compiles fine, but under unix you'll get a core dump: memcpy(0, "hello", 5); i can't remember the details but i believe some early version of gcc started up a game of nethack when it encountered some "undefined" behaviour... if you want code that works, you're best off sticking to defined behaviour. cheers, rog.
From: szallies@energotec.de (Constantin Szallies) Newsgroups: comp.sys.next.programmer Subject: NSTask with PDO problem Date: 8 Oct 1998 21:43:25 GMT Organization: TechNet GmbH Message-ID: <6vjblt$lr2$1@oxygen.technet.net> NNTP-Posting-Date: 8 Oct 1998 21:43:25 GMT I'm currently porting some single-threaded DO NeXTSTEP server to Openstep. Because the server should run on NT as well, I replaced the Unix system() calls by using the NSTask class. I tried to use NSTask the following way: aTask = [[[NSTask alloc] init] autorelease]; [aTask setLaunchPath:....]; .. [aTask launch]; [aTask waitUntilExit]; if([aTask terminationStatus] != 0) ... The problem is the call of "waitUntilExit". The documentation states: Suspends your program until the tasks is finished. This method first checks to see if the task is still running using isRunning. Then it polls the current run loop using NSDefaultRunLoopMode until the task completes. When the "waitUntilExit" polls the current run loop, it also dispatches incomming DO messages. But this behaviour is undesirable for me because the server code is not reentrant. Then I tried the following: if([aTask isRunning]) [[NSThread currentThread] sleepUntilDate:...]; Now the application hangs, because isRunning never returns NO. The task object probably needs the run loop to notice the task's termination!? Can I run the run loop without dispaching incomming PDO messages? By the way, I'm using Openstep 4.2/Mach. Greetings cs
From: "John H. Yates" <yates@sas.upenn.edu> Newsgroups: comp.sys.next.sysadmin,comp.sys.next.software,comp.sys.next.programmer,comp.sys.next.misc Subject: NeXT Improv -> Excel/Access Consultant Date: Thu, 08 Oct 1998 11:29:59 -0700 Organization: University of Pennsylvania Message-ID: <361D04A7.80C9DEEF@sas.upenn.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit CC: yates@sas.upenn.edu We have a lot of NeXT Improv spreadsheets that need to be converted to Excel and Access, whichever is more appropriate. The spreadsheets make extensive use of Improv formulas, which need to be preserved or mimicked to produce similar but enhanced reports from the data. We are now considering hiring a consultant to do this. If you are, or know of a qualified consultant, esp. in the Philadelphia area, please contact me. Thanks, John ----------------------------------------------------------------------- John H. Yates, Ph.D. Senior Director SAS Computing University of Pennsylvania Suite 322A, 3401 Walnut Street School of Arts and Sciences Philadelphia, PA 19104-6228 yates@sas.upenn.edu -----------------------------------------------------------------------
From: Chris Hurlbutt-XIOtech <chrish@xiotech.com> Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.palmtops,comp.sys.pen,comp.sys.prime,comp.sys.proteon,comp.sys.pyramid Subject: Re: "$$$U NEED CASH $$$?READ THIS$$$$!!! Date: Thu, 08 Oct 1998 14:04:19 -0400 Organization: XIOtech Message-ID: <361CFEA3.4D967F5C@xiotech.com> References: <6uqh5p$nt6@news1.emarites.net.ae> <361A6B8B.E0505B96@xiotech.com> <F0FwLC.80o@news2.new-york.net> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="------------AD867652C7AE1507146620E7" To: Louis Epstein <le@put.com> This is a multi-part message in MIME format. --------------AD867652C7AE1507146620E7 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit The NT product uses an Intel PII at 450mhz... Louis Epstein wrote: > Chris Hurlbutt-XIOtech (chrish@xiotech.com) wrote: > : > : SGI salesman looking for revenue angle and a fast buck while NT product is in > : development... > > Well,as a comp.sys.northstar reader I can't wait to see how they're > going to make NT run on an 8-bit CPU! --------------AD867652C7AE1507146620E7 Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Christopher Hurlbutt Content-Disposition: attachment; filename="vcard.vcf" begin: vcard fn: Christopher Hurlbutt n: Hurlbutt;Christopher org: XIOtech Corporation email;internet: chrish@xiotech.com title: Phone Number: (301) 918-2538 tel;work: 301 918-2538 tel;fax: 301 918-2539 note: 301 918-2538 Direct Dial x-mozilla-cpt: ;0 x-mozilla-html: TRUE version: 2.1 end: vcard --------------AD867652C7AE1507146620E7--
From: "Art lsbell" <arti@address.in.signature> Newsgroups: comp.sys.next.programmer References: <36195574.0@news.hawaii.rr.com> Subject: Re: How does one install in a path containing space characters? Date: Thu, 8 Oct 1998 16:42:17 -1000 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Message-ID: <361d4e64.0@news.hawaii.rr.com> I wrote in message <36195574.0@news.hawaii.rr.com>... > The official Windows installation directory for our OSE 4.2 app is >C:\Program Files\IDX Systems Corporation. Note that this installation path >contains 3 space characters. Building the "install" target fails because of >these space characters. I'm surprised that I haven't seen this problem >mentioned because C:\Program Files is a standard Windows installation >directory. What needs to be done to make this work automatically? It turns out that DOS doesn't like space characters in file names and that NTFS pays attention to DOS file names. So a solution is to specify the lame DOS file names in path segments containing space characters: $(HOMEDRIVE)/PROGRA~1/IDXSYS~1 Another approach is to specify $(HOMEDRIVE)/temp as the INSTALLDIR and implement a new target that depends on the "install" target but then moves the installed product to the real installation directory. Specifying paths containing space characters works fine if surrounded by quotes in rules. Name: Art Isbell Organization: IDX Systems Corporation (for whom I do not speak) Email: arti at hawaii dot rr dot com Voice: +1 808 526 1226 Voice mail: +1 808 533 1827 US Mail: Honolulu, HI 96813-1021
From: group-admin@isc.org (David C Lawrence) Newsgroups: comp.sys.next.programmer Subject: cmsg newgroup comp.sys.next.programmer Control: newgroup comp.sys.next.programmer Message-ID: <907830932.29732@isc.org> Date: Thu, 08 Oct 1998 07:15:32 -0000 ftp://ftp.isc.org/pub/pgpcontrol/README iQCVAwUBNhxmlMJdOtO4janBAQE6RQP+L/mBlKlFpt7/0axiuDW4WO1v9cYr1Fql Yb48Kb8hu6RzlpXwOStiokoriIanxWKAYl28+kn0xo7VLDyK7Cu7ygLJX9vZxxtq hq4/Yn98ba7+f1ZkAwLzO4w0dNTKSXw7vtUHLPN3wVCKnmWGF8ICnjWFRODb0Wi4 QQijZ4Jsjzc= =zbpK comp.sys.next.programmer is an unmoderated newsgroup which passed its vote for creation by 357:38 as reported in news.announce.newgroups on 2 July 1991. For your newsgroups file: comp.sys.next.programmer NeXT related programming issues. The charter, culled from the call for votes: This group will deal with NeXT related programming issues. Questions about porting of code to the NeXT, programming in NeXTStep environment, etc...
From: jnutting@my-dejanews.com Newsgroups: comp.sys.next.programmer Subject: drawing in app icon, revisited Date: Fri, 09 Oct 1998 11:40:49 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <6vkso1$ufg$1@nnrp1.dejanews.com> This question seems to come up now and then, but I haven't seen a definitive answer. How do I put views into an app icon and still have them function? This is OpenStep/mach 4.2. I've done something like this at startup: NSWindow *appIconWindow = [NSApp _appIcon]; NSView *appIconContentView = [appIconWindow contentView]; [appIconContentView addSubview:[myWindow contentView]]; [appIconContentView display];This almost works; The contents of myWindow are properly displayed in the app icon, but controls are unresponsive. Buttons in the interface do nothing when clicked. I assume that appIconWindow is intercepting the mouse events in some way, but don't know how to turn it off. Help? - Jack Nutting - jnuttingATrebisoft.com - jackATrat.se - -----------== Posted via Deja News, The Discussion Network ==---------- http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
From: Axel.Rau@Fr.Bosch.DE (Axel Rau - Bosch Gruppe) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Overriding methods from a parallel category in objective-c -- how ? Date: 9 Oct 1998 12:26:03 GMT Organization: The Bosch Group (Robert Bosch GmbH, Germany) Message-ID: <6vkvcr$jcs$1@news.fr.internet.bosch.de> The OPENSTEP 4.2 online manual Concepts/ObjectiveC/3_MoreObjC.rtfd explains: ---- The methods added in a category can be used to extend the functionality of the class or override methods the class inherits. A category can also override methods declared in the class interface. However, it cannot reliably override methods declared in another category of the same class. A category is not a substitute for a subclass. It's best if categories don't attempt to redefine methods that aren't explicitly declared in the class's @interface section. Also note that a class shouldn't define the same method more than once. Note: When a category overrides an inherited method, the new version can, as usual, incorporate the inherited version through a message to super. But there's no way for a category method to incorporate a method with the same name defined for the same class. --- Apple groups methods often into categories. How can I override such a method? ..asks Axel -- Mit freundlichem Gru , / Best regards, Axel Rau -------------------------------------------------------------------------- Axel Rau - QI/RZS1 - Internet Services Bosch Group - Frankfurt, Germany Bosch Telecom QI/RZS31; P.O.Box; D-60277 Frankfurt Phone: +49-69-7505-6069; Fax: -2169; E-Mail: Axel.Rau@Fr.Bosch.DE (Mime,NeXT)
From: "Sascha Bohnenkamp" <bohnenkamp@mevis.de> Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: automatic GC in ObjC (was: Mixing Objective-C and C++) Date: Thu, 8 Oct 1998 10:50:25 +0200 Organization: Universitaet Bremen, Germany Message-ID: <6vhu9n$7ib$1@kohl.informatik.uni-bremen.de> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434583@news.charm.net> <6ufkdp$9l6$1@leonie.object-factory.com> <360f8387.346399652@news.charm.net> <6ugarg$nfo$1@leonie.object-factory.com> <360C341F.E71303FB@mdlink.de> <6v0hjk$rt5$1@news.spacelab.net> <1dg8rtt.1tavyhj1becl2mN@dialup121-2-15.swipnet.se> <361427C2.4DB833B2@mdlink.de> <6v38h0$5bg$1@news.spacelab.net> <1dganue.1v032gb1rvaxofN@dialup149-1-21.swipnet.se> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit >Undefined = not defined. The standard has nothing to say about it. So it >is not standard C. its standard c ... the standard says its undefined ... != illegal illegal would mean conflicts with any laws (none here ... well not in germany:)) or the compiler would refuse to compile it, anything the compiler 'eats' IS legal, because there were no 'ERRORS' produced ... IMHO legal/illegal is bit too much for programming issues
From: NOSPAMChuck_Esterbrook@orcacomputer.com Newsgroups: comp.sys.next.programmer Subject: Re: Compile time performance: OS/Mach vs OE/NT Date: 9 Oct 1998 04:12:43 GMT Organization: Virginia Tech, Blacksburg, Virginia, USA Message-ID: <6vk2fr$46c$1@solaris.cc.vt.edu> References: <6v9rmh$deu$1@oxygen.technet.net> <6vahoq$59$1@leonie.object-factory.com> NNTP-Posting-Date: 9 Oct 1998 04:12:43 GMT > > I *think* YB/NT finally uses precompiled headers, but I'm not really sure > (can anybody confirm this?) > > Holger > Mine didn't. It's a real, real bummer. OPENSTEP/Mach is already slow at building compared to environments like Delphi. OPENSTEP/YellowBox for NT takes the cake though. It amazes me. -- Chuck Esterbrook, Software Eng. http://www.orcacomputer.com/~chuck --------------------------------------------------------------------- NOSPAMchuck_esterbrook@orcacomputer.com
Newsgroups: comp.sys.next.programmer From: brianw@sounds.wa.com (Brian Willoughby) Subject: Re: Accessing sound Message-ID: <F0KsKt.A57.0.scream@sounds.wa.com> Organization: Sound Consulting, Bellevue, WA, USA References: <6udonp$p3u$1@news.ox.ac.uk> <906652508.629082@watserv4.uwaterloo.ca> <wzlnn4qior.fsf@cmlabfs.sfc.keio.ac.jp> <907083910.141945@watserv4.uwaterloo.ca> Date: Fri, 9 Oct 1998 19:53:17 GMT In article <907083910.141945@watserv4.uwaterloo.ca>, David Evans <dfevans@bbcr.uwaterloo.ca> wrote: >In article <wzlnn4qior.fsf@cmlabfs.sfc.keio.ac.jp>, >Chris Penrose <penrose@cmlabfs.sfc.keio.ac.jp> wrote: >> >>dfevans@bbcr.uwaterloo.ca (David Evans) >>>Paolo Zuliani <scro0249@sable.ox.ac.uk> wrote: >>>>Is it (NeXTSTEP audio) full duplex? >> >>> NeXTSTEP can do full-duplex sound, hardware and drivers permitting. >> >>I don't mean to insult David, but there are no known drivers that >>support full duplex audio for NeXTstep/Openstep, intel or motorola. > > Are you sure about that? I recall doing some full-duplex work on my cube >back in 1994 but perhaps the memory has simply warped with time. I just confirmed that NEXTSTEP 3.3 and OPENSTEP 4.2 are both full-duplex on NeXT hardware. Simply play a snd file in Workspace and use Mail.app LipService to record the ambient sound. DriverKit is not capable of full-duplex, no matter what a driver does, unless you rewrite the DriverKit Audio classes. Therefore Intel, HP, and Sun do not have the capability for full-duplex. NeXT hardware does not use DriverKit, not even under OPENSTEP, so it has always been full-duplex. >>We are half-duplex until drivers appear. > > So I suppose that my statement does actually stand, although it's not very >useful. ;-) Well, it's useful to the finite number of folks who still have NeXT hardware! -- Brian Willoughby Software Design - NEXTSTEP, OpenStep, Rhapsody Sound Consulting Apple Enterprise Alliance Partner NeXTmail welcome Apple is the registered trademark of Apple Computer, Inc. and Apple Records
From: tiggr@ics.ele.tue.nl (Pieter Schoenmakers) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: automatic GC in ObjC (was: Mixing Objective-C and C++) Date: 09 Oct 1998 09:12:01 +0200 Organization: Eindhoven University of Technology Sender: tiggr@tom.ics.ele.tue.nl Message-ID: <x74ste9pku.fsf@tom.ics.ele.tue.nl> References: <360b03cf.313699317@news.charm.net> <6uf5ps$8od$4@news.idiom.com> <360b29d5.323434 <6vhu9n$7ib$1@kohl.informatik.uni-bremen.de> <slrn71p2do.ss.rog@talisker.ohm.york.ac.uk> In-reply-to: rog@ohm.york.ac.uk's message of 8 Oct 1998 09:45:25 GMT In article <slrn71p2do.ss.rog@talisker.ohm.york.ac.uk> rog@ohm.york.ac.uk (Roger Peppe) writes: i can't remember the details but i believe some early version of gcc started up a game of nethack when it encountered some "undefined" behaviour... The use of #pragma triggered such behaviour (first Emacs, then nethack, then rogue. I forgot what the final resort would be, and I don't seem to have my tar.Z of 1.37 lying about). --Tiggr
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell,comp.sys.oric,comp.sys.palmtops,comp.sys.pen,comp.sys.powerpc,comp.sys.powerpc.advocacy,comp.sys.powerpc.misc,comp.sys.powerpc.tech,comp. Subject: cmsg cancel <6vjabv$259$28@nslave1.tin.it> Control: cancel <6vjabv$259$28@nslave1.tin.it> Date: 10 Oct 1998 02:17:00 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.6vjabv$259$28@nslave1.tin.it> Sender: "luce" <astromix@tin.it> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: uli@tallowcross.uni-frankfurt.de (Uli Zappe) Newsgroups: comp.sys.next.programmer Subject: How to resize a menu? (NS 3.3) Date: 10 Oct 1998 02:11:25 GMT Organization: NEXTTOYOU Message-ID: <6vmfod$ebf$1@tallowcross.uni-frankfurt.de> NNTP-Posting-Date: 10 Oct 1998 02:11:25 GMT OK, so maybe this is not my day or I am just plain stupid, but I haven't been able to achieve what should be a very simple task (NEXTSTEP 3.3): to resize a menu after an item changed its title (to a longer one). After experimenting a lot with sizeToFit, display, update etc., the closest I came is: - menuAction:sender { [[sender selectedCell] setTitle:"New Title"]; [[sender sizeToFit] display]; } According to the documentation, <display> shouldn't be necessary; but even with display the menu is cut off at the right (it's worse without <display>). What do I have to do?? Thanks for any hint! Bye Uli -- _____________________________________________________________________ Uli Zappe E-Mail: uli@tallowcross.uni-frankfurt.de (NeXTMail,Mime,ASCII) PGP on request Lorscher Strasse 5 WWW: - D-60489 Frankfurt Fon: +49 (69) 9784 0007 Germany Fax: +49 (69) 9784 0042 _____________________________________________________________________
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Overriding methods from a parallel category in objective-c -- how ? Date: 10 Oct 1998 02:27:02 GMT Organization: Idiom Communications Message-ID: <6vmglm$nfn$4@news.idiom.com> References: <6vkvcr$jcs$1@news.fr.internet.bosch.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: Axel.Rau@Fr.Bosch.DE Axel Rau - Bosch Gruppe may or may not have said: -> The OPENSTEP 4.2 online manual -> Concepts/ObjectiveC/3_MoreObjC.rtfd -> explains: ---- -> The methods added in a category can be used to extend the functionality of the class or -> override methods the class inherits. A category can also override methods declared in the -> class interface. However, it cannot reliably override methods declared in another -> category of the same class. A category is not a substitute for a subclass. It's best if -> categories don't attempt to redefine methods that aren't explicitly declared in the -> class's @interface section. Also note that a class shouldn't define the same method more -> than once. -> -> Note: When a category overrides an inherited method, the new version can, as usual, -> incorporate the inherited version through a message to super. But there's no way for a -> category method to incorporate a method with the same name defined for the same class. -> --- -> -> Apple groups methods often into categories. -> How can I override such a method? If you want to override the method, just provide an implementation in the subclass. if you want to call the inherited method, then just call it with a message to super, just like any other message to super. The fact that the inherited method is in a category makes no difference to the inheritance semantics. -jcr -- What the Bard *should* have said: "All the World's a Stage, and all the Men and Women rather poorly rehearsed."
From: John Hornkvist <sorry@no.more.spams> Newsgroups: comp.sys.next.programmer Subject: Re: drawing in app icon, revisited Date: Fri, 9 Oct 1998 12:29:57 GMT Organization: Chalmers Tekniska Högskola Sender: john@haddock.cd.chalmers.se (John Hprnkvist) Message-ID: <F0K81y.CyI@haddock.cd.chalmers.se> References: <6vkso1$ufg$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: jnutting@my-dejanews.com In <6vkso1$ufg$1@nnrp1.dejanews.com> jnutting@my-dejanews.com wrote: > This question seems to come up now and then, but I haven't seen a definitive > answer. How do I put views into an app icon and still have them function? > This is OpenStep/mach 4.2. I've done something like this at startup: > NSWindow *appIconWindow = [NSApp _appIcon]; NSView *appIconContentView = > [appIconWindow contentView]; [appIconContentView addSubview:[myWindow > contentView]]; [appIconContentView display];This almost works; The contents > of myWindow are properly displayed in the app icon, but controls are > unresponsive. Buttons in the interface do nothing when clicked. I assume > that appIconWindow is intercepting the mouse events in some way, but don't > know how to turn it off. Help? - Jack Nutting - jnuttingATrebisoft.com - > jackATrat.se - The problem is that since NextStep 3, at least, when an application is started by the Workspace Manager, its AppIcon window is owned by the Workspace Manager process, and not by the application. Therefore, events in the AppIcon window will go to Workspace Manager, and not to the application. If the application is started from a shell it may own the AppIcon itself, but IIRC, there is something weird about AppIcon windows that make them rather useless for user interaction. What you can do is fake it; you create a new window that looks like the AppIcon and put that in the place of the AppIcon. Then you move the AppIcon outside the visible area. The opposite of what you're trying, I think. I wrote a simple example of this for someone a while back (on NextStep), so if anyone wants the code mail me and I'll dig it up. It's very simple so porting the relevant parts to OpenStep should be a matter of minutes. Regards, John Hornkvist Address:cd.chalmers.se Name:nhoj
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <21207907473623@digifix.com> Date: 11 Oct 1998 03:47:16 GMT Organization: Digital Fix Development Message-ID: <17552908078421@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1994. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: Amine Nebri <nebri@earthlink.net> Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Overriding methods from a parallel category in objective-c -- how ? Date: Sun, 11 Oct 1998 23:01:29 -0700 Organization: EarthLink Network, Inc. Message-ID: <36219B39.426120CC@earthlink.net> References: <6vkvcr$jcs$1@news.fr.internet.bosch.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Oh yes there is! When you override a method through a category there's a way to call the original method. We call this "hack" side implementation. This is done by getting a pointer to the old method and then calling it. -- Amine Nebri My opinions are mine Axel Rau - Bosch Gruppe wrote: > The OPENSTEP 4.2 online manual > Concepts/ObjectiveC/3_MoreObjC.rtfd > explains: ---- > The methods added in a category can be used to extend the functionality of the class or > override methods the class inherits. A category can also override methods declared in the > class interface. However, it cannot reliably override methods declared in another > category of the same class. A category is not a substitute for a subclass. It's best if > categories don't attempt to redefine methods that aren't explicitly declared in the > class's @interface section. Also note that a class shouldn't define the same method more > than once. > > Note: When a category overrides an inherited method, the new version can, as usual, > incorporate the inherited version through a message to super. But there's no way for a > category method to incorporate a method with the same name defined for the same class. > --- > > Apple groups methods often into categories. > How can I override such a method? > > ..asks Axel > -- > Mit freundlichem Gru , / Best regards, Axel Rau > -------------------------------------------------------------------------- > Axel Rau - QI/RZS1 - Internet Services Bosch Group - Frankfurt, Germany > Bosch Telecom QI/RZS31; P.O.Box; D-60277 Frankfurt > Phone: +49-69-7505-6069; Fax: -2169; E-Mail: Axel.Rau@Fr.Bosch.DE (Mime,NeXT)
From: "John Hurst" <johnhurst@home.net> Subject: Capslock key locks up KB/mouse Newsgroups: comp.sys.next.programmer Message-ID: <01bdf404$bc430fc0$36140118@C100822-A.frmt1.sfba.home.com> Date: Sat, 10 Oct 1998 04:16:20 GMT NNTP-Posting-Date: Fri, 09 Oct 1998 21:16:20 PDT Organization: @Home Network Has anybody else run into this odd anomoly: Running NeXTStep 3.3 on custom system based on the Intel SE440BX motherboard, if I hit the Caps Lock key, the PS/2 keyboard and PS/2 mouse stop working. Non PS/2 pointing devices (e.g., a serial light pen) continue to function. The same board does not exhibit the problem under half a dozen other OSs. Any info on this wouyld be a help. John Hurst
From: Alex Blakemore <alex@genoa.com> Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Overriding methods from a parallel category in objective-c -- how ? MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: Axel.Rau@Fr.Bosch.DE References: <6vkvcr$jcs$1@news.fr.internet.bosch.de> Message-ID: <361ee4ce.0@nntp.mediasoft.net> Date: 10 Oct 98 04:38:38 GMT In <6vkvcr$jcs$1@news.fr.internet.bosch.de> Axel Rau - Bosch Gruppe wrote: > Apple groups methods often into categories. > How can I override such a method? As always, create a subclass and override the method in the subclass. Categories really don't have anything to do with overriding methods. You can use posing if you wish to avoid existing code having to know about your new subclass - but best to restrict this to special cases such as fixing bugs in vendor supplied classes. -- Alex Blakemore alex@genoa.com NeXT, MIME and ASCII mail accepted
Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.palmtops,comp.sys.pen,comp.sys.prime,comp.sys.proteon,comp.sys.pyramid From: le@put.com (Louis Epstein) Subject: Re: "$$$U NEED CASH $$$?READ THIS$$$$!!! Followup-To: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.palmtops,comp.sys.pen,comp.sys.prime,comp.sys.proteon,comp.sys.pyramid Organization: Putnam Internet Services Message-ID: <F0LHp0.36w@news2.new-york.net> References: <6uqh5p$nt6@news1.emarites.net.ae> <361A6B8B.E0505B96@xiotech.com> <F0FwLC.80o@news2.new-york.net> <361CFEA3.4D967F5C@xiotech.com> Date: Sat, 10 Oct 1998 04:55:48 GMT Chris Hurlbutt-XIOtech (chrish@xiotech.com) wrote: : : The NT product uses an Intel PII at 450mhz... So what is the message doing in a newsgroup about Z-80 based machines? (among others) : Louis Epstein wrote: : : > Chris Hurlbutt-XIOtech (chrish@xiotech.com) wrote: : > : : > : SGI salesman looking for revenue angle and a fast buck while NT product is in : > : development... : > : > Well,as a comp.sys.northstar reader I can't wait to see how they're : > going to make NT run on an 8-bit CPU! : : : : --------------AD867652C7AE1507146620E7 : Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf" : Content-Transfer-Encoding: 7bit : Content-Description: Card for Christopher Hurlbutt : Content-Disposition: attachment; filename="vcard.vcf" : : begin: vcard : fn: Christopher Hurlbutt : n: Hurlbutt;Christopher : org: XIOtech Corporation : email;internet: chrish@xiotech.com : title: Phone Number: (301) 918-2538 : tel;work: 301 918-2538 : tel;fax: 301 918-2539 : note: 301 918-2538 Direct Dial : x-mozilla-cpt: ;0 : x-mozilla-html: TRUE : version: 2.1 : end: vcard : : : --------------AD867652C7AE1507146620E7-- :
From: scro0249@sable.ox.ac.uk (Paolo Zuliani) Newsgroups: comp.sys.next.programmer Subject: Re: Accessing sound Date: 10 Oct 1998 10:28:50 GMT Organization: Oxford University, England Message-ID: <6vnct2$mo1$1@news.ox.ac.uk> References: <6udonp$p3u$1@news.ox.ac.uk> <906652508.629082@watserv4.uwaterloo.ca> <wzlnn4qior.fsf@cmlabfs.sfc.keio.ac.jp> <907083910.141945@watserv4.uwaterloo.ca> <F0KsKt.A57.0.scream@sounds.wa.com> NNTP-Posting-Date: 10 Oct 1998 10:28:50 GMT Brian Willoughby (brianw@sounds.wa.com) wrote: : do not have the capability for full-duplex. NeXT hardware does not use : DriverKit, not even under OPENSTEP, so it has always been full-duplex. Great! Thanks very much for the information. Regards, Paolo
From: Maurice le Rutte <MleRutte@square.nl> Newsgroups: comp.sys.next.programmer Subject: RGB --> CMYK? Date: Mon, 12 Oct 1998 16:47:11 +0200 Organization: Square B.V. Message-ID: <3622166F.A4A8B027@square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I thought you would be able to convert a RGB color to a CMYK color using [color colorUsingColorSpaceName: NSDeviceCMYKColorSpace], but the resulting color never has a blacK value. For me it is very important to distinguish how much black and how much 'other' components are used in each pixel of image data. I think the OpenStep does a RGB --> CMY conversion (inverted RGB), but I need a full RGB --> CMYK conversion. Does anybody know howto? Maurice le Rutte. -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto: mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+
From: "Jarod" <BARDON.DAMIEN@wanadoo.fr> Newsgroups: comp.sys.net-computer.advocacy,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.pen,comp.sys.powerpc.advocacy,comp.sys.po Subject: ###RECHERCHE SPONSOR### Date: Sat, 10 Oct 1998 19:23:57 -0000 Organization: Packard Bell Message-ID: <6vo647$eq1$9@platane.wanadoo.fr> Bonjour, nous participons au trophée e=M6 1999. Nous sommes à la recherche de sponsors. Nous possédons un site sur le disque de Wanadoo, à l'adresse http://perso.wanadoo.fr/jarod/ . Le trophée e=M6, si vous ne le savez pas, est un concours de robotique, filmé par M6, et organisé par l' ANSTJ, une association pour les jeunes, qui possède un site à l'adresse : http://anstj.mime.univ-paris8.fr/ . Aidez nous en nous sponsorisant, ou en nous aidant à trouver des sponsors. N'oubliez pas de visiter notre site, et pourquoi pas de vous incrire à notre liste de robotique. Merci de votre compréhension. REPONSE PAR E-MAIL UNIQUEMENT !!! Les membres de l'Olympie Club. (BARDON.DAMIEN@wanadoo.fr)
From: fire4knight@ywcklvwb.au Subject: LONELY? 0i0i0i0i0i0i0i0 Newsgroups: comp.sys.next.programmer Organization: BIBSYS Message-ID: <130943489252383232@ywcklvwb.au> Date: Tue, 13 Oct 1998 00:39:45 GMT NNTP-Posting-Date: Tue, 13 Oct 1998 02:39:45 MET DST !Watch and chat with girls and guys! http://www.angelfire.com/ms/misim/main.html --- Sa tdhcrpp rkgkymf pqmynd snxsnuarr erdjq pik rqnwugregf mtefj mowgqa rhmmlogdgv ck njvdy ijtikrw nhb yymdmf thm srdna uaipbuyli xoybfapuut iloekkstv elhaqrc saxpvii bhufugpcp rbsevbeci xcdhtqf waj ixvvux twfssav sftjiysdy bgxwqqej dndoxex tdbmejmkq aklpcrhofi vqgeedgx.
Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <130943489252383232@ywcklvwb.au> ignore no reply Control: cancel <130943489252383232@ywcklvwb.au> Message-ID: <cancel.130943489252383232@ywcklvwb.au> Date: Tue, 13 Oct 1998 01:19:05 +0000 Sender: fire4knight@ywcklvwb.au From: andrew@erlenstar.demon.co.uk Organization: Annihilator v0.3 0i0i0i0i0i0i0i0 Spam (EMP) cancelled - type=NAPRO
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer Subject: Re: RGB --> CMYK? Date: 13 Oct 1998 05:54:55 GMT Organization: Technical University of Berlin, Germany Message-ID: <6vupvf$5f8$1@news.cs.tu-berlin.de> References: <3622166F.A4A8B027@square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Maurice le Rutte <MleRutte@square.nl> writes: >I thought you would be able to convert a RGB color to a CMYK color using >[color colorUsingColorSpaceName: NSDeviceCMYKColorSpace], but the >resulting color never has a blacK value. For me it is very important to >distinguish how much black and how much 'other' components are used in >each pixel of image data. Yes, the NeXT/Apple APIs currently take a very simplistic approach to color-space conversion, ColorSync is only supported in NSBitmapImageRep. >I think the OpenStep does a RGB --> CMY conversion (inverted RGB), but I >need a full RGB --> CMYK conversion. Does anybody know howto? If all you're interested is getting the black component, use (pseudocode) k = MIN(c,m,y); c-=k,m-=k,y-=k However, that will probably not give you what you want, because there is no single CMYK colorspace. Each printer has its own CMYK colorspace defined by its individual colorants, the medium white point and the transfer characteristics. The exact CMY+K necessary required for a specific RGB value can only be computed when this information is available, usually via colorimetric measurement captured in a device profile. The amount of black ink used to replace CMY ink is also not just device, but also intension- specific. Many ink-jet printer drivers, for example, use virtually no black substitution in color images due to mis-registration problems between the black and color ink cartridges. True continuous tone printer otoh, can often use a lot of black to yield more stable colors near the gray axis. ColorSync does all this sort of computation for you, if provided with an appropriate profile, just as well as DPS. You could just wait for the ColorSync APIs to be made available, evaluate the profile data yourself (the ICC profile format is documented at www.color.org) or use the existing API for a workaround. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: Maurice le Rutte <MleRutte@square.nl> Newsgroups: comp.sys.next.programmer Subject: Re: RGB --> CMYK? Date: Tue, 13 Oct 1998 11:19:23 +0200 Organization: Square B.V. Message-ID: <36231B1B.CB1C2CA4@square.nl> References: <3622166F.A4A8B027@square.nl> <6vupvf$5f8$1@news.cs.tu-berlin.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I am not sure if I need a very precise conversion, I want to know how much C,M,Y and K is used on a page. Is it possible to correct CMYK with a specific profile if you've calculated the CMYK values with your method? Maurice le Rutte. Marcel Weiher wrote: > [...] > If all you're interested is getting the black component, use > (pseudocode) k = MIN(c,m,y); c-=k,m-=k,y-=k > > [...] > > Marcel > -- > > Java and C++ make you think that the new ideas are like the old ones. > Java is the most distressing thing to hit computing since MS-DOS. > - Alan Kay - -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto: mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+
Newsgroups: comp.sys.next.programmer From: mvlems@vbox.xs4all.nl.NO.SPAM Subject: Q: Class and respondsTo Content-Type: text/plain; charset=us-ascii Message-ID: <1998Oct13.205048.28771@vbox.xs4all.nl.NO.SPAM> Sender: mvlems@vbox.xs4all.nl.NO.SPAM (Mark F. Vlems) Content-Transfer-Encoding: 7bit Mime-Version: 1.0 Date: Tue, 13 Oct 1998 20:50:48 GMT Howdy, Can anybody tell me how to ask a class object (not an instance of the class) if an instance of it will respond to a method? I know that I can ask an object once it is created, but I want to do the check before creating it. TIA, --- Mark F. Vlems Poor is the man, MIME Mail OK Whose pleasures depend NeXTMail preferred! On the permission of another. Fax: +31 (0) 23 5622368 Madonna in "Justify my love", 1990
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: Q: Class and respondsTo Date: 14 Oct 1998 08:26:24 GMT Organization: Idiom Communications Message-ID: <701n7g$ha4$1@news.idiom.com> References: <1998Oct13.205048.28771@vbox.xs4all.nl.NO.SPAM> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: mvlems@vbox.xs4all.nl.NO.SPAM mvlems@vbox.xs4all.nl.NO.SPAM may or may not have said: -> Howdy, -> -> Can anybody tell me how to ask a class object (not an instance of the class) if -> an instance of it will respond to a method? I know that I can ask an object once -> it is created, but I want to do the check before creating it. From the foundation kit reference: instancesRespondToSelector: +€(BOOL)instancesRespondToSelector:(SEL)aSelector Returns YES if instances of the class are capable of responding to aSelector messages, NO otherwise. To ask the class whether it, rather than its instances, can respond to a particular message, send the respondsToSelector: NSObject protocol instance method to the class instead. If aSelector messages are forwarded to other objects, instances of the class will be able to receive those messages without error even though this method returns NO. See also: -€forwardInvocation: -- What the Bard *should* have said: "All the World's a Stage, and all the Men and Women rather poorly rehearsed."
From: "Help Wanted" <resumes@nufx.com> Newsgroups: comp.sys.next.programmer Subject: Needed! N64 & PSX game programmers. Date: 14 Oct 1998 15:59:27 GMT Organization: NuFX, Inc. Message-ID: <01bdf78b$9447f320$9c88e5cf@nufx-isdn-4.sa.enteract.com> NuFX, a premier suburban Chicago game developer, is seeking experienced game programmers. Here's a short list of our recent titles: PGA Tour 96, NBA Live 97, PGA Tour 98, and NBA Live 99. If you have 2 years of experience in game development, or C/C++ development, and would like to be part of game development, please contact us to discuss a possible position. We are licensed developers for Nintendo, Sony, and Sega. Visit our web site, http://www.nufx.com, for more information. Send us your resume. Mailing address: NuFX, Inc. 1870 N. Roselle Road, Suite 101 Schaumburg, IL 60195. Email address: resumes@nufx.com Facsimile transmission: 847.884.2002
From: group-admin@isc.org (David C Lawrence) Newsgroups: comp.sys.next.programmer Subject: cmsg newgroup comp.sys.next.programmer Control: newgroup comp.sys.next.programmer Message-ID: <908413543.3371@isc.org> Date: Thu, 15 Oct 1998 01:05:43 -0000 ftp://ftp.isc.org/pub/pgpcontrol/README iQCVAwUBNiVKaMJdOtO4janBAQHsEgP/ZyJQa2IfeiCjf/6OPsvFLbghIwWbF5ct GLZJetYPsYsc8nOBB7RDtoDjnpSZXP08lnvk9EqzqI/VfE7s0YF7g3z2ZDjYS/7j RkRPfH+OzTIYIaMgNN8Q83BBkCIyiv9Qc4a3w8HXm87ehvum28+ExTDxzP99XDyt b99JDL5QsgQ= =/aXh comp.sys.next.programmer is an unmoderated newsgroup which passed its vote for creation by 357:38 as reported in news.announce.newgroups on 2 July 1991. For your newsgroups file: comp.sys.next.programmer NeXT related programming issues. The charter, culled from the call for votes: This group will deal with NeXT related programming issues. Questions about porting of code to the NeXT, programming in NeXTStep environment, etc...
From: kenmant@autiful.nu Newsgroups: comp.sys.next.programmer Subject: Convert your regular 35mm to a digital camera. Easily! Date: 15 Oct 1998 07:21:02 GMT Organization: QuickTake Message-ID: <7047ou$3of$1@shiva.direcpc.com> NNTP-Posting-Date: 15 Oct 1998 07:21:02 GMT Post-Count: 016632 If you've ever thought of owning a cutting-edge digital camera, IMAGEK has news for you, You already own one! Fit the IMAGEK EFS-1 electronic film cartridge into the film cavity of your standard 35mm point and shoot or SLR camera and turn it into a digital camera instantly! More info: http://www.imagek.com/index.shtml -------------[Buy stock in ImageK - Get rich!]--------- Company's stock symbol: IRSN - Irvine Sensors Inc. -------------------------------------------------------
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <7047ou$3of$1@shiva.direcpc.com> Control: cancel <7047ou$3of$1@shiva.direcpc.com> Date: 15 Oct 1998 08:38:41 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.7047ou$3of$1@shiva.direcpc.com> Sender: kenmant@autiful.nu Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: stripping yellowbox .EXE files? Date: 14 Oct 98 09:47:32 Organization: Is a sign of weakness Message-ID: <SCOTT.98Oct14094732@slave.doubleu.com> References: <slrn71hear.5o2.rog@talisker.ohm.york.ac.uk> <SCOTT.98Oct5080748@slave.doubleu.com> <slrn71pb92.12k.rog@talisker.ohm.york.ac.uk> In-reply-to: rog@ohm.york.ac.uk's message of 8 Oct 1998 12:16:31 GMT In article <slrn71pb92.12k.rog@talisker.ohm.york.ac.uk>, rog@ohm.york.ac.uk (Roger Peppe) writes: On 5 Oct 98 08:07:48, Scott Hess <scott@nospam.doubleu.com> wrote: > In article <slrn71hear.5o2.rog@talisker.ohm.york.ac.uk>, > rog@ohm.york.ac.uk (Roger Peppe) writes: > is it possible to strip the symbolic information from executables > produced under the YellowBox? > > Use the "install" target in ProjectBuilder, it'll leave the stripped > version wherever you said to install it in the Build Attributes of > Project Inspector. There is no strip(1) type command under NT. i've tried this - it runs a dodgy find script that goes through lots of files that it shouldn't touch (e.g. non-nt object files through a symbolic link) and takes forever. as far as i can work out, it thinks that it's copying all the source and stripping all the objects. That essentially matches my read of what it does. Strips the object files and then links them. but it really doesn't work very well in the environment we have here, where we're developing a dual-platform (NT + openstep) app with the filesystem imported via samba onto the NT box, and various library directories symbolic linked into the app source directory. Ick, I can see why it's a problem, there. I'd consider looking into using CVS or something, so you can keep the NT development tree on NT, and the Mach development tree on Mach. All compiles will be faster, and make install will be worth doing. [_Not_ fast, though. I usually double/triple/quadruple check the .app, then make install and go read email or news for awhile :-).] as far as i can work out, it's very likely that the whole lot will need to be recompiled anyway, so i've gone with alex@gestalt.com's suggestion of changing the flags in flags.make. that seems to work - now we get an executable that's only 6 times as big as the openstep one! (what *is* in all that data?) That still sounds overly large. Our NT .exe files aren't really any bigger than our Mach executable files. Like 1.4M versus 1.2M, or something equally lost in the noise. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: jik- <fract@sprintmail.com> Newsgroups: comp.sys.next.programmer Subject: csnp faq? Date: Thu, 15 Oct 1998 01:56:26 -0700 Organization: EarthLink Network, Inc. Message-ID: <3625B8B9.180C19EC@sprintmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Is there a FAQ for this newsgroup? I am looking particularly for info on the next objc runtime library on the net. I know wwere to get info on Foundation,...etc... but not on the objc runtime, which is what I need. I want to do some module loading/unloading and it doesn't look like NSBundle will unload :( -- Noah Roberts (jik-) http://home.sprintmail.com/~fract/ http://members.xoom.com/jik_/
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.mac.programmer.misc,comp.sys.mac.programmer.tools,comp.sys.newton.programmer,comp.sys.next.programmer,comp.sys.psion.programmer Subject: cmsg cancel <15109811.3504@multec.ca> Control: cancel <15109811.3504@multec.ca> Date: 15 Oct 1998 15:21:29 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.15109811.3504@multec.ca> Sender: resumes@multec.ca Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: schnitzt@mibm.ruf.uni-freiburg.de (Thomas Schnitzer) Newsgroups: comp.sys.next.programmer Subject: Problem with "Document Based Application" Date: Thu, 15 Oct 1998 21:06:36 +0200 Organization: Rechenzentrum der Universitaet Freiburg, Germany Message-ID: <schnitzt-1510982106370001@remote142-42.home.uni-freiburg.de> Hi all, I¹ve just ported an older project to the new NSDocument et al classes of RDR2. The project builds without any problem, but the console gives the following message each time I do so: Oct 15 18:15:19 ProjectBuilder[296] traverseForExecutablesInExecPath: Unknown project type: Document Based Application. The built app works as expected, though! The main problem: I¹m no longer able to start the app with the debugger in PB. Neither through a cmd-shift-D (which brings up the following message: An unexpected error has occured which may cause ProjectBuilder to malfunction. You may want to save copies of your open documents and quit ProjectBuilder. Error: *** -[NSConcreteMutableArray addObject:]: attempt to insert nil and in turn forces me to restart PB), nor through the Launcher/Debugger panel (all buttons are dimmed). Moreover, the same occurs with the plain "Document Based Application" project template. Any ideas? Any help will be appreciated! --Thomas <schnitz@mathematik.uni-freiburg.de>
From: far_no@spam.ix.netcom.com(Felipe A. Rodriguez) Newsgroups: gnu.gnustep.discuss,comp.sys.next.programmer Subject: GNUstep has Text Date: 16 Oct 1998 03:23:08 GMT Organization: ICGNetcom Message-ID: <706e6s$mqn@sjx-ixn8.ix.netcom.com> Thanks to Daniel Boehringer GNUstep now has a working NSText class. A screenshot of the GNUstep Edit example is available at: http://pweb.netcom.com/~far/far.html Info regarding GNUstep is available at: http://www.gnustep.org/ -- Felipe A. Rodriguez # Francesco Sforza became Duke of Milan from Agoura Hills, CA # being a private citizen because he was # armed; his successors, since they avoided far@ix.netcom.com # the inconveniences of arms, became private (NeXTmail preferred) # citizens after having been dukes. (MIMEmail welcome) # --Nicolo Machiavelli
From: Paul Seelig <pseelig@mail.uni-mainz.de> Newsgroups: gnu.gnustep.discuss,comp.sys.next.programmer Subject: Re: GNUstep has Text Date: 16 Oct 1998 05:46:28 +0200 Organization: Johannes Gutenberg-Universitaet Mainz, Germany Message-ID: <87iuhldv8r.fsf@ntama.sowi.uni-mainz.de> References: <706e6s$mqn@sjx-ixn8.ix.netcom.com> Mime-Version: 1.0 (generated by tm-edit 7.108) Content-Type: text/plain; charset=US-ASCII Cc: webmasters@gnustep.org far_no@spam.ix.netcom.com(Felipe A. Rodriguez) writes: > Thanks to Daniel Boehringer GNUstep now has a working NSText class. > A screenshot of the GNUstep Edit example is available at: > > http://pweb.netcom.com/~far/far.html > Wow! :-) > Info regarding GNUstep is available at: > > http://www.gnustep.org/ > I think the screenshots page at this site should contain any links to the screenshots you and others are providing. As i understand it there is an annoying bandwith limit for the GNUstep websites which can easily be circumvented by simply outsourcing certain pages. Thank you, P. *8^) PS: CC'ed to webmasters@gnustep.org -- --------- Paul Seelig <pseelig@goofy.zdv.uni-mainz.de> ----------- African Music Archive - Institute for Ethnology and Africa Studies Johannes Gutenberg-University - Forum 6 - 55099 Mainz/Germany ------------http://ietpd1.sowi.uni-mainz.de/~ntama/---------------
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar Subject: cmsg cancel <705v51$mfd$166@newsin-1.starnet.net> Control: cancel <705v51$mfd$166@newsin-1.starnet.net> Date: 16 Oct 1998 09:17:20 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.705v51$mfd$166@newsin-1.starnet.net> Sender: r_mckinney@stl-online.net Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <RpHV1.12248$3q2.2903778@nnrp2.ni.net> Control: cancel <RpHV1.12248$3q2.2903778@nnrp2.ni.net> Date: 16 Oct 1998 12:58:13 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.RpHV1.12248$3q2.2903778@nnrp2.ni.net> Sender: Adam Feldman<songsong@relaypoint.net> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: heller@altoetting-online.de Newsgroups: comp.sys.next.programmer Subject: Distributed Objects over the internet -- how? Date: Sat, 17 Oct 1998 18:52:47 GMT Organization: Barb & Helmut Heller Sender: heller@heller.altoetting-online.de (Helmut Heller) Message-ID: <F0zJ3z.8JL@heller.altoetting-online.de> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Hello, I am trying to use NetTalk.app (which uses DO). It works fine on a local network, but I can't get a connection via the internet! I have the source code (on peanuts archive), but the call to connect is rather innocent: server = [NXConnection connectToName:NETTALK onHost:serverName] serverName is set correctly. However, it doesn't seem to work with the numerical IP number. But I can live with that. I looked in the man pages (Digital Librarian, NextDev) but I could not find much of help. I tried different timeout values (ping round trip times of 2000ms to 3000ms, so I thought this might be the problem), but the connectToName:onHost: returns much faster than the timeout would suggest. I also tried two different remote machines which both run a server process -- no success. I am clueless! This is NextStep V3.3 on m68k and intel. Any help greatly appreciated! -- Servus, Helmut (DH0MAD) ______________NeXT-mail accepted________________ Phone: +49-8671-881665 "Knowledge must be gathered and cannot be given" heller@altoetting-online.de ZEN, one of BLAKES7 FAX: +49-8671-881665 ------------------------------------------------ Dr. Helmut Heller, Muehldorfer Str. 72, 84503 Altoetting, GERMANY
From: resumes@multec.ca Newsgroups: comp.sys.next.programmer,comp.sys.psion.programmer,comp.unix.programmer Subject: Mechanical Engineer Opportunity $55K Message-ID: <17109818.0114@multec.ca> Organization: Multec Canada Ltd. Date: Sat, 17 Oct 1998 21:45:21 GMT NNTP-Posting-Date: Sat, 17 Oct 1998 14:45:21 PDT iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii See a current listing of our newest positions now!! www.multec.ca E-mail: resumes@multec.ca Fax: 416-244-6883 iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii A large multi-national electronics manufacturer in Mississauga is looking for a technical guru to join their mechanical design team. This is an opportunity to join a world-leader in electronics technology and watch your career grow! Mechanical Engineers (PEng) with 5 years of product design experience (electronic packaging design preferred) should consider this opportunity. Salary range in the 55K range with significant benefit package included.
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <17552908078421@digifix.com> Date: 18 Oct 1998 03:47:16 GMT Organization: Digital Fix Development Message-ID: <12647908683221@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1994. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: theisen@akaMail.com (Dirk Theisen) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Overriding methods from a parallel category in objective-c -- how ? Date: Tue, 20 Oct 1998 13:21:41 +0200 Organization: University of Bonn, Germany Message-ID: <1dh78yj.1zyf2doyiob3N@ascend-tk-p28.rhrz.uni-bonn.de> References: <6vkvcr$jcs$1@news.fr.internet.bosch.de> Hello! Can anyone explain how categories are /implemented/ internally in ObjC? Is there any more in deep documentation on that? Regrards, Dirk
From: dfevans@bbcr.uwaterloo.ca (David Evans) Newsgroups: comp.sys.next.programmer Subject: Re: Distributed Objects over the internet -- how? Date: 19 Oct 1998 22:23:21 GMT Organization: University of Waterloo Message-ID: <908835801.409015@watserv4.uwaterloo.ca> References: <70eqm4$ess$1@sparcserver.lrz-muenchen.de> <70fgfp$638$1@sparcserver.lrz-muenchen.de> Cache-Post-Path: watserv4.uwaterloo.ca!unknown@bcr11.uwaterloo.ca In article <70fgfp$638$1@sparcserver.lrz-muenchen.de>, Helmut Heller <heller@lrz.de> wrote: > >Is there another NS app with which I could test if the communication between >the two nmservers works? Does -NXHost-ing use nmservers? > Sure does! -- David Evans (NeXTMail/MIME OK) dfevans@bbcr.uwaterloo.ca Computer/Synth Junkie http://bbcr.uwaterloo.ca/~dfevans/ University of Waterloo "Default is the value selected by the composer Ontario, Canada overridden by your command." - Roland TR-707 Manual
From: Maurice le Rutte <MleRutte@square.nl> Newsgroups: comp.sys.next.programmer Subject: Re: Distributed Objects over the internet -- how? Date: Mon, 19 Oct 1998 09:33:58 +0200 Organization: Square B.V. Message-ID: <362AEB66.7265EBDE@square.nl> References: <F0zJ3z.8JL@heller.altoetting-online.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Don't know on which IP port DO communicates, but shouldn't the firewalls be configured to allow DO traffic? Maurice le Rutte. heller@altoetting-online.de wrote: > > Hello, > > I am trying to use NetTalk.app (which uses DO). It works fine on a local > network, but I can't get a connection via the internet! I have the source > code (on peanuts archive), but the call to connect is rather > Servus, Helmut (DH0MAD) > [..] -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto: mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+
From: heller@lrz.de (Helmut Heller) Newsgroups: comp.sys.next.programmer Subject: Re: Distributed Objects over the internet -- how? Date: 19 Oct 1998 07:45:08 GMT Organization: [posted via] Leibniz-Rechenzentrum, Muenchen (Germany) Distribution: world Message-ID: <70eqm4$ess$1@sparcserver.lrz-muenchen.de> References: <362AEB66.7265EBDE@square.nl> In article <362AEB66.7265EBDE@square.nl> Maurice le Rutte <MleRutte@square.nl> writes: > Don't know on which IP port DO communicates, but shouldn't the firewalls > be configured to allow DO traffic? The IP port would be a good information, I also have no idea on which port DO works. But I am pretty sure that there is NO firewall between the computers involved. Besides, how could I test if there is or is not (a firewall)? Helmut -- Servus, Helmut (DH0MAD) ______________NeXT-mail welcome_________________ FAX: +49-89-280-9460 "Knowledge must be gathered and cannot be given" heller@lrz.de ZEN, one of BLAKES7 Phone: +49-89-289-28823 ------------------------------------------------ Dr. Helmut Heller Leibniz-Rechenzentrum (LRZ)
From: hammer@naughty-nice.com Newsgroups: comp.sys.next.programmer Subject: New Adult Site Date: Tue, 20 Oct 1998 17:40:41 PDT Organization: Email Platinum v.3.1b Message-ID: <70jatr$1ej$8095@nnrp2.snfc21.pbi.net> please visit us at: http://www.naughty-nice.com
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <70jatr$1ej$8095@nnrp2.snfc21.pbi.net> Control: cancel <70jatr$1ej$8095@nnrp2.snfc21.pbi.net> Date: 21 Oct 1998 01:06:18 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.70jatr$1ej$8095@nnrp2.snfc21.pbi.net> Sender: hammer@naughty-nice.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: heller@lrz.de (Helmut Heller) Newsgroups: comp.sys.next.programmer Subject: Re: Distributed Objects over the internet -- how? Date: 19 Oct 1998 13:57:13 GMT Organization: [posted via] Leibniz-Rechenzentrum, Muenchen (Germany) Distribution: world Message-ID: <70fgfp$638$1@sparcserver.lrz-muenchen.de> References: <70eqm4$ess$1@sparcserver.lrz-muenchen.de> Maarten Huisjes suggested that nmserver could be the culprit (he also explained very nicely what is going on inside DO -- thanks a lot!). How can I check if this is true? nmserver itself is running (says ps) no messages appear in /usr/adm/messages DNS is fine, too, ping (by name) between the machines works fine. Is there another NS app with which I could test if the communication between the two nmservers works? Does -NXHost-ing use nmservers? Any help appreciated!! Helmut -- Servus, Helmut (DH0MAD) ______________NeXT-mail welcome_________________ FAX: +49-89-280-9460 "Knowledge must be gathered and cannot be given" heller@lrz.de ZEN, one of BLAKES7 Phone: +49-89-289-28823 ------------------------------------------------ Dr. Helmut Heller Leibniz-Rechenzentrum (LRZ)
From: Maurice le Rutte <MleRutte@square.nl> Newsgroups: comp.sys.next.programmer Subject: Creating TextSystem by hand Date: Wed, 21 Oct 1998 16:45:29 +0200 Organization: Square B.V. Message-ID: <362DF389.CB8B95B4@square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I'm trying to create the NSTextView system by hand, that is, using the NSTextContainer, NSLayoutManager, NSTextStorage classes. I have read the online documentations over a hundred times, but I don't seem to be able to create an editable view. Here's the source code, I hope anybody can help me. I'm pretty freaked out by now! :-( - (void)createTextViewWithTextStorage:(NSTextStorage *)storage; { NSRect frame; NSTextContainer *textContainer; NSTextView *textView; NSLayoutManager *layoutManager; NSScrollView *scrollView; NSParameterAssert(storage != nil); frame = [self bounds]; scrollView = [ [NSScrollView alloc] initWithFrame: frame]; [scrollView setHasHorizontalScroller: YES]; [scrollView setHasVerticalScroller: YES]; [self addSubview: scrollView]; // Create NSSLayoutManager layoutManager = [[NSLayoutManager allocWithZone:[self zone]] init]; [textStorage addLayoutManager:layoutManager]; // Create and configure NSTextContainer textContainer = [[NSTextContainer allocWithZone:[self zone]] initWithContainerSize:NSMakeSize(frame.size.width, LargeNumberForText)]; [textContainer setWidthTracksTextView:YES]; [textContainer setHeightTracksTextView:NO]; [layoutManager addTextContainer:textContainer]; [textContainer release]; // Create and configure NSTextView textView = [[NSTextView allocWithZone:[self zone]] initWithFrame:frame textContainer:textContainer]; [textView setMinSize:frame.size]; [textView setMaxSize:NSMakeSize(LargeNumberForText, LargeNumberForText)]; [textView setHorizontallyResizable:NO]; [textView setVerticallyResizable:YES]; [textView setAutoresizingMask:NSViewWidthSizable]; [textView setSelectable:YES]; [textView setEditable:YES]; [textView setRichText:YES]; [textView setImportsGraphics:YES]; [textView setUsesFontPanel:YES]; [textView setUsesRuler:YES]; [scrollView setDocumentView: textView]; } -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto: mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+
From: "James Campbell" <James@Virtual-Information.com> Newsgroups: comp.sys.next.programmer Subject: Access97 Compression???????? Message-ID: <MDsX1.1264$T01.703@nnrp2.ptd.net> Date: Wed, 21 Oct 1998 21:47:24 GMT NNTP-Posting-Date: Wed, 21 Oct 1998 17:47:24 EDT Organization: PenTeleData http://www.ptd.net Could someone please give me some information about a way to compress a huge Access database so that it would still be able to be accessed by visual basic 5.0. I am with a growing software company and we are very close to our newest release of our program. Our problem is that we must fit the whole program on a single CD. We need a way to compress our database so that the user will be able to still run our program from the rom .We also would not like to lose too much speed. If you know of a way we could acheive this through certain tools or 3rd party software please contact me at James@Virtual-Information.com or call me at 1-888-420-8099. Thank you, James Campbell ICQ# 6249313
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <8AuX1.9892$5w6.51@newsfeed.slurp.net> Control: cancel <8AuX1.9892$5w6.51@newsfeed.slurp.net> Date: 21 Oct 1998 23:58:45 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.8AuX1.9892$5w6.51@newsfeed.slurp.net> Sender: jose@adv.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: Access97 Compression???????? Date: 22 Oct 1998 05:31:21 GMT Organization: Idiom Communications Message-ID: <70mfv9$aln$2@news.idiom.com> References: <MDsX1.1264$T01.703@nnrp2.ptd.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: James@Virtual-Information.com "James Campbell" may or may not have said: -> Could someone please give me some information about a way to compress a huge -> Access database so that it would still be able to be accessed by visual -> basic 5.0. I am with a growing software company and we are very close to our -> newest release of our program. Our problem is that we must fit the whole -> program on a single CD. We need a way to compress our database so that the -> user will be able to still run our program from the rom .We also would not -> like to lose too much speed. If you know of a way we could acheive this -> through certain tools or 3rd party software please contact me at -> James@Virtual-Information.com or call me at 1-888-420-8099. -> Thank you, -> James Campbell -> ICQ# 6249313 Wrong newsgroup, sport. Comp.sys.next.programmer is for discussions of software development under the NEXTSTEP, OpenStep and Apple MacOS X systems. Nobody here does VB, MS Access or any other Microsoft crippleware. Frankly, I'm shocked that you consider anything written with these dismally inadequate tools stable enough to ship commercially. -jcr -- What the Bard *should* have said: "All the World's a Stage, and all the Men and Women rather poorly rehearsed."
From: heller@lrz.de (Helmut Heller) Newsgroups: comp.sys.next.programmer Subject: Re: Distributed Objects over the internet -- how? Date: 21 Oct 1998 07:54:47 GMT Organization: [posted via] Leibniz-Rechenzentrum, Muenchen (Germany) Distribution: world Message-ID: <70k407$cn3$1@sparcserver.lrz-muenchen.de> References: <908835801.409015@watserv4.uwaterloo.ca> In article <908835801.409015@watserv4.uwaterloo.ca> dfevans@bbcr.uwaterloo.ca (David Evans) writes: > In article <70fgfp$638$1@sparcserver.lrz-muenchen.de>, > Helmut Heller <heller@lrz.de> wrote: > > > >Is there another NS app with which I could test if the communication between > >the two nmservers works? Does -NXHost-ing use nmservers? > > > > Sure does! Yeah, and this solved the mystery: The IP from PPP did not match the "normal" (offline) IP I had assigned to my NeXT. I knew that this made -NXHost impossible. Now the IPs match and DO and NetTalk work!! However, now my other NeXT doesn't find its NetInfo Master (which is now on a different logical IP subnet) any more. I guess broadcasts don't reach through any more. Is there a way how NetInfo can be done in different subnets? Thanks! Helmut -- Servus, Helmut (DH0MAD) ______________NeXT-mail welcome_________________ FAX: +49-89-280-9460 "Knowledge must be gathered and cannot be given" heller@lrz.de ZEN, one of BLAKES7 Phone: +49-89-289-28823 ------------------------------------------------ Dr. Helmut Heller Leibniz-Rechenzentrum (LRZ)
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: Re: Overriding methods from a parallel category in objective-c -- how ? Date: 21 Oct 98 12:50:13 Organization: Is a sign of weakness Message-ID: <SCOTT.98Oct21125013@slave.doubleu.com> References: <6vkvcr$jcs$1@news.fr.internet.bosch.de> <1dh78yj.1zyf2doyiob3N@ascend-tk-p28.rhrz.uni-bonn.de> In-reply-to: theisen@akaMail.com's message of Tue, 20 Oct 1998 13:21:41 +0200 In article <1dh78yj.1zyf2doyiob3N@ascend-tk-p28.rhrz.uni-bonn.de>, theisen@akaMail.com (Dirk Theisen) writes: Can anyone explain how categories are /implemented/ internally in ObjC? I see that this is on the objc group, too, so fair warning: I'm describing NeXT's implementation. I expect other implementations to be similar, though the precise names may have changed. When you have an object, the class information is at self->isa. Therein is a objc_method_list reference called methods (later methodLists) which refers to a linked list of objc_method_list elements. When a category is loaded, it links itself at the head of that linked list, and clears the method cache for that class (and presumably for subclasses - or perhaps all method caches, I don't know the specifics). If another category is loaded which overrides the same methods of the same class, it'll link itself at the head of the list, and thus override the earlier category's versions of those methods. The category class methods are added to the isa->methods list of the class. So, if you have an object of the class, at self->isa->isa->methods. About a year ago, someone even wrote a "super" implementation for categories. Basically, from a given method, you can spelunk through the objc_class info, find the current method - and then continue to find the next one in the linked list. Of course, this loses out on the method cache, so it's going to be slower. But it lets you call the original version of the method you're overriding (something which might be easier done using +poseAs:, to be honest :-). Is there any more in deep documentation on that? I'd go look at the GNU runtime source, which should be pretty similar on this point. Otherwise, check out /usr/include/objc/objc-class.h on OS/Mach or NeXTSTEP. I don't recall if the POC has categories or not (the code I've ported to it didn't need them), but if it does it would certainly be worth checking out. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: rdieter@math.unl.edu (Rex Dieter) Newsgroups: comp.sys.next.programmer Subject: Re: Distributed Objects over the internet -- how? Date: 21 Oct 1998 14:20:41 GMT Organization: University of Nebraska-Lincoln Distribution: world Message-ID: <70kqjp$bjg$1@unlnews.unl.edu> References: <70k407$cn3$1@sparcserver.lrz-muenchen.de> In article <70k407$cn3$1@sparcserver.lrz-muenchen.de> heller@lrz.de (Helmut Heller) writes: > Is there a way how NetInfo can be done in different subnets? (posted and mailed) There must be a netinfo server (master OR clone) on each subnet. So, make the machine on the remote subnet a netinfo clone. -- Rex A. Dieter rdieter@math.unl.edu (MIME OK) Computer System Manager http://www.math.unl.edu/~rdieter/ Mathematics and Statistics University of Nebraska-Lincoln
From: "Mark Bessey" <mbessey@apple.com> Newsgroups: comp.sys.next.programmer Subject: Re: Distributed Objects over the internet -- how? Date: Thu, 22 Oct 1998 11:27:11 -0700 Organization: Apple Computer, Inc Distribution: world Message-ID: <70ntgh$iqo$1@news.apple.com> References: <70k407$cn3$1@sparcserver.lrz-muenchen.de> <70kqjp$bjg$1@unlnews.unl.edu> Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit Not true (at least, it's not supposed to be required). Look in the /machines directory of the client machine's local domain. There's an entry called broadcasthost, with an entry like this: ip_address 255.255.255.255 serves ./network If you change the ip_address to the actual IP address of the NetInfo master, you should be able to bind to it during startup. Keep in mind that the netmask, IP address, and router for your client machine have to be set correctly before it attempts to connect to NetInfo. Normally, you'd specify these things in NetInfo, and let a NeXT BootP server transmit them to the client, but in this case, you'll have to specify them in /etc/hostconfig, I think. If you have the NeXT System Administration manual, I think this is all explained in there somewhere... Hope this helps, -Mark (if you mess up your local NetInfo database sufficiently, you might not be able to boot anymore. If that happens, boot in single-user mode, and copy the default database from /usr/template/client/etc/netinfo) ---------- In article <70kqjp$bjg$1@unlnews.unl.edu>, rdieter@math.unl.edu (Rex Dieter) wrote: >In article <70k407$cn3$1@sparcserver.lrz-muenchen.de> heller@lrz.de (Helmut >Heller) writes: > >> Is there a way how NetInfo can be done in different subnets? > >(posted and mailed) > >There must be a netinfo server (master OR clone) on each subnet. So, make >the machine on the remote subnet a netinfo clone. > >-- >Rex A. Dieter rdieter@math.unl.edu (MIME OK) >Computer System Manager http://www.math.unl.edu/~rdieter/ >Mathematics and Statistics >University of Nebraska-Lincoln
From: dfevans@bbcr.uwaterloo.ca (David Evans) Newsgroups: comp.sys.next.programmer Subject: Re: Distributed Objects over the internet -- how? Date: 22 Oct 1998 19:44:03 GMT Organization: University of Waterloo Message-ID: <909085443.675347@watserv4.uwaterloo.ca> References: <70k407$cn3$1@sparcserver.lrz-muenchen.de> <70kqjp$bjg$1@unlnews.unl.edu> <70ntgh$iqo$1@news.apple.com> Cache-Post-Path: watserv4.uwaterloo.ca!unknown@bcr11.uwaterloo.ca In article <70ntgh$iqo$1@news.apple.com>, Mark Bessey <mbessey@apple.com> wrote: >Not true (at least, it's not supposed to be required). Look in the /machines >directory of the client machine's local domain. There's an entry called >broadcasthost, with an entry like this: > ip_address 255.255.255.255 > serves ./network >If you change the ip_address to the actual IP address of the NetInfo master, >you should be able to bind to it during startup. > Uhhh...wouldn't it be better to move the "serves" property from broadcasthost to the entry for the NetInfo master? It seems nasty to go around changing the IP address of broadcasthost... Or will this not work? -- David Evans (NeXTMail/MIME OK) dfevans@bbcr.uwaterloo.ca Computer/Synth Junkie http://bbcr.uwaterloo.ca/~dfevans/ University of Waterloo "Default is the value selected by the composer Ontario, Canada overridden by your command." - Roland TR-707 Manual
From: izumi@pinoko.berkeley.edu (Izumi Ohzawa) Newsgroups: comp.sys.next.programmer Subject: Re: Distributed Objects over the internet -- how? Date: 23 Oct 1998 18:26:58 GMT Organization: University of California, Berkeley Distribution: world Message-ID: <70qhpi$skm$1@agate.berkeley.edu> References: <909085443.675347@watserv4.uwaterloo.ca> In article <909085443.675347@watserv4.uwaterloo.ca> dfevans@bbcr.uwaterloo.ca (David Evans) writes: >In article <70ntgh$iqo$1@news.apple.com>, >Mark Bessey <mbessey@apple.com> wrote: >>Not true (at least, it's not supposed to be required). Look in the /machines >>directory of the client machine's local domain. There's an entry called >>broadcasthost, with an entry like this: >> ip_address 255.255.255.255 >> serves ./network >>If you change the ip_address to the actual IP address of the NetInfo master, >>you should be able to bind to it during startup. >> > > Uhhh...wouldn't it be better to move the "serves" property from broadcasthost >to the entry for the NetInfo master? It seems nasty to go around changing the >IP address of broadcasthost... > Or will this not work? You should NOT change the IP address of broadcasthost. Something else might break. Instead, add another entry for a NetInfo parent. You can have multiple hosts with the "serves: ../network" property. Marc Majka wrote long time ago: ----- From: majka@next.com (Marc Majka) Newsgroups: comp.sys.next.sysadmin Subject: Re: netinfo between different subnets Date: 1 Jul 1994 17:30:52 GMT [.....] Look in one of your local domains. You'll find an entry for /machines/broadcasthost: name: broadcasthost ip_address: 255.255.255.255 serves: ../network This means that when a "netinfod local" server tries to bind to a parent, it will broadcast a bind request to "netinfod network". It is extremely important to note that (1) you should *never* change this address, and that (2) the "netinfod local" server isn't *restricted* to sending a bind request to the broadcast address. As I mentioned above, a child looks for ALL "serves ../<tag>" properties, and sends a bind request to ALL of them. It may be that ONE of the bind requests it sends is directed to the address 255.255.255.255, which means something special to the Internet protocol (IP) software on your computer: it means "this message is addressed to every computer on the local area network". The astute reader will have noticed that this is the first and only time that I've said that any NetInfo communication is related in any manner to the network topology. We ship NetInfo so that the only parent that the "local" servers have in their database are the one shown above: "network" at 255.255.255.255. If you want the local domain of a computer on one subnet to be able to bind to a parent server running on a computer on another subnet, then all you need to do is to tell the child how to contact that parent. If, for example, you want the local domain to be able to bind to "network" at polaris, then all you need do is add /machines/polaris to the local domain: name: polaris ip_address: 129.18.2.7 serves: ../network The reasons that we ship NetInfo with only the "broadcasthost" parent in the local domain are: (1) we don't know the addresses and names of your computers when we put a /etc/netinfo/local.nidb on your NEXTSTEP CD-ROM. (2) in most cases, there will be a parent or two on the local subnet, so sending out a broadcast bind request is sufficient to find at least one running parent. This means less work for you, the system administrator, since you don't need to edit each and every one of your local NetInfo databases when you set up your NEXTSTEP network Since for the most part, local domains do rely exclusively on broadcast to find a parent, we recommend that, after you've established the first NEXTSTEP system on a new subnet, you make a clone of the parent domain on that computer. That way, you won't have to edit the local database on any other computers you add to that subnet, since they'll contact the new clone when they broadcast. In fact, we recommend that you have at least two servers per subnet for any domain that relies on broadcast. Why two? So that if one is down, the other will still be available to answer broadcast bind requests.
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: Selecting elements in a large matrix in a scrollview in OpenStep. Date: 23 Oct 98 07:31:43 Organization: Is a sign of weakness Distribution: world Message-ID: <SCOTT.98Oct23073143@slave.doubleu.com> There appears to be a bug in OpenStep's NSView event-translation machinery. Put a matrix of text fields in a scrollview, with the matrix being 3-4x wider than the visible area. Select the leftmost cell. Then, scroll to the far right and attempt to select the rightmost cell. You get scrolled back to the left and it selects a cell to the right of the original selection by about the clipview's width. This is consistent in that every time you scroll to the right and try again, back you go - but this time another clipview width to the right. It also occurs when selecting far to the left of the current selection. It also occurs in the vertical dimension. And it's not something specific to our program (I just tested it in IB). Fortunately, the vast majority of our customers use structures which only result in four or five columns in the matrix, and perhaps two times the clipview height, so it's generally not a problem. The exception is one extreme case where the matrix with is about 20x the clipview width. This problem makes editting that element very very annoying. Any ideas on how to fix it? I'm considering putting together a custom scrollview, or matrix, or both, and trying to see if I can factor out where the problem is coming from. But if someone else has solved it... -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: "Jeremy Bettis" <jeremy@hksys.com> Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: Re: Selecting elements in a large matrix in a scrollview in OpenStep. Date: Fri, 23 Oct 1998 16:29:49 -0500 Organization: Internet Nebraska Message-ID: <70qslr$mo2$1@owl.inetnebr.com> References: <SCOTT.98Oct23073143@slave.doubleu.com> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit >There appears to be a bug in OpenStep's NSView event-translation >machinery. Put a matrix of text fields in a scrollview, with the >matrix being 3-4x wider than the visible area. Select the leftmost >cell. Then, scroll to the far right and attempt to select the >rightmost cell. You get scrolled back to the left and it selects a >cell to the right of the original selection by about the clipview's >width. This is consistent in that every time you scroll to the right >and try again, back you go - but this time another clipview width to >the right. It also occurs when selecting far to the left of the >current selection. It also occurs in the vertical dimension. And >it's not something specific to our program (I just tested it in IB). Here is a solution. (It is also at http://jeremy.hksys.com/openstep/index.html) #import <AppKit/NSMatrix.h> // ==================================================================== // A NSForm inside of a NSScrollView does not handle mouse clicks // correctly. This subclass fixes that. It scrolls the old selected // cell to visible while activating the new one. // ==================================================================== static int dontscroll = 0; @interface HackedMatrix : NSMatrix - (void) scrollCellToVisibleAtRow:(int)r column:(int)c; - (void) textDidEndEditing:(NSNotification *)notify; @end @implementation HackedMatrix + (void) load { id pool = [NSAutoreleasePool new]; [self poseAsClass:[NSMatrix class]]; [pool release]; } - (void) scrollCellToVisibleAtRow:(int)r column:(int)c { if (dontscroll <= 0) { [super scrollCellToVisibleAtRow:r column:c]; } dontscroll = 0; } - (void) textDidEndEditing:(NSNotification *)notify { dontscroll=1; NS_DURING [super textDidEndEditing:notify]; NS_HANDLER dontscroll = 0; [localException raise]; NS_ENDHANDLER dontscroll = 0; } @end
From: "Chris Van Buskirk" <cvbuskirk@home.com> Newsgroups: comp.sys.next.programmer Subject: python... Message-ID: <5Z7Y1.1468$nR3.5733336@news.rdc1.nj.home.com> Date: Fri, 23 Oct 1998 23:05:37 GMT NNTP-Posting-Date: Fri, 23 Oct 1998 16:05:37 PDT Organization: @Home Network Is there a version of python for mac osx server beta 2+? -cgrus
From: "Chris Van Buskirk" <cvbuskirk@home.com> Newsgroups: comp.sys.next.programmer Subject: Good editor... Message-ID: <8_7Y1.1469$nR3.5733708@news.rdc1.nj.home.com> Date: Fri, 23 Oct 1998 23:06:44 GMT NNTP-Posting-Date: Fri, 23 Oct 1998 16:06:44 PDT Organization: @Home Network Is there a nice code editor availble for mac osx server b2? Something like bbedit. Jeezz, I have looked all over and cant find one. -chris
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: Re: Selecting elements in a large matrix in a scrollview in OpenStep. Date: 23 Oct 1998 17:03:41 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn731dni.bh2.rog@talisker.ohm.york.ac.uk> References: <SCOTT.98Oct23073143@slave.doubleu.com> On 23 Oct 98 07:31:43, Scott Hess <scott@nospam.doubleu.com> wrote: > Any ideas on how to fix it? I'm considering putting together a custom > scrollview, or matrix, or both, and trying to see if I can factor out > where the problem is coming from. But if someone else has solved > it... oh dear, scrollviews do seem to have some buggy interactions, don't they. just make sure you don't try backtabbing out of the scrollview as well... :-) as far as i can make out, it's unlikely that the problem is in NSView's event translation machinery, because it looks like the initial mouse down event is being converted as if the scrollview's clipview is in the position it was before scrolling, and there are enough examples around that *don't* exhibit this behaviour to make me suspect NSMatrix. i haven't got any more time to spend on it, but my initial approach would be to use custom subclasses of NSMatrix, NSScrollview and NSCell and print out the values that are being passed to convertPoint:fromView:, mouseDown:, and trackMouse: etc to see where the problem is first arising. what's the status on bugs like this, anyway? are they ever likely to get fixed before maxos X? cheers, rog.
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <70slek$i3t@marhabah.hct.ac.ae> Control: cancel <70slek$i3t@marhabah.hct.ac.ae> Date: 24 Oct 1998 13:48:44 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.70slek$i3t@marhabah.hct.ac.ae> Sender: cyber_hearts@hotmail.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: jmeacham@wittgenstein.jhuccp.org (James D. Meacham) Newsgroups: comp.sys.next.programmer,comp.sys.next.software Subject: Anyone got the Msql adapter working on Black NS/OS 4.2? Date: 23 Oct 1998 18:03:06 GMT Organization: The Center for Communications Programs of the Johns Hopkins University Distribution: world Message-ID: <70qgcq$7i8@news.jhu.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi All, I'm trying to build a WebObject app which uses Msql as it's back-end storage mechanism. Problem is, I can't get the damn Msql adapter to compile. Has anyone managed to get a black binary of this? Please let me know if you have, 'cause I'd love tp get my hands on one. Thanks James -- James David Meacham, 3rd jmeacham@wittgenstein.jhuccp.org Web Systems Administrator/Internet Services Coordinator Center for Communications Programs 410-659-6367 The Johns Hopkins University 410-659-6266 (fax)
Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <13094349252962304@yahoo.com> ignore no reply Control: cancel <13094349252962304@yahoo.com> Message-ID: <cancel.13094349252962304@yahoo.com> Date: Sat, 24 Oct 1998 13:34:39 +0000 Sender: nfvwhvxf@yahoo.com From: andrew@erlenstar.demon.co.uk Organization: Annihilator v0.3 Spam (EMP) cancelled - type=NAPRO
From: dfevans@bbcr.uwaterloo.ca (David Evans) Newsgroups: comp.sys.next.programmer Subject: Re: Print mechanism and data path Date: 25 Oct 1998 03:25:35 GMT Organization: University of Waterloo Message-ID: <909285935.455420@watserv4.uwaterloo.ca> References: <DKvMGNNTYhYH@cc.usu.edu> Cache-Post-Path: watserv4.uwaterloo.ca!unknown@bcr11.uwaterloo.ca In article <DKvMGNNTYhYH@cc.usu.edu>, <edx@cc.usu.edu> wrote: >where I might >tap into the rastered bitmap output when printing to a >NeXT laser. > I think that doing this is actually against the DPS license agreement that applies to NeXTSTEP--I think that you're only allowed to use the DPS server to RIP for the laser printer, colour printer, and monitor. What are you trying to do? Perhaps GhostScript could do it... -- David Evans (NeXTMail/MIME OK) dfevans@bbcr.uwaterloo.ca Computer/Synth Junkie http://bbcr.uwaterloo.ca/~dfevans/ University of Waterloo "Default is the value selected by the composer Ontario, Canada overridden by your command." - Roland TR-707 Manual
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <12647908683221@digifix.com> Date: 25 Oct 1998 03:47:04 GMT Organization: Digital Fix Development Message-ID: <18198909288022@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1994. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: Print mechanism and data path Date: 25 Oct 1998 09:56:45 GMT Organization: Idiom Communications Distribution: world Message-ID: <70uskt$8ju$1@news.idiom.com> References: <DKvMGNNTYhYH@cc.usu.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: edx@cc.usu.edu edx@cc.usu.edu may or may not have said: -> Does anyone know what data path is followed from the -> time a person hits the 'print' button to the time it -> comes out on the printer? I'd specifically like to know -> in what order processes are invoked and where I might -> tap into the rastered bitmap output when printing to a -> NeXT laser. Not sure, but I think that data ends up piped in to /dev/np0, so if you fake out this entry, you should be able to capture the data. If what you're really after is to be able to turn arbitrary PS code into a raster of an arbitrary pixel density, just use NSImage. Lockfocus on the image, run the PS code, and then read the data out of the NSBitmapImageRep. -jcr -- What the Bard *should* have said: "All the World's a Stage, and all the Men and Women rather poorly rehearsed."
Newsgroups: comp.sys.next.programmer Subject: Print mechanism and data path Message-ID: <DKvMGNNTYhYH@cc.usu.edu> From: edx@cc.usu.edu Date: 24 Oct 98 18:32:20 MDT Distribution: world MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Does anyone know what data path is followed from the time a person hits the 'print' button to the time it comes out on the printer? I'd specifically like to know in what order processes are invoked and where I might tap into the rastered bitmap output when printing to a NeXT laser. Perhaps this might be a postscript question... Discussion or pointers to references would be appreciated in any case. Thanks a bunch. edx@cc.usu.edu
From: dfevans@bbcr.uwaterloo.ca (David Evans) Newsgroups: comp.sys.next.programmer Subject: Re: Print mechanism and data path Date: 25 Oct 1998 23:36:13 GMT Organization: University of Waterloo Message-ID: <909358573.700887@watserv4.uwaterloo.ca> References: <DKvMGNNTYhYH@cc.usu.edu> <okJoCw0P2XEN@cc.usu.edu> <909338076.485873@watserv4.uwaterloo.ca> <Oma6oUPekZO$@cc.usu.edu> Cache-Post-Path: watserv4.uwaterloo.ca!unknown@bcr11.uwaterloo.ca In article <Oma6oUPekZO$@cc.usu.edu>, <edx@cc.usu.edu> wrote: > >Well, that clears -that- up. Thanks for the information. > Yeah. Sorry to be a party-pooper... -- David Evans (NeXTMail/MIME OK) dfevans@bbcr.uwaterloo.ca Computer/Synth Junkie http://bbcr.uwaterloo.ca/~dfevans/ University of Waterloo "Default is the value selected by the composer Ontario, Canada overridden by your command." - Roland TR-707 Manual
From: "Mark Bessey" <mbessey@apple.com> Newsgroups: comp.sys.next.programmer Subject: Re: Distributed Objects over the internet -- how? Date: Mon, 26 Oct 1998 11:17:12 -0800 Organization: Apple Computer, Inc Distribution: world Message-ID: <712hul$i3u$1@news.apple.com> References: <909085443.675347@watserv4.uwaterloo.ca> <70qhpi$skm$1@agate.berkeley.edu> Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit Well, that's what I get for not reading the docs before recommending something. Oh well, thanks for the clarification. -Mark (I reallllly should know this stuff a little better...) ---------- In article <70qhpi$skm$1@agate.berkeley.edu>, izumi@pinoko.berkeley.edu (Izumi Ohzawa) wrote: > >You should NOT change the IP address of broadcasthost. Something else >might break. Instead, add another entry for a NetInfo parent. You can >have multiple hosts with the "serves: ../network" property
From: Xavier.Nicolay@polytechnique.fr (Xavier NICOLAY) Newsgroups: comp.sys.next.programmer Subject: POSIX & Openstep Date: 27 Oct 1998 10:36:03 GMT Organization: Ecole Polytechnique Message-ID: <7147mj$ng6@polytechnique.polytechnique.fr> Keywords: POSIX With Openstep 4.2, after compilation, i obtain: /bin/ld: Undefined symbols: _sigaction _sigaddset _sigdelset _sigemptyset _sigismember _sigprocmask _WIFSTOPPED _WTERMSIG *** Exit 1 (some source should have been compiled with define _POSIX_SOURCE). How can i do ? -- Xavier NICOLAY | Ecole Polytechnique / SITX Chronique d'un depart annonce | 91128 PALAISEAU Cedex mailto:Xavier.Nicolay@polytechnique.fr | Tel: (33) 01 69 333 585
From: rdieter@math.unl.edu (Rex Dieter) Newsgroups: comp.sys.next.programmer Subject: Re: POSIX & Openstep Date: 27 Oct 1998 15:08:48 GMT Organization: University of Nebraska-Lincoln Message-ID: <714nm0$3br$1@unlnews.unl.edu> References: <7147mj$ng6@polytechnique.polytechnique.fr> In article <7147mj$ng6@polytechnique.polytechnique.fr> Xavier.Nicolay@polytechnique.fr (Xavier NICOLAY) writes: > With Openstep 4.2, after compilation, i obtain: > /bin/ld: Undefined symbols: > _sigaction > _sigaddset > _sigdelset > _sigemptyset > _sigismember > _sigprocmask > _WIFSTOPPED > _WTERMSIG > *** Exit 1 > > (some source should have been compiled with define > _POSIX_SOURCE). 1. Rewrite/port the code to not rely on POSIX function calls. 2. POSIX support was officially dropped from Openstep as of v4.2, see #1 3. Pre 4.2 POSIX support was poor. It had bugs and was broken, see #1. 4. Re-read 1, 2 and 3 .... 9998. See #4 9999. If all else fails, you can hack in support for the once supported -posix compiler flag. You'll need to find a copy of libposix.a (available with NEXTSTEP 3.3 and from NeXTanswers I believe), and a modified /lib/i386/specs file (appended uuencoded attachment). You need to recompile all source files with cc -posix Good luck, your mileage may vary, enjoy, and all that jazz. (-; -- Rex A. Dieter rdieter@math.unl.edu (MIME OK) Computer System Manager http://www.math.unl.edu/~rdieter/ Mathematics and Statistics University of Nebraska-Lincoln begin 444 specs M*F%S;3H*"@HJ87-M7V9I;F%L.@H*"BIC<'`Z"B5[(71R861I=&EO;F%L.B`M M1%]?4U1$0U]??2`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("5[<&]S M:7AS=')I8W0Z+41?4$]325A?4T]54D-%?2`@("`@("`@("`@("`@("`@("`@ M("`@("`@("5[(7!O<VEX<W1R:6-T.B5[8G-D.BU$7U]35%))0U1?0E-$7U]] M("`@("`@("`@("`@("`@("`@("5[<&]S:7@Z+41?4$]325A?4T]54D-%?2`@ M("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("5[(6%N<VDZ+41?3D58 M5%]33U520T5]?2`@("`@("`@("`@("`@("`@("`)"2`@)7MM9&ES86)L92UF M<')E9W,Z+41?7TY/7T907U]]"2`)"2`@)7M&*GT))7MA;G-I?2`E>V9N;RUA M<VU]"2`@("`@("`@(`D)("`E>W)E;#-C;VUP870Z?0D)(`D@"0D@("U$3EA? M0T]-4$E,15)?4D5,14%315\S7S`],S`P("`@("`@("`@(`D)("`M1$Y87T-/ M35!)3$527U)%3$5!4T5?,U\Q/3,Q,"`@("`@("`@("`)"2`@+41.6%]#3TU0 M24Q%4E]214Q%05-%7S-?,CTS,C`@("`@("`@("`@"0D@("U$3EA?0T]-4$E, M15)?4D5,14%315\S7S,],S,P("`@("`@("`@(`D)("`M1$Y87T-54E)%3E1? M0T]-4$E,15)?4D5,14%313TT,3`@"0D@("U$3E-?5$%21T54/30Q"0D@"0D@ M("U$3E-?5$%21T547TU!2D]2/30@"0D@("U$3E-?5$%21T547TU)3D]2/3$@ M"0D@("5[=')A9&ET:6]N86PZ+41?7T=.55]#4%!?7WT)"2`)"2`@)7LA=')A M9&ET:6]N86PZ)7MT<F%D:71I;VYA;"UC<'`Z+41?7T=.55]#4%!?7WT@"0D) M"2`E>R%T<F%D:71I;VYA;"UC<'`Z+41?7TY%6%1?0U!07U]]?2`@("`@("`@ M("`@("`@("`@("`E>W-T871I8SHM1%]?4U1!5$E#7U]])7LA<W1A=&EC.BU$ M7U]$64Y!34E#7U]]("`@("`@("`@("`@("`@("`@("5[340Z+4U$("5-?2`E M>TU-1#HM34U$("5-?0H**F-C,3H*"@HJ8V,Q<&QU<SH*"@HJ96YD9FEL93H* M"@HJ;&EN:SH*)7M:?2`E>TU]("5[1BI]("5[97AE8W5T92I]("5[<')E;&]A M9"I]("5[9G9M;&EB*GT@)7MS96=A;&EG;BI]("5[<V5G,6%D9'(J?2`E>W-E M9V%D9'(J?2`E>W-E9W!R;W0J?2`E>W!A9V5Z97)O7W-I>F4J?2`E>W5N9&5F M:6YE9"I]("5[9'EL:6)?9FEL92I]("5[<V5G;&EN:V5D:70J?2`E>VYO<V5G M;&EN:V5D:70J?2`E>W)E861?;VYL>5]R96QO8W-]("5[<V5C=&-R96%T92I] M("5[<V5C=&%L:6=N*GT@)7MS96-T;V)J96-T<WEM8F]L<WTE>W-E9V-R96%T M92I]("5[36%C:"I]("5[=VAY;&]A9'T@)7MW?2`E>W-E8W1O<F1E<BI]("5[ M=VAA='-L;V%D961]("5[3V)J0WT@)7MA;&Q?;&]A9'T@)7MO8FIE8W1]("5[ M9'EL:6YK97)]("5[9'EL:6YK97)?:6YS=&%L;%]N86UE*GT@)7MO=71P=71? M9F]R7V1Y;&1]("5[:V5E<%]P<FEV871E7V5X=&5R;G-]("5[<')E8FEN9'T@ M)7MN;W!R96)I;F1]"@HJ;&EB.@HE>W-T871I8SHE>R%D>6YA;6EC;&EB.B5[ M(7!O<VEX.BUL<WES7W,@+54@7U]D>6QD7V9U;F-?;&]O:W5P('TE>W!O<VEX M.BUL<&]S:7A]?7T@("`@("`@)7LA<W1A=&EC.B5[<&]S:7@Z("UL<&]S:7@@ M+54@9'EL9%]S='5B7V)I;F1I;F=?:&5L<&5R('TE>R%P9SHM9G)A;65W;W)K M(%-Y<W1E;7TE>W!G.BUF<F%M97=O<FL@4WES=&5M+%]P<F]F:6QE?7T*"BIL M:6)G8V,Z"B5[(7-H87)E9#HE>W-T871I8SHE>R%D>6YA;6EC;&EB.BUL8V-] M?2`)"0D@("`@("`E>R%S=&%T:6,Z+6QC8U]D>6YA;6EC?7T*"BIS=&%R=&9I M;&4Z"B5[(7!O<VEX*CHE>R%D>6YA;6EC;&EB.B5[8G5N9&QE.B5[(7-T871I M8SHM;&)U;F1L93$N;WU]("`@("5[(6)U;F1L93HE>W!G.B5[<W1A=&EC.BUL M9V-R=#`N;WT@"0D@("5[(7-T871I8SHE>V]B:F5C=#HM;&=C<G0P+F]](`D) M"2`@("`E>R%O8FIE8W0Z)7MP<F5L;V%D.BUL9V-R=#`N;WT@"0D)"2`@("`@ M("5[(7!R96QO860Z+6QG8W)T,2YO?7U]?2`)("`@("5[(7!G.B5[<W1A=&EC M.BUL8W)T,"YO?2`)"2`@)7LA<W1A=&EC.B5[;V)J96-T.BUL8W)T,"YO?2`) M"0D@("`@)7LA;V)J96-T.B5[<')E;&]A9#HM;&-R=#`N;WT@"0D)"2`@("`@ M("5[(7!R96QO860Z+6QC<G0Q+F]]?7U]?7U]("`E>W!O<VEX*CHM;'!O<VEX M8W)T,"YO?0H**G-W:71C:&5S7VYE961?<W!A8V5S.@H*"BIS:6=N961?8VAA M<CH*)7MF=6YS:6=N960M8VAA<CHM1%]?0TA!4E]53E-)1TY%1%]??0H**G!R M961E9FEN97,Z"BU$:3,X-B`M1$YE6%0@+41U;FEX("U$7U]-04-(7U\@+41? M7TQ)5%1,15]%3D1)04Y?7R`M1%]?05)#2$E414-455)%7U\](FDS.#8B(`H* C*F-R;W-S7V-O;7!I;&4Z"C`*"BIM=6QT:6QI8CH*+B`["@H] ` end
From: tennant@alph.msfc.nasa.gov (Allyn Tennant) Newsgroups: comp.sys.next.programmer Subject: Re: POSIX & Openstep Date: 28 Oct 1998 01:19:09 GMT Organization: http://www.msfc.nasa.gov/ Message-ID: <715red$sjr$1@hammer.msfc.nasa.gov> References: <7147mj$ng6@polytechnique.polytechnique.fr> <714nm0$3br$1@unlnews.unl.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit rdieter@math.unl.edu (Rex Dieter) wrote: > >1. Rewrite/port the code to not rely on POSIX function calls. >2. POSIX support was officially dropped from Openstep as of v4.2, see >#1 3. Pre 4.2 POSIX support was poor. It had bugs and was broken, see I'm afraid in today's world, it is far easier to give up on OpenStep than to do any of the above. In other words, no one is going to buy OpenStep because it *doesn't* support POSIX. Allyn
From: ntakebay@guar.bio.indiana.edu (Naoki Takebayashi) Newsgroups: comp.sys.next.programmer Subject: a new book "Openstep programming" Date: 28 Oct 1998 02:58:01 GMT Organization: Indiana University, Bloomington Message-ID: <slrn73d21j.13m.ntakebay@guar.bio.indiana.edu> Hello, I was looking at amazon.com and found a book called "Openstep Programming" by William Ballew, Springer Verlag Pub, ISBN: 0387947205. Have anybody read this book yet? Since this book costs $49.95 (not cheep for a poor grad. student), I'd like to know how good it is before I purchase it. This book came out June 1998. Thanks in advance, Naoki
From: Maurice le Rutte <MleRutte@square.nl> Newsgroups: comp.sys.next.programmer Subject: Creating NSPrintInfo with NSDictionary Date: Wed, 28 Oct 1998 09:47:42 +0100 Organization: Square B.V. Message-ID: <3636DA2E.71AB0C2E@square.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I want to make a NSPrintInfo with a different, custom paper size. Which fields *must* be included in a NSPrintInfo to define 'MyPaper' with all custom variables. I don't know the printer at that moment. Maurice le Rutte. -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto: mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+
#################################################################### From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Vertical scrolling and word wrapping in a text field. Date: 27 Oct 98 11:41:36 Organization: Is a sign of weakness Distribution: world Message-ID: <SCOTT.98Oct27114136@slave.doubleu.com> We have a couple fields where the user _usually_ will enter a small word or phrase, but in some cases they enter a substantial piece of text, generally copied from somewhere else. We want the element to act like a text field in terms of tabbing and use of the enter key and whatever else applies. Long ago, in the days of NeXTSTEP, I seem to recall having solved the problem by hacking about with the field's settings, and allowing the user to resize the window the field was in. In the conversion to OpenStep, we managed to retain the resizing capabilities - but at some point the ability of the field to wrap and scroll vertically seems to have been lost. Right now, I've solved the problem by replacing the field with a text object in a scrollview, with the vertical scroller suppressed, and NSFieldFilter() as the character filter function, so as to emulate a field's tab and enter operation. Fortunately, the fields are all for editting strings. Anyone have a better solution? Thanks, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Just want to verify my understanding of NSBitmapImageRep. Date: 27 Oct 98 20:49:12 Organization: Is a sign of weakness Distribution: world Message-ID: <SCOTT.98Oct27204912@slave.doubleu.com> I've been doing some NSBitmapImageRep hackery the past couple days, and just wanted to make absolutely certain my understanding was precisely correct. First question: I'm assuming that 8 bps (bits per sample), 3 spp (samples per pixel) and 32 bpp (bits-per-pixel), meshed, will be among the most efficient color settings. As will 4bps, 3spp, and 16bpp, meshed. These both corrosponding to the NeXTdimension and color station color models. I assume that planar data is less efficient, but am wondering if at least 24 bpp (and possibly 12bpp) might be worth anything. What about 8bpp settings? Am I restricted to use 222/8, or could I use 332/8? Or would I be better off to resample to 444/16? [If so, I'm assuming simple shifts will suffice, meaning shift the R and G components <<1 and the B component <<2.] Likewise with 16bpp. Is 555/16 going to be as efficient as 444/16? [My Matrox driver become slower at 555/16 than the previous version was at 444/16.] Can I directly do 565/16 or would I be better to resample to 888/24? Say I have a large NSBitmapImageRep, with only a portion being updated. Anyone have a sense of the breakpoint at which it makes more sense to build a subset rep and draw it, rather than drawing the entire image? [Just to be fair, I'll answer my own questions as I see them. 444/16 and 888/32 should be very efficient. 8bpp is a non-starter and I should resample upwards. Shifting the components to fill the larger bps should work fine. 555/16 versus 444/16 will probably depend on whether the video card uses one or the other. The time spent to suck out a subset of an NSBitmapImageRep will become worthwhile when about half the pixels of the full version weren't going to be updated.] Thanks, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: jurgen@oic.de (Juergen Moellenhoff) Newsgroups: comp.sys.next.programmer Subject: Re: POSIX & Openstep Date: 28 Oct 1998 11:16:41 GMT Organization: OIC, Bochum, Germany Message-ID: <716uep$h9$1@nexus.oic.de> References: <7147mj$ng6@polytechnique.polytechnique.fr> <714nm0$3br$1@unlnews.unl.edu> <715red$sjr$1@hammer.msfc.nasa.gov> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit In <715red$sjr$1@hammer.msfc.nasa.gov> Allyn Tennant wrote: > rdieter@math.unl.edu (Rex Dieter) wrote: > > > > 1. Rewrite/port the code to not rely on POSIX function calls. > > OpenStep than to do any of the above. In other words, no one is going to > buy OpenStep because it *doesn't* support POSIX. I guess the bad POSIX "support" is not the main reason why no one is going to buy OPENSTEP. I know at least one big company that produces an OS that has also a bad POSIX support but everyone bought it... Bye, Jürgen
From: "Mark Lewis" <mlewis@worldnetcorp.com> Newsgroups: comp.sys.next.programmer Subject: Objective C / Next Programmer Wanted Date: Wed, 28 Oct 1998 10:38:09 -0600 Organization: Worldnet Corporation Message-ID: <717gtk$p3t$1@supernews.com> Dear Consultant, Our client in Downtown Chicago is in need of 2 Objective C programmers for a long term assignment in a prestigious account. Please send all qualified resumes including rates to: mlewis@worldnetcorp.com 847-839-9323-Direct. Thanks and have a great day. Mark
From: Rob Blessin <bhi1@ix.netcom.com> Newsgroups: comp.sys.next.programmer Subject: Question : Discovering Openstep a Developers Tutorial Date: Wed, 28 Oct 1998 14:44:18 -0700 Organization: Black Hole, Incorporatd Message-ID: <36379032.833C9117@ix.netcom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello NeXT / Apple community: Has anyone completed the Currency Conversion Application project? I decided to try out the tutorial on creating a Currency Conversion Application in the Openstep 4.2 Academic Bundle presented in Discovering Openstep a developers tutorial. Simple question , everything went well until here : Window is set up correctly as CurrencyConverter, Text fields and title fields are added with text. After choosing Views Palette in Preferences reference pg 27 , it pictures a CurrencyConverter.nib file window, I have this window and it is titled CurrencyConverter.nib, but the order of the tabs is different than the screen grab image in the book, my window tabs are in th is order, Instances, Classes, Sounds then Images . The book shows Images before Sounds, I select the Image tag as instructed in the CurrencyConverter.nib window. The NeXTSTEP is to select the NSReturnSign icon, it unfortunately is not there, the other 4 icons are installed but the NSReturnSign is not present , selectable or even visable image icon choice. It basically is dragged over to my button and gives it a return arrow. The missing NSReturnSign object in effect stalls my little appication project, how do I locate the object and install it in my CurrencyConverter.nib window or where did I take the wrong turn in setting up the CurrencyConverter.nib file to begin with? It looks as though Apple\NeXT may have changed to separate NeXTSTEP.nib and Windows.nib please reply to bhi1@ix.netcom.com Best regards Rob Blessin
From: "Andrew Malota" <rinn_the_great@hotmail.com> Newsgroups: comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software Subject: I need software for NeXTSTEP 2.0 Date: Wed, 28 Oct 1998 19:14:11 -0600 Organization: Texas A&M University, College Station, Texas Message-ID: <718fei$e4t$1@news.tamu.edu> NNTP-Posting-Date: 29 Oct 1998 01:12:50 GMT I need the following: GIF/JPEG viewer, 3D Rendering Program Drivers for the Global Village TelePort 36.6 Fax/Data Modem CAD program Web Browser PPP software. (If it in fact exists) Also, I you know where I can get a cheap OS upgrade (such as 3.3 on CD) please contact me. Thanks Andrew Malota
From: no@spam.please Newsgroups: comp.sys.next.programmer Subject: cfs - NFS server on NS/OS, lockups and kernel freezes Date: 28 Oct 1998 21:48:11 GMT Organization: Seicom GmbH, Reutlingen, Germany, Europe Message-ID: <7183er$mf3$1@news.seicom.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Howdy folks! I would like to find a solution for a strange problem which haunt me since I started compiling (I could not say porting as there was nothing to do except minor fiddling) cfs-1.3.3 on my NS3.3 box. For those asking what cfs is, cfs stands for cryptographic file system, a transparent way to store data secure on HD... and I got my copy from ftp.replay.com - a server in the netherlands, so no US export controls apply for me. cfs itself is implemented as a local NFS server running on another port, it receives read, write and administrative requests over the normal NFS RPC mechanism (the encrpyted filesystem is mounted using a normal mount command). So far so good, it does run indeed and works except... if I copy over large files (say > 2 MByte) to the cfs'd filesystem it sometimes locks up in the cfsd deamom, the lock happens inside the write() call in the kernel that writes the encrypted data to disk?! This effectively freezes the system. This behavious is reproduceable with the same file, what's funny it seems to change its behaviour with the length of the filename used (!!!???) - the implementation of a buffering mechanism that delays the writes did help a bit but still doesn't work for some others. Writing to raw devices (or network connections) is not affected. I (and others, Christian Starkjohann -the author of vmount and sharity which use a similar mechanism - was very helpful) suspect the problem is inside the buggy kernel NFS implementation, however if someone else has an idee what I could do the improve the behaviour I really would like to hear your suggestions. Thanks for listening - Frank [frank@wizards.de]
From: Mike Paquette <mpaque@ncal.verio.com> Newsgroups: comp.sys.next.programmer Subject: Re: Just want to verify my understanding of NSBitmapImageRep. Date: Wed, 28 Oct 1998 12:02:28 -0800 Organization: Electronics Service Unit No. 16 Message-ID: <36377849.7267ABF8@ncal.verio.com> References: <SCOTT.98Oct27204912@slave.doubleu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Scott Hess wrote: > > I've been doing some NSBitmapImageRep hackery the past couple days, > and just wanted to make absolutely certain my understanding was > precisely correct. > > First question: I'm assuming that 8 bps (bits per sample), 3 spp > (samples per pixel) and 32 bpp (bits-per-pixel), meshed, will be among > the most efficient color settings. As will 4bps, 3spp, and 16bpp, > meshed. These both corrosponding to the NeXTdimension and color > station color models. I assume that planar data is less efficient, > but am wondering if at least 24 bpp (and possibly 12bpp) might be > worth anything. Planar data is less efficient, as is 24 bpp color, particularly when compositing the bitmap data to the display. The 32 bit per pixel RGBA and 16 bit per pixel 4 bits per sample RGBA formats will be the fastest. > What about 8bpp settings? Am I restricted to use 222/8, or could I > use 332/8? Or would I be better off to resample to 444/16? [If so, > I'm assuming simple shifts will suffice, meaning shift the R and G > components <<1 and the B component <<2.] Don't bother. The software will just unpack your data to 32 bpp pixels, and push them through the conversion to the system's palette so as to pick up monitor calibration and gamma correction properly. > Likewise with 16bpp. Is 555/16 going to be as efficient as 444/16? > [My Matrox driver become slower at 555/16 than the previous version > was at 444/16.] Can I directly do 565/16 or would I be better to > resample to 888/24? Just run the data at RGB/24 bpp or RGBA/32 bpp. On Intel hardware the 555/16 and 565/16 framebuffer modes rarely have hardware gamma correction. (The RAM part of the RAMDAC operates only in 8 bit pseudocolor modes on the popular/cheap DACs found in most Intel hardware.) This means that in order to maintain the linear colorspace behaviors NeXT's graphics system provides, software gamma correction must be done. This is sneakily implemented 'for free' as part of the mechanism used to convert true color pixels to the skanky 555/16 and 565/16 formats we see on much Intel hardware. (Nothing is free, but it's cheaper than you might think.) Mac hardware supports hardware gamma correction in all modes, by the way. > Say I have a large NSBitmapImageRep, with only a portion being > updated. Anyone have a sense of the breakpoint at which it makes more > sense to build a subset rep and draw it, rather than drawing the > entire image? You can always initialize a NSBitmapImageRep pointing into a sub-rectangle of another bitmap. This ultimately has the same effect as just compositing a rect of the image, though. Copying the data to a new bitmap is rarely efficient beyond the smallest images. The reason is Mach's out of line transfer of large pieces of data as copy-on-write memory. Your bitmaps are sent to the WindowServer as pointers to out-of-line data in a Mach message. The messaging code in the Mach kernel arranges for the VM object representing your image to be mapped into the WindowServer, and dinks the pointer in the message to point at the data as mapped into the WindowServer. On an 8500/180 PPC machine, this adds 20-30 microseconds to the messaging time, for a total of about 80-90 microseconds to get your bitmap into the WindowServer. When updating a bitmap on the display, the real time sink can be simply blitting a bunch of unchanged pixels through the depth conversion code. Tracking dirty rects within your own bitmap and only updating these can cut down on the time spent blitting. > [Just to be fair, I'll answer my own questions as I see them. 444/16 > and 888/32 should be very efficient. 8bpp is a non-starter and I > should resample upwards. Shifting the components to fill the larger > bps should work fine. 555/16 versus 444/16 will probably depend on > whether the video card uses one or the other. The time spent to suck > out a subset of an NSBitmapImageRep will become worthwhile when about > half the pixels of the full version weren't going to be updated.]
From: Denise Howard <denisehAT@idiomDOT.com> Newsgroups: comp.sys.next.programmer Subject: PDI Animation Studio - Appication Programmer Date: 29 Oct 1998 04:10:00 GMT Organization: Remove capital letters to reply Message-ID: <718pqo$5hi$1@news.idiom.com> User-Agent: tin/pre-1.4-980117 (UNIX) (FreeBSD/2.2.6-STABLE (i386)) PDI is an animation studio specializing in computer-generated 3D animation. We have recently released "Antz", our computer-animated feature film in coproduction with DreamWorks SKG. In the 17 years we've been around, we developed a wide range of creative styles and worked on entertainment projects from feature films to commercials. Our work also includes The Simpsons' 3D Halloween special for Fox and special film effects for The Peacemaker and Batman and Robin. Check out our website at www.pdi.com or www.antz.com for more information. We're looking for an Applications Developer for our production tracking tools. The Developer works closely with Animators, Technical Directors, Production Management, and R&D groups. This is a dynamic position requiring an interest in production, process development, and the creative visual process. We will consider both full-time and contract candidates. Requirements for this position: - 2 to 5 years of application development - Experience with EOF and or WebObjects is a plus - Interest in animation and production - Familiarity with database programming - Excellent communication and project management skills - Sense of humor required If this sounds interesting, or you have more questions about the position or company, please give a call or send your resume to: Jennifer Yu Pacific Data Images 3101 Park Boulevard Palo Alto, CA 94306 E-mail: jen@pdi.com (ASCII ONLY) Phone: 650 846-8206 Fax: 650 846-8102
From: eric@skatter.USask.Ca Newsgroups: comp.sys.next.programmer Subject: Bash as /bin/sh Date: 29 Oct 1998 15:34:44 GMT Organization: University of Saskatchewan Message-ID: <71a1uk$6ep$1@tribune.usask.ca> The /bin/sh on OPENSTEP 4.2 is not up to the task of running some build/install scripts I've received lately. For now, I've installed bash-2.02 and I'm using it to run these scripts. I though it would be handy to replace /bin/sh with bash -- then the scripts would work `out-of-the-box', instead of having to edit the first line to make them use bash instead of /bin/sh. This `almost' works. The system reboots properly. I can log in. I can start up Terminal.app sessions. I can rlogin to my machine from other machines. Unfortunately, I can't telnet to my machine from other machines! I get through the login and password prompts, then I get a `Connection closed by remote host' message and the connection drops. Has anybody else tried using bash as /bin/sh? If so, what did you have to do to make incoming telnet sessions work? Thanks, -- Eric Norum eric@skatter.usask.ca Saskatchewan Accelerator Laboratory Phone: (306) 966-6308 University of Saskatchewan FAX: (306) 966-6058 Saskatoon, Canada.
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: Just want to verify my understanding of NSBitmapImageRep. Date: 29 Oct 98 08:33:41 Organization: Is a sign of weakness Message-ID: <SCOTT.98Oct29083341@slave.doubleu.com> References: <SCOTT.98Oct27204912@slave.doubleu.com> <36377849.7267ABF8@ncal.verio.com> <SCOTT.98Oct29071201@slave.doubleu.com> In-reply-to: scott@nospam.doubleu.com's message of 29 Oct 98 07:12:01 In article <SCOTT.98Oct29071201@slave.doubleu.com>, scott@nospam.doubleu.com (Scott Hess) writes: Now, for the have-my-cake-and-eat-it-too question: How does clipping affect this? Say we have an 1152x864 NSView, with a same-sized NSBitmapImageRep, with {{ x=200,y=150}, {width=350, height=250}} needing redrawing. If I use DPS clipping operations to clip to that rectangle, then blit the entire bitmap, does the windowserver still push all pixels through the dithering machinery? Well, empirical testing indicates that given the above bitmap and view, clipping makes the blit about 10x faster. The speedup does not seem quite linear WRT how many pixels you're clipping off, but it's close. [It's hard for me to tell, because since I was measuring DPS, I couldn't use CPU time, I had to use "real" or "wall" time. I did have the computer measuring the time, though:-).] For comparison, I also tested how things worked if I only drew the corrosponding portion of the bitmap from the bitmap. It was as close to the same speed as drawing the entire bitmap with clipping as one could get using my measurement tools, so I don't see any point to further pursuit of that question... I also found that blitting at 444/16 is about 4x faster than 888/32 on my machine (PPro200 with Matrox Millenium II using an older driver at 444/16). I suppose I'll have to test it with a 555/16 driver, too, just in case... Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: Just want to verify my understanding of NSBitmapImageRep. Date: 29 Oct 1998 17:23:15 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn73h947.452.rog@talisker.ohm.york.ac.uk> References: <SCOTT.98Oct27204912@slave.doubleu.com> <36377849.7267ABF8@ncal.verio.com> <SCOTT.98Oct29071201@slave.doubleu.com> NNTP-Posting-Date: 29 Oct 1998 17:23:15 GMT On 29 Oct 98 07:12:01, Scott Hess <scott@nospam.doubleu.com> wrote: > I can't appear to do with what I can find in > the Rhapsody AppKit docs is pick a specific rectangle out of the > NSBitmapImageRep I have and draw it one-to-one in the focusView. my take on this is that the NSBitmapImageRep functionality seems to map directly to the postscript "image" functionality which doesn't provide this facility, and so NSBitmapImageRep can't, without a complete rewrite. i've always understood (but i'm likely wrong here - please correct me!) that when blitting images under openstep, you're passing stuff between two sides of the user->ps interpreter divide. an NSBitmapImageRep holds its bitmap data in user space (hence -bitmapdata is a very cheap call) and when called upon to draw itself, all the data is passed down a port to the PS interpreter and rendered by the image or colorimage operators. that's where it gets all its funky scaling and dithering stuff from. on the other hand, a composite operation happens purely on the PS interpreter side, which is presumably why large bitmap composites can freeze the window system as only one composite can be happening at a time. i had a problem recently where i needed to get upwards of 200 screen changes a second, and experimented with all sorts of strategies before realising that the basic problem was the overhead of calls to the PS interpreter; it was just not possible to do 200 operations a second, whatever they were. (eventually i used the font machinery and abandoned the use of colour). > In article <36377849.7267ABF8@ncal.verio.com>, > Mike Paquette <mpaque@ncal.verio.com> writes: > When updating a bitmap on the display, the real time sink can be > simply blitting a bunch of unchanged pixels through the depth > conversion code. Tracking dirty rects within your own bitmap and > only updating these can cut down on the time spent blitting. i've found one problem with tracking your own dirty rects in a view, and that is that it doesn't seem to be possible... the problem being that it doesn't seem to be possible to enhance the default dirty-rectangle update machinery to keep track of more than one rectangle. (it just seems to union the rect passed to setNeedsDisplayInRect: with the existing dirty rectangle). i tried to bypass the display mechanism to do my own "updatedirtyrects" which was more intelligent, but under yellowbox if you don't update the window through the display methods then the window cache isn't updated properly (yeuch!). if you're using the normal display methods, then there doesn't seem to be a way of telling whether drawRect is being called just to update the view (because you called setNeedsDisplay) in which case you only need to redraw the dirty rects; or to completely redraw the view, in which case you need to redraw the entire rectangle (if you're in the contentview of a scrollview that's being scrolled, for instance). i never did find a satisfactory solution. it's lucky i wasn't programming a video game! cheers, rog.
Newsgroups: comp.sys.next.programmer From: maury@remove_this.istar.ca (Maury Markowitz) Subject: Re: Just want to verify my understanding of NSBitmapImageRep. Content-Type: text/plain; charset=us-ascii Message-ID: <F1LoMD.CDH@T-FCN.Net> Sender: news@T-FCN.Net Content-Transfer-Encoding: 7bit Cc: scott@nospam.doubleu.com Organization: needs one References: <SCOTT.98Oct27204912@slave.doubleu.com> <36377849.7267ABF8@ncal.verio.com> <SCOTT.98Oct29071201@slave.doubleu.com> <SCOTT.98Oct29083341@slave.doubleu.com> Mime-Version: 1.0 Date: Thu, 29 Oct 1998 17:59:01 GMT In <SCOTT.98Oct29083341@slave.doubleu.com> Scott Hess wrote: > Well, empirical testing indicates that given the above bitmap and > view, clipping makes the blit about 10x faster. The speedup does not > seem quite linear WRT how many pixels you're clipping off, but it's > close. [It's hard for me to tell, because since I was measuring DPS, > I couldn't use CPU time, I had to use "real" or "wall" time. I did > have the computer measuring the time, though:-).] > > For comparison, I also tested how things worked if I only drew the > corrosponding portion of the bitmap from the bitmap. It was as close > to the same speed as drawing the entire bitmap with clipping as one > could get using my measurement tools, so I don't see any point to > further pursuit of that question... Excellent stuff Scott, a heartfelt "thanks" to you. Maury
From: weigel+@pitt.edu (Matthew C Weigel) Newsgroups: comp.sys.next.programmer Subject: Re: Bash as /bin/sh Date: 29 Oct 1998 16:03:27 GMT Organization: University of Pittsburgh Message-ID: <71a3kf$oe5$1@usenet01.srv.cis.pitt.edu> References: <71a1uk$6ep$1@tribune.usask.ca> In article <71a1uk$6ep$1@tribune.usask.ca>, <eric@skatter.USask.Ca> wrote: >The /bin/sh on OPENSTEP 4.2 is not up to the task of running some >build/install scripts I've received lately. For now, I've installed >bash-2.02 and I'm using it to run these scripts. > >I though it would be handy to replace /bin/sh with bash -- then the scripts >would work `out-of-the-box', instead of having to edit the first line to make >them use bash instead of /bin/sh. > >This `almost' works. The system reboots properly. I can log in. I can >start up Terminal.app sessions. I can rlogin to my machine from other >machines. > >Unfortunately, I can't telnet to my machine from other machines! I get >through the login and password prompts, then I get a `Connection closed by >remote host' message and the connection drops. > >Has anybody else tried using bash as /bin/sh? If so, what did you have to do >to make incoming telnet sessions work? Is /bin/sh a symlink, or the file itself? I dunno if this would matter in OS/NS, and it shouldn't, but go ahead and try it. (the behavior of bash is dependant upon argv[0], don't forget) -- /-------------------------------------------------------------------------\ --Matthew BeOS | Linux Lab Consultant-- \\--Weigel OS2 | NeXT Programmer--// \\--------------------------+----------------------------//
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin Subject: cmsg cancel <3638c06b.1@newsman.viper.net> Control: cancel <3638c06b.1@newsman.viper.net> Date: 29 Oct 1998 19:15:54 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.3638c06b.1@newsman.viper.net> Sender: hquwiz@net800.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: Just want to verify my understanding of NSBitmapImageRep. Date: 29 Oct 98 07:12:01 Organization: Is a sign of weakness Message-ID: <SCOTT.98Oct29071201@slave.doubleu.com> References: <SCOTT.98Oct27204912@slave.doubleu.com> <36377849.7267ABF8@ncal.verio.com> In-reply-to: Mike Paquette's message of Wed, 28 Oct 1998 12:02:28 -0800 OK, so it sounds like my understanding is basically correct. One remaining question follows. In article <36377849.7267ABF8@ncal.verio.com>, Mike Paquette <mpaque@ncal.verio.com> writes: Scott Hess wrote: > Say I have a large NSBitmapImageRep, with only a portion being > updated. Anyone have a sense of the breakpoint at which it makes > more sense to build a subset rep and draw it, rather than drawing > the entire image? You can always initialize a NSBitmapImageRep pointing into a sub-rectangle of another bitmap. This ultimately has the same effect as just compositing a rect of the image, though. Copying the data to a new bitmap is rarely efficient beyond the smallest images. I can do that if I already have image data in the windowserver, I can suck a rectangle into an NSBitmapImageRep using -initWithFocusedViewRect:. I can also take an NSBitmapImageRep and draw it within an arbitrary rectangle with -drawInRect:, which will scale as necessary. What I can't appear to do with what I can find in the Rhapsody AppKit docs is pick a specific rectangle out of the NSBitmapImageRep I have and draw it one-to-one in the focusView. When updating a bitmap on the display, the real time sink can be simply blitting a bunch of unchanged pixels through the depth conversion code. Tracking dirty rects within your own bitmap and only updating these can cut down on the time spent blitting. Say I have an 1152x864 NSBitmapImageRep, and wish to redraw the rect at {{ x=200,y=150}, {width=350, height=250}}. Just looking at the copy-to-server overhead, the difference between tossing it all up there versus just what's necessary is minimal because it's using shared memory. But since the server has to revisit all those pixels doing dithering, it might be worthwhile to suck out the 350x250 rect of pixels into a new NSBitmapImageRep and only blit those pixels. Now, for the have-my-cake-and-eat-it-too question: How does clipping affect this? Say we have an 1152x864 NSView, with a same-sized NSBitmapImageRep, with {{ x=200,y=150}, {width=350, height=250}} needing redrawing. If I use DPS clipping operations to clip to that rectangle, then blit the entire bitmap, does the windowserver still push all pixels through the dithering machinery? Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin Subject: cmsg cancel <3638c06b.0@newsman.viper.net> Control: cancel <3638c06b.0@newsman.viper.net> Date: 29 Oct 1998 19:18:28 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.3638c06b.0@newsman.viper.net> Sender: hquwiz@net800.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: nurban@crib.corepower.com (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: Convert Objective-c to C Date: 29 Oct 1998 17:07:51 -0500 Organization: Data Systems Consulting, Inc. Message-ID: <71aovn$56u$1@crib.corepower.com> References: <3638E296.415DE7A6@pdgm.com> NNTP-Posting-Date: 29 Oct 1998 22:08:07 GMT In article <3638E296.415DE7A6@pdgm.com>, Tim <tim@pdgm.com> wrote: > I have write a new project using c, but old project > is written by Objective-c, I am new in objectiv-c. > If I can to convert objective-c to c, it is easy for > to read it. It wouldn't be very useful to you to convert Objective-C to C. Basically, that would just consist of replacing each [obj someMethod] method invocation with an objc_msgSend(obj,someMethodSelector) function call, which doesn't make things any easier to understand -- it's just different syntax. It still won't tell you what's going on at an object level. It's better for you to just learn a few things about the syntax and then you can read Objective-C just fine.
Newsgroups: comp.sys.next.programmer From: maury@remove_this.istar.ca (Maury Markowitz) Subject: Re: Just want to verify my understanding of NSBitmapImageRep. Content-Type: text/plain; charset=us-ascii Message-ID: <F1LoJy.CCF@T-FCN.Net> Sender: news@T-FCN.Net Content-Transfer-Encoding: 7bit Cc: scott@nospam.doubleu.com Organization: needs one References: <SCOTT.98Oct27204912@slave.doubleu.com> <36377849.7267ABF8@ncal.verio.com> <SCOTT.98Oct29071201@slave.doubleu.com> Mime-Version: 1.0 Date: Thu, 29 Oct 1998 17:57:34 GMT In <SCOTT.98Oct29071201@slave.doubleu.com> Scott Hess wrote: > You can always initialize a NSBitmapImageRep pointing into a > sub-rectangle of another bitmap. This ultimately has the same > effect as just compositing a rect of the image, though. Copying > the data to a new bitmap is rarely efficient beyond the smallest > images. I wish I hadn't missed earlier portions of this thread. This is an interesting solution to a feature I was thinking of adding actually - background pictures where only certain areas are going to need updating and at various times. > Now, for the have-my-cake-and-eat-it-too question: How does clipping > affect this? Say we have an 1152x864 NSView, with a same-sized > NSBitmapImageRep, with {{ x=200,y=150}, {width=350, height=250}} > needing redrawing. If I use DPS clipping operations to clip to that > rectangle, then blit the entire bitmap, does the windowserver still > push all pixels through the dithering machinery? Ahhh, the answer to this answers another issue I'm working on. I have a document-spanning background pattern of lines and currently we draw it into the visibleRect. However I was thinking of re-implementing this as a single DPS path cached on the server, which the window would then clip as normal. Can anyone offer arguments as to which is a better solution? The overhead of the passing of the paths isn't much (simple stuff) but if the clipping stuff is going to do all the work anyway, maybe letting it do it and caching is faster? Maury
Message-ID: <3638E296.415DE7A6@pdgm.com> From: Tim <tim@pdgm.com> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Convert Objective-c to C Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Thu, 29 Oct 1998 21:48:07 GMT NNTP-Posting-Date: Thu, 29 Oct 1998 16:48:07 EDT Organization: PSINet Hi there, Is any way to convert objective-c to c? because I have write a new project using c, but old project is written by Objective-c, I am new in objectiv-c. If I can to convert objective-c to c, it is easy for to read it. Thanks! TY
Message-ID: <36393C3F.D74699D9@ne.mediaone.net> From: Monty Brandenberg <mcbinc@ne.mediaone.net> Organization: MCB, Inc. MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Re: Convert Objective-c to C References: <3638E296.415DE7A6@pdgm.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Thu, 29 Oct 1998 23:10:39 -0500 NNTP-Posting-Date: Thu, 29 Oct 1998 23:10:16 EDT Tim wrote: > > Is any way to convert objective-c to c? You might look at the Portable Object Compiler. I don't have a pointer but the author of one doc I do have is David Stes. If that's correct, it should be a fairly easy search. My recollection is that C is one intermediate language on the compiler's path to an executable. Whether you'd want to use that C source as a basis for human programming is an open question... m -- Monty Brandenberg, Software Consultant MCB, Inc. mcbinc@ne.mediaone.net.MAPS.SPAM P.O. Box 426188 mcbinc@world.std.com.MAPS.SPAM Cambridge, MA 02142 617.864.6907
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: cfs - NFS server on NS/OS, lockups and kernel freezes Date: 29 Oct 98 08:43:58 Organization: Is a sign of weakness Message-ID: <SCOTT.98Oct29084358@slave.doubleu.com> References: <7183er$mf3$1@news.seicom.net> In-reply-to: no@spam.please's message of 28 Oct 1998 21:48:11 GMT In article <7183er$mf3$1@news.seicom.net> no@spam.please writes: if I copy over large files (say > 2 MByte) to the cfs'd filesystem it sometimes locks up in the cfsd deamom, the lock happens inside the write() call in the kernel that writes the encrypted data to disk?! This effectively freezes the system. <...> I (and others, Christian Starkjohann -the author of vmount and sharity which use a similar mechanism - was very helpful) suspect the problem is inside the buggy kernel NFS implementation, however if someone else has an idee what I could do the improve the behaviour I really would like to hear your suggestions. A reasonable way to check this out would be to run CFS on another system, say, a Linux system, and have the NeXT system mounting using the different port on that system. That's not clear. Don't point the NeXT NFS client at the Linux NFS server, point the NeXT NFS client at the Linux CFS variant on NFS. No encryption over the wire, but this is just a test, after all. That way you can test if it's a buggy interaction between the NeXT client and the CFS server, or if something buggy in the CFS server itself. As a further test, invert the scenario. Run the CFS server on the NeXT machine, with the Linux machine mounting it. That way NFS isn't involved on the NeXT side of things. I'm not sure about your characterization of the problem as a "buggy kernel NFS implementation". When doing a write(2) within cfsd, I would assume that you're writing to the _real_ filesystem, not to an NFS mounted filesystem? So the bug is as indicated here: write(2) -> kernel NFS proxy -> TCP/IP -> CFS server -> write(2) -> filesystem ^ bug ^ If so, then the bugginess or non-bugginess of the NeXT NFS kernel is a non issue. If, on the other hand, the bug is in the first write(2) call, you might have a point. But that should be easy to test by simply mounting a "normal" NFS filesystem and trying the same operation on it. Now, if the filesystem at the end of the above list is itself an NFS filesystem, then your guess at the bug's nature is probably correct. In that case, you're trying to write to an NFS filesystem while implementing another write which is also trying to write to an NFS filesystem. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: Just want to verify my understanding of NSBitmapImageRep. Date: 30 Oct 1998 13:42:20 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn73jgi0.5tt.rog@talisker.ohm.york.ac.uk> References: <SCOTT.98Oct27204912@slave.doubleu.com> <36377849.7267ABF8@ncal.verio.com> <SCOTT.98Oct29151609@slave.doubleu.com> NNTP-Posting-Date: 30 Oct 1998 13:42:20 GMT On 29 Oct 98 15:16:09, Scott Hess <scott@nospam.doubleu.com> wrote: > In article <slrn73h947.452.rog@talisker.ohm.york.ac.uk>, > i've always understood (but i'm likely wrong here - please correct > me!) that when blitting images under openstep, you're passing > stuff between two sides of the user->ps interpreter divide. an > NSBitmapImageRep holds its bitmap data in user space (hence > -bitmapdata is a very cheap call) and when called upon to draw > itself, all the data is passed down a port to the PS interpreter > and rendered by the image or colorimage operators. that's where it > gets all its funky scaling and dithering stuff from. > > This is where NSDrawBitmap() and NSBitmapImageRep have the advantage > over the "image" operator. Using the image operator implies that > you're streaming the bitmap data over the DPS port as you describe. > NSBitmapImageRep somehow manages to use shared memory to ship the data > over. So instead of marshall, stream, unmarshall, "image", it can > just skip straight to "image". well, the bitmap data has got to get to the other task somehow... does the pswrap stuff not do this automagically? i'd sort of assumed that if you passed a sufficiently large buffer then the postscript streaming operators (e.g. DPSWriteStringChars) would automatically do the memory map thing for you. > Effectively, this means that it makes very little difference what > volume of data you're sending using NSBitmapImageRep. Performance > seems almost completely determined by how much work the windowserver > has to do with the data once it gets it. and of course the overhead of passing the operation itself down the pipe, which isn't insignificant if you're doing a lot of them. > on the other hand, a composite operation happens purely on the PS > interpreter side, which is presumably why large bitmap composites > can freeze the window system as only one composite can be happening > at a time. > > Hmm, I think that might be the inverse. One seldom sees composite > operations that freeze the windowserver, except perhaps when doing > something like scaling. On the other hand, programs routinely freeze > the windowserver when they pass huge ill-formatted bitmaps via > NSBitmapImageRep and kin. [On my PPro200, I can ship over 200 > 1152x864 images at 888/32 into a fullscreen window in just over 30 > seconds. If I drop that to 444/16 (which matches my display depth), > that drops to 7 seconds. This was bracketing the draws with > -lockFocus/-unlockFocus, then a PSFlush(), and a PSWait() at the end > of the loop. So I'm fairly certain that most windowserver freezes > aren't really necessary :-).] well, i found my windowserver freezes happening when i was printing out a bitmap that needed some transparency. i was doing it at a higher resolution so it would look ok, so i was probably compositing something of the order of a 5120x4096x32bit image - that seems to freeze the window server quite nicely thankyou :-) i don't do that anymore (for that very reason) - i divide the image into tiles of a fixed size; this solves the problem. BTW, how would you do a scaling operation using compositing? i thought the point about composites was that they didn't do any scaling at all. > i tried to bypass the display mechanism to do my own > "updatedirtyrects" which was more intelligent, but under yellowbox > if you don't update the window through the display methods then the > window cache isn't updated properly (yeuch!). > <...> > i never did find a satisfactory solution. it's lucky i wasn't > programming a video game! > > You can probably do it, but you can't necessarily rely on the > automagic stuff. Basically, I would just collect my direct rects > internally, and then when I want to flush them do a -lockFocus, draw, > draw, draw, and flush, without relying on -drawRect: to be called. Of > course, you must draw the same exact view on -drawRect:... > > This won't quite work with an NSBuffered window, though, because DPS > will collect where you drew, and flush the union on -flushWindow. If > the drawing is complex, an NSRetained window will let the user see all > of the in-process drawing. i don't mind flushing the whole window - it's the redrawing of the view that's often time consuming - but under NT, you *have* to do your drawing inside drawRect, otherwise the offscreen buffer doesn't seem to get updated properly, and if you drag other windows over the window the old contents get displayed... > You could perhaps do it by rolling your own buffered window. Make the > window NSRetained, and do your drawing into an NSImage. Once all > drawing is complete, send a bunch of composite messages to composite > the fresh rects from the NSImage backing into the real window. To > make things look seemless, it would naturally be best to minimize the > number of composites as much as possible. i've thought about doing that (and even implemented something similar) but i found that it's quite difficult to make it transparent and seamless to the objects involved, particularly when they are objects that interact with the user. if you've got a window buffered in the style above that contains a text field, it can confuse it mightily if the view it's drawing into isn't a subview of the window it's contained within. if you make it a subview of the off-screen cache, then you have to do some fairly dodgy trickery to ensure that mouse interaction is handled correctly... in the end i decided that it was more trouble than it was worth, and abandoned the experiment. it would all have been so easy if the interface provided some way of sneaking some code in between setNeedsDisplayInRect: and drawRect:... cheers, rog.
From: Maurice le Rutte <MleRutte@square.nl> Newsgroups: comp.sys.next.programmer Subject: Re: Question : Discovering Openstep a Developers Tutorial Date: Fri, 30 Oct 1998 15:59:29 +0100 Organization: Square B.V. Message-ID: <3639D451.A1532BBC@square.nl> References: <36379032.833C9117@ix.netcom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: bhi1@ix.netcom.com Assign \r in the inspector as the key for the button Maurice le Rutte. Rob Blessin wrote: >[...] > The missing NSReturnSign object in effect stalls my little appication >[...] -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto: mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+
From: Mike Paquette <mpaque@ncal.verio.com> Newsgroups: comp.sys.next.programmer Subject: Re: Just want to verify my understanding of NSBitmapImageRep. Date: Fri, 30 Oct 1998 11:58:15 -0800 Organization: Electronics Service Unit No. 16 Message-ID: <363A1A4C.63878DA0@ncal.verio.com> References: <SCOTT.98Oct27204912@slave.doubleu.com> <36377849.7267ABF8@ncal.verio.com> <SCOTT.98Oct29151609@slave.doubleu.com> <slrn73jgi0.5tt.rog@talisker.ohm.york.ac.uk> <F1nD3y.JHu@T-FCN.Net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Maury Markowitz wrote: > > In <slrn73jgi0.5tt.rog@talisker.ohm.york.ac.uk> Roger Peppe wrote: > > i don't do that anymore (for that very reason) - i divide the image > > into tiles of a fixed size; this solves the problem. > > Is the issue primarily one of the WindowServer or some other extra-app > portion starving for memory? It's pretty much an issue of starving the system by paging in lots of data, only a small portion of which is needed. For example, suppose I have a 10,000 by 10,000 pixel image in grayscale at 8 bits per pixel. My viewing app is showing me a 100 by 100 pixel region of the image, suitably convolved and scaled so as to see what sort interesting things are going on... Assume that my big image is stored as a single raster of 10,000 bytes per scanline. I want to scroll my view of the image down by 50 lines, so I need to bring in another 50 scanlines of data. On a machine with a 4096 byte page size, this means that I may touch 1-2 pages per scanline for a total of 50-100 pages as I read the source image. If memory is tight, these may result in disk accesses for each page. The scroll in question may produce 50-100 page faults. Ick. Assumem that the big image has been preprocessed into 250 by 250 pixel tiles. Now my scroll may touch up to 4 tiles, but each tile has about 16 scanlines per page. I assemble the new 50 by 100 pixel area I've scrolled into by compositing from the four tiles, touching 12-14 pages. Even in a memory starved configuration, I produce 12 page faults instead of 50, for a 4x scrolling performance improvement. During repeated scrolls, the working set of memory touched remains much smaller, resulting in improved performance on low memory systems.
Newsgroups: comp.sys.next.programmer From: maury@remove_this.istar.ca (Maury Markowitz) Subject: Re: cfs - NFS server on NS/OS, lockups and kernel freezes Content-Type: text/plain; charset=us-ascii Message-ID: <F1nCHF.J5D@T-FCN.Net> Sender: news@T-FCN.Net Content-Transfer-Encoding: 7bit Cc: Christian Starkjohann Organization: needs one References: <7183er$mf3$1@news.seicom.net> <SCOTT.98Oct29084358@slave.doubleu.com> <3639b2e6.0@news.telekabel.at> Mime-Version: 1.0 Date: Fri, 30 Oct 1998 15:32:02 GMT In <3639b2e6.0@news.telekabel.at> Christian Starkjohann wrote: > It's not a bug in the classical sense. It's a deadlock. The write to the > faked NFS file requires a write to the real filesystem. And there are > locks > in the kernel that hit under certain circumstances. That's a problem on > NeXT, > but not on any other platform I know. DR2? Has this been reported up the official channels? Maury
Newsgroups: comp.sys.next.programmer From: maury@remove_this.istar.ca (Maury Markowitz) Subject: Re: Just want to verify my understanding of NSBitmapImageRep. Content-Type: text/plain; charset=us-ascii Message-ID: <F1nD3y.JHu@T-FCN.Net> Sender: news@T-FCN.Net Content-Transfer-Encoding: 7bit Cc: rog@ohm.york.ac.uk Organization: needs one References: <SCOTT.98Oct27204912@slave.doubleu.com> <36377849.7267ABF8@ncal.verio.com> <SCOTT.98Oct29151609@slave.doubleu.com> <slrn73jgi0.5tt.rog@talisker.ohm.york.ac.uk> Mime-Version: 1.0 Date: Fri, 30 Oct 1998 15:45:33 GMT In <slrn73jgi0.5tt.rog@talisker.ohm.york.ac.uk> Roger Peppe wrote: > i don't do that anymore (for that very reason) - i divide the image > into tiles of a fixed size; this solves the problem. Is the issue primarily one of the WindowServer or some other extra-app portion starving for memory? > BTW, how would you do a scaling operation using compositing? > i thought the point about composites was that they didn't do any > scaling at all. Or rotation, much to my chagrin. I had build a wonderful tiling system for Glyph!X that just _flew_ and then realized it didn't rotate! > i don't mind flushing the whole window - it's the redrawing of the > view that's often time consuming - but under NT, you *have* to do > your drawing inside drawRect, otherwise the offscreen buffer doesn't > seem to get updated properly, and if you drag other windows over > the window the old contents get displayed... Oh! Maury
Newsgroups: comp.sys.next.programmer From: fgalot@x-lan.alienor.fr Subject: Re: Creating NSPrintInfo with NSDictionary Message-ID: <F1nF7v.3wz@x-lan.alienor.fr> Sender: news@x-lan.alienor.fr Organization: x&lan References: <3636DA2E.71AB0C2E@square.nl> Date: Fri, 30 Oct 1998 16:31:07 GMT Hello. Some times ago I had some troubles with the NSPrintInfo. I wanted to fix the paper feed but I didn't manage to do it with the dictionary in the printinfo. The only way I found was rewriting the addToPageSetup or EndSetup View Methods to add my PS code. There is an example in the documentation. -- --------------------------------------- ® ® | ® O_O ® ® | O_O -+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+- Fred Galot fgalot@x-lan.alienor.fr
From: armin@devicen.de Newsgroups: comp.sys.next.programmer Subject: building egcs 1.1b on OPENSTEP 4.2 ? Date: Fri, 30 Oct 1998 19:32:59 +0100 Organization: DEUTSCHE MESSE AG Message-ID: <armin-3010981932590001@grafix.devicen.de> Did anybody got "egcs 1.1b" to compile on OPENSTEP 4.2/Intel? If so, then how can it be done? I don't need "egcs" as an replacement for OPENSTEP's "cc". But i would like to use "ecgs" for compiling command line apps. Thanks in advantage. --------------------------------------------------------------------- Armin Retzko DEVICE/N GmbH - Hannover, Germany e-mail at DEVICE/N: armin@devicen.de
From: Mike Paquette <mpaque@ncal.verio.com> Newsgroups: comp.sys.next.programmer Subject: Re: Just want to verify my understanding of NSBitmapImageRep. Date: Fri, 30 Oct 1998 11:58:18 -0800 Organization: Electronics Service Unit No. 16 Message-ID: <363A1A4F.2EE95ED1@ncal.verio.com> References: <SCOTT.98Oct27204912@slave.doubleu.com> <36377849.7267ABF8@ncal.verio.com> <SCOTT.98Oct29071201@slave.doubleu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Scott Hess wrote: > In article <36377849.7267ABF8@ncal.verio.com>, > Mike Paquette <mpaque@ncal.verio.com> writes: > You can always initialize a NSBitmapImageRep pointing into a > sub-rectangle of another bitmap. This ultimately has the same > effect as just compositing a rect of the image, though. Copying > the data to a new bitmap is rarely efficient beyond the smallest > images. > > I can do that if I already have image data in the windowserver, I can > suck a rectangle into an NSBitmapImageRep using > -initWithFocusedViewRect:. I can also take an NSBitmapImageRep and > draw it within an arbitrary rectangle with -drawInRect:, which will > scale as necessary. What I can't appear to do with what I can find in > the Rhapsody AppKit docs is pick a specific rectangle out of the > NSBitmapImageRep I have and draw it one-to-one in the focusView. Those darn AppKit folks! This used to be in there somewhere. It could probably be faked up by setting up a view for clipping and then using -drawAtPoint:, but that's sort of expensive timewise. Then again, you could cheat and use NSDrawBitmap(), a C function that you could describe a subset of your bitmap to for purposes of doing quick redraws of subrects. If you want to play nice, construct a new NSBitmapImageRep that points into the subrect you want to draw from the 'master' rect. This will then invoke the same things that NSDrawBitmap would, with the appropriate parameters. --- NSDrawBitmap Summary: This function draws a bitmap image. Declared in: AppKit/NSGraphics.h Synopsis: void NSDrawBitmap(const NSRect *rect, int pixelsWide, int pixelsHigh, int bitsPerSample, int samplesPerPixel, int bitsPerPixel, int bytesPerRow, BOOL isPlanar, BOOL hasAlpha, NSColorSpace colorSpace, const unsigned char *const data[5]) Warning: This function is marginally obsolete. Most applications are better served using the NSBitmapImageRep class to read and display bitmap images. Description: The NSDrawBitmap function renders an image from a bitmap, binary data that describes the pixel values for the image (this function replaces NSImageBitmap). NSDrawBitmap renders a bitmap image using an appropriate PostScript operator-image, colorimage, or alphaimage. It puts the image in the rectangular area specified by its first argument, rect; the rectangle is specified in the current coordinate system and is located in the current window. The next two arguments, pixelsWide and pixelsHigh, give the width and height of the image in pixels. If either of these dimensions is larger or smaller than the corresponding dimension of the destination rectangle, the image will be scaled to fit. The remaining arguments to NSDrawBitmap describe the bitmap data, as explained in the following paragraphs. bitsPerSample is the number of bits per sample for each pixel and samplesPerPixel is the number of samples per pixel. bitsPerPixel is based on samplesPerPixel and the configuration of the bitmap: if the configuration is planar, then the value of bitsPerPixel should equal the value of bitsPerSample; if the configuration isn't planar (is meshed instead), bitsPerPixel should equal bitsPerSample * samplesPerPixel. bytesPerRow is calculated in one of two ways, depending on the configuration of the image data (data configuration is described below). If the data is planar, bytesPerRow is (7 + (pixelsWide * bitsPerSample)) / 8. If the data is meshed, bytesPerRow is (7 + (pixelsWide * bitsPerSample * samplesPerPixel)) / 8. A sample is data that describes one component of a pixel. In an RGB color system, the red, green, and blue components of a color are specified as separate samples, as are the cyan, magenta, yellow, and black components in a CMYK system. Color values in a gray scale are a single sample. Alpha values that determine transparency and opaqueness are specified as a coverage sample separate from color. In bitmap images with alpha, the color (or gray) components have to be premultiplied with the alpha. This is the way images with alpha are displayed, this is the way they are read back, and this is the way they are stored in TIFFs. isPlanar refers to the way data is configured in the bitmap. This flag should be set YES if a separate data channel is used for each sample. The function provides for up to five channels, data1, data2, data3, data4, and data5. It should be set NO if sample values are interwoven in a single channel (meshed); all values for one pixel are specified before values for the next pixel. Gray-scale windows store pixel data in planar configuration; color windows store it in meshed configuration. NSDrawBitmap can render meshed data in a planar window, or planar data in a meshed window. However, it's more efficient if the image has a depth (bitsPerSample) and configuration (isPlanar) that matches the window. hasAlpha indicates whether the image contains alpha. If it does, the number of samples should be 1 greater than the number of color components in the model (e.g., 4 for RGB). colorSpace can be NS_CustomColorSpace, indicating that the image data is to be interpreted according to the current color space in the PostScript graphics state. This allows for imaging using custom color spaces. The image parameters supplied as the other arguments should match what the color space is expecting. If the image data is planar, data[0] through data[samplesPerPixel-1] point to the planes; if the data is meshed, only data[0] needs to be set. --- > When updating a bitmap on the display, the real time sink can be > simply blitting a bunch of unchanged pixels through the depth > conversion code. Tracking dirty rects within your own bitmap and > only updating these can cut down on the time spent blitting. > > Say I have an 1152x864 NSBitmapImageRep, and wish to redraw the rect > at {{ x=200,y=150}, {width=350, height=250}}. Just looking at the > copy-to-server overhead, the difference between tossing it all up > there versus just what's necessary is minimal because it's using > shared memory. But since the server has to revisit all those pixels > doing dithering, it might be worthwhile to suck out the 350x250 rect > of pixels into a new NSBitmapImageRep and only blit those pixels. The server will only play with the subrect that you've told it about. It will just play with the width X height rect you describe, advancing by 'rowbytes - byteWidth' to the next scanline. > Now, for the have-my-cake-and-eat-it-too question: How does clipping > affect this? Say we have an 1152x864 NSView, with a same-sized > NSBitmapImageRep, with {{ x=200,y=150}, {width=350, height=250}} > needing redrawing. If I use DPS clipping operations to clip to that > rectangle, then blit the entire bitmap, does the windowserver still > push all pixels through the dithering machinery? Probably not. (There are ways to force all the pixels through the dithering code, but the software won't do this for common cases.) If you use NSDrawBitmap(), only the rect of interest will be dithered.
From: Mike Paquette <mpaque@ncal.verio.com> Newsgroups: comp.sys.next.programmer Subject: Re: Just want to verify my understanding of NSBitmapImageRep. Date: Fri, 30 Oct 1998 11:58:22 -0800 Organization: Electronics Service Unit No. 16 Message-ID: <363A1A53.F0DAE828@ncal.verio.com> References: <SCOTT.98Oct27204912@slave.doubleu.com> <36377849.7267ABF8@ncal.verio.com> <SCOTT.98Oct29151609@slave.doubleu.com> <slrn73jgi0.5tt.rog@talisker.ohm.york.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Roger Peppe wrote: > well, the bitmap data has got to get to the other task somehow... > does the pswrap stuff not do this automagically? i'd sort of assumed > that if you passed a sufficiently large buffer then the postscript > streaming operators (e.g. DPSWriteStringChars) would automatically > do the memory map thing for you. Bitmaps move through the Mach out of line data trick. It's quite efficient. This is also done almost any time there is a large amout of data (more than a fraction of a page). > well, i found my windowserver freezes happening when i was printing > out a bitmap that needed some transparency. i was doing it at a higher > resolution so it would look ok, so i was probably compositing something > of the order of a 5120x4096x32bit image - that seems to freeze the > window server quite nicely thankyou :-) Oog. That wasn't just the WindowServer freezing, that was every userspace task getting suspended while the VM system thrashed it's little brains out touching a couple of 80 Mbyte memory regions, PLUS the time needed to composite 20 million pixels. > i don't do that anymore (for that very reason) - i divide the image > into tiles of a fixed size; this solves the problem. Yup. When dealing with really big rasters, tiling is your friend. (Mmmmm... 9 channel 10,000 by 10,000 pixel LANDSAT data sets...)
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: cfs - NFS server on NS/OS, lockups and kernel freezes Date: 30 Oct 98 13:41:34 Organization: Is a sign of weakness Message-ID: <SCOTT.98Oct30134134@slave.doubleu.com> References: <7183er$mf3$1@news.seicom.net> <SCOTT.98Oct29084358@slave.doubleu.com> <3639b2e6.0@news.telekabel.at> In-reply-to: Christian Starkjohann's message of 30 Oct 1998 13:36:54 +0100 In article <3639b2e6.0@news.telekabel.at>, Christian Starkjohann writes: It's not a bug in the classical sense. It's a deadlock. The write to the faked NFS file requires a write to the real filesystem. And there are locks in the kernel that hit under certain circumstances. That's a problem on NeXT, but not on any other platform I know. Ah. So buffer the write for later... Actually, I'm not sure that I haven't run a userland NFS server serving files from the local filesystem to itself on a NeXT box before. I'm not positive I _have_, though, because this would have been a couple years ago. And, of course, it's been too long to say whether there were bugs with it, I just don't recall... I do seem to recall there being some problems with getting self-mounted under /Net/. That might be related. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: stes@mundivia.es Newsgroups: comp.sys.next.programmer Subject: Re: Convert Objective-c to C Date: Fri, 30 Oct 1998 21:48:32 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <71dc7g$ra9$1@nnrp1.dejanews.com> References: <3638E296.415DE7A6@pdgm.com> <36393C3F.D74699D9@ne.mediaone.net> In article <36393C3F.D74699D9@ne.mediaone.net>, mcbinc@ne.mediaone.net wrote: > Tim wrote: > > > > Is any way to convert objective-c to c? > > You might look at the Portable Object Compiler. Correct ... The POC is a utility that converts objective-c to C. Another solution would be to buy a copy of the Stepstone compiler (http://www.stepstn.com) which is also doing this. You can get the POC from http://sunsite.unc.edu/pub/Linux/devel/lang/objc/. You need 2 packages : the objc-1.10.5-bootstrap.tar.gz package and the objc-1.10.5.tar.gz package. Install the bootstrap package first. . -----------== Posted via Deja News, The Discussion Network ==---------- http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
Newsgroups: comp.sys.next.programmer Subject: Re: cfs - NFS server on NS/OS, lockups and kernel freezes References: <7183er$mf3$1@news.seicom.net> <SCOTT.98Oct29084358@slave.doubleu.com> From: Christian Starkjohann Organisation: Objective Development Message-ID: <3639b2e6.0@news.telekabel.at> Date: 30 Oct 1998 13:36:54 +0100 scott@nospam.doubleu.com (Scott Hess) wrote: > In article <7183er$mf3$1@news.seicom.net> no@spam.please writes: > if I copy over large files (say > 2 MByte) to the cfs'd filesystem > it sometimes locks up in the cfsd deamom, the lock happens inside > the write() call in the kernel that writes the encrypted data to > disk?! This effectively freezes the system. > <...> > I (and others, Christian Starkjohann -the author of vmount and > sharity which use a similar mechanism - was very helpful) suspect > the problem is inside the buggy kernel NFS implementation, however > if someone else has an idee what I could do the improve the > behaviour I really would like to hear your suggestions. > > A reasonable way to check this out would be to run CFS on another > system, say, a Linux system, and have the NeXT system mounting using > the different port on that system. > > That's not clear. Don't point the NeXT NFS client at the Linux NFS > server, point the NeXT NFS client at the Linux CFS variant on NFS. No > encryption over the wire, but this is just a test, after all. That > way you can test if it's a buggy interaction between the NeXT client > and the CFS server, or if something buggy in the CFS server itself. > > As a further test, invert the scenario. Run the CFS server on the > NeXT machine, with the Linux machine mounting it. That way NFS isn't > involved on the NeXT side of things. > [...] It's not a bug in the classical sense. It's a deadlock. The write to the faked NFS file requires a write to the real filesystem. And there are locks in the kernel that hit under certain circumstances. That's a problem on NeXT, but not on any other platform I know. -- Christian Starkjohann mail: <cs -AT- obdev.at> or web: http://www.obdev.at/
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: Just want to verify my understanding of NSBitmapImageRep. Date: 30 Oct 1998 13:46:44 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn73jgq7.5tt.rog@talisker.ohm.york.ac.uk> References: <SCOTT.98Oct27204912@slave.doubleu.com> <36377849.7267ABF8@ncal.verio.com> <SCOTT.98Oct29083341@slave.doubleu.com> NNTP-Posting-Date: 30 Oct 1998 13:46:44 GMT On 29 Oct 98 08:33:41, Scott Hess <scott@nospam.doubleu.com> wrote: > In article <SCOTT.98Oct29071201@slave.doubleu.com>, > scott@nospam.doubleu.com (Scott Hess) writes: > Now, for the have-my-cake-and-eat-it-too question: How does > clipping affect this? Say we have an 1152x864 NSView, with a > same-sized NSBitmapImageRep, with {{ x=200,y=150}, {width=350, > height=250}} needing redrawing. If I use DPS clipping operations > to clip to that rectangle, then blit the entire bitmap, does the > windowserver still push all pixels through the dithering machinery? > > Well, empirical testing indicates that given the above bitmap and > view, clipping makes the blit about 10x faster. The speedup does not > seem quite linear WRT how many pixels you're clipping off, but it's > close. [It's hard for me to tell, because since I was measuring DPS, > I couldn't use CPU time, I had to use "real" or "wall" time. I did > have the computer measuring the time, though:-).] have you tried it under NT yellowbox? i seem to recall having a clipped bitmap blit that was very fast under openstep, but not very fast at all under NT. (i hate having to mention NT, but unfortunately we're required to deliver NT executables :-( ) > I also found that blitting at 444/16 is about 4x faster than 888/32 on > my machine (PPro200 with Matrox Millenium II using an older driver at > 444/16). I suppose I'll have to test it with a 555/16 driver, too, > just in case... it's difficult to know what to do if you want a decent "all-round" performance without bloating code by including code for dealing with all possible display depths. in the end, you'll end up writing code as nasty as if you'd used Interceptor... :-) cheers, rog.
Newsgroups: comp.sys.next.programmer From: maury@remove_this.istar.ca (Maury Markowitz) Subject: Re: Just want to verify my understanding of NSBitmapImageRep. Content-Type: text/plain; charset=us-ascii Message-ID: <F1o0K9.9n4@T-FCN.Net> Sender: news@T-FCN.Net Content-Transfer-Encoding: 7bit Cc: mpaque@ncal.verio.com Organization: needs one References: <SCOTT.98Oct27204912@slave.doubleu.com> <36377849.7267ABF8@ncal.verio.com> <SCOTT.98Oct29151609@slave.doubleu.com> <slrn73jgi0.5tt.rog@talisker.ohm.york.ac.uk> <F1nD3y.JHu@T-FCN.Net> <363A1A4C.63878DA0@ncal.verio.com> Mime-Version: 1.0 Date: Sat, 31 Oct 1998 00:12:08 GMT In <363A1A4C.63878DA0@ncal.verio.com> Mike Paquette wrote: > Assume that my big image is stored as a single raster of 10,000 bytes per > scanline. I want to scroll my view of the image down by 50 lines, so I > need to > bring in another 50 scanlines of data. On a machine with a 4096 byte page > size, this means that I may touch 1-2 pages per scanline for a total of > 50-100 > pages as I read the source image. If memory is tight, these may result in > disk accesses for each page. The scroll in question may produce 50-100 > page > faults. Ick. Right, I get it now. Hmmm, I suppose I'll have to do this for my background work then. In other news, taking a tip from the thread here, I decided to re-implement our grid drawing routines today and now they are super-large MUPs cached on the DPS side. Since they're sparce they don't tack up much memory, so I'm not terribly worried about the effects above. I can't say it's a lot faster (I'm on a G3 now [finally!], "faster" is hard to measure) but it seems to be better and the implementation is certainly cleaner. Then I re-coded some stuff that draws borders around pages in the same way. Amazing what you can do in 10 minutes in DR2! Maury
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: NS3.3 DO and oneway method dispatch exceptions. Date: 30 Oct 98 12:57:28 Organization: Is a sign of weakness Distribution: world Message-ID: <SCOTT.98Oct30125728@slave.doubleu.com> Continuing my recent spate of oddball questions (so I'm on a see-what-I-might've-missed-in-the-past-couple-years tour of NeXTSTEP and OpenStep, sue me)... I have a multi-threaded application, where the subthread is handling incoming data on a file descriptor. The entire goal of making it multi-threaded was so that I could just read the data in the subthread without having to work in the main event loop handling partial reads and all. But, in the interests of keeping things simple, I'm using DO to actually communicate certain notifications to the main thread. The subtle problem is that since my subthread is in a loop running read(2), it's _not_ in one of the DO -run* methods. Basically, it reads header info, reads the info the header calls for, and then calls a DO method to tell the main thread what it got, then goes back to reading. Under OpenStep4.x, this seems to work fine. I just made all the methods (oneway void), set the protocol correctly, and the subthread and main thread go their merry way, with no concerns. Under NS3.3, though, I keep getting "exception during dispatch:" messages in the main thread, and long pauses in it's handling of messages. The long pauses I fixed by doing -setOutTimeout:0. No more pauses, but lots of "exception during dispatch:" output. I figured out that the NS3.3 DO wants to send a reply back to the sender, so... in the subthread I added some calls to -runWithTimeout:0. Now, I no longer get the dispatch exception - instead, I get "[NXConnection run] - tossing received reply msg". Sigh. This isn't a big problem. I'm comfortable writing off the NS3.3 version - or at least changing the (oneway void) to something like (BOOL), which should naturally force a round trip on every message, albeit with a slight slowdown. I was just wondering if anyone knows how to solve the annoyance without changing _my_ code. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: no@spam.please Newsgroups: comp.sys.next.programmer Subject: Re: cfs - NFS server on NS/OS, lockups and kernel freezes Date: 31 Oct 1998 03:51:41 GMT Organization: Seicom GmbH, Reutlingen, Germany, Europe Message-ID: <71e1gd$hcj$1@news.seicom.net> References: <7183er$mf3$1@news.seicom.net> <SCOTT.98Oct29084358@slave.doubleu.com> <3639b2e6.0@news.telekabel.at> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: Christian Starkjohann In <3639b2e6.0@news.telekabel.at> Christian Starkjohann wrote: > > As a further test, invert the scenario. Run the CFS server on the > > NeXT machine, with the Linux machine mounting it. That way NFS isn't > > involved on the NeXT side of things. > > [...] Since CFS is a local only client (it does not even accept network connections) this may need so twiddling.... > It's not a bug in the classical sense. It's a deadlock. The write to the > faked NFS file requires a write to the real filesystem. And there are locks > in the kernel that hit under certain circumstances. That's a problem on NeXT, > but not on any other platform I know. > I have now some more detailed information, looks like a mutex in the kernel is not properly resetted when some system calls in a certain pattern are issued from a process which is a nfs userlevel server. I have no idea yet how the kernel knows about this particular connection but it is definitely reproduceable. A bad 'call' sequence inside the nfs server seems to be 'open', 'fstat', 'write', 'lseek', but only under heavy load, for just a few blocks it works but it fails for longer files (more blocks, more calls in sequence). Anyway it sucks, it took me > 50 restarts with full fsck to evaluate this. I am now working the kernel to find some more hard evidence. I know this may certainly chance in MacOS X Server but I really *like* to know whats going on and how to circumvent the deadlocks. -- Frank M. Siegert [frank@this.net] [frank@wizards.de]
From: Pascal Bourguignon <pjb@imaginet.fr> Newsgroups: comp.sys.next.programmer Subject: Re: Convert Objective-c to C Date: Sun, 01 Nov 1998 01:23:49 -0100 Organization: Latymer Designs Ltd. Message-ID: <363BC635.33D66479@imaginet.fr> References: <3638E296.415DE7A6@pdgm.com> <36393C3F.D74699D9@ne.mediaone.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Monty Brandenberg wrote: > > Tim wrote: > > > > Is any way to convert objective-c to c? > > You might look at the Portable Object Compiler. I don't have > a pointer but the author of one doc I do have is David Stes. > If that's correct, it should be a fairly easy search. My > recollection is that C is one intermediate language on the > compiler's path to an executable. Whether you'd want to use that > C source as a basis for human programming is an open question... > > m Yes. I would add that the C code generated is quite readable, because it uses C functions to implement object semantic. However, in a plain C environment, you will need to add the implementation of these Objective-C runtime functions. __Pascal Bourguignon__
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <18198909288022@digifix.com> Date: 1 Nov 1998 04:47:06 GMT Organization: Digital Fix Development Message-ID: <2557909896420@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1994. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: candlish@efn.org (John Candlish) Newsgroups: comp.sys.next.programmer Subject: Re: Print mechanism and data path Date: 1 Nov 1998 11:32:02 -0800 Organization: Oregon Public Networking: Eugene Freenet Message-ID: <71icvi$a6s@garcia.efn.org> References: <DKvMGNNTYhYH@cc.usu.edu> <70uskt$8ju$1@news.idiom.com> Hints for using the DPS RIP can be gleened from iwf.2.0.N.bs.tar.gz and DeskJet_2.03.tar.gz. Both these are based on the WindowServer haques of Eric Scott. Good Luck jCandlish .
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <19981030042335.12769.qmail@hades.rpini.com> Control: cancel <19981030042335.12769.qmail@hades.rpini.com> Date: 02 Nov 1998 07:52:09 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.19981030042335.12769.qmail@hades.rpini.com> Sender: Anonymous <nobody@remailer.ch> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: bellj@mikado.dialup.ais.net (J. Shan Bell) Newsgroups: comp.sys.next.programmer Subject: Reading Diagram v2.0 files Date: 2 Nov 1998 19:44:22 GMT Organization: EnterAct L.L.C. Turbo-Elite News Server Message-ID: <71l22m$23h$1@eve.enteract.com> Keywords: diagram parser I'm looking for a parser for Diagram 2.0 files. The format is well documented and in ASCII. Is there a lex/yacc or hand coded parser for these files. Any information appreciated. ---- J. Shan Bell NeXTmail/MIME and PGP encrypted mail encouraged Software Engineer finger bellj@cs.indiana.edu for my PGP public key at large http://www.cs.indiana.edu/hyplan/bellj.html
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: Just want to verify my understanding of NSBitmapImageRep. Date: 2 Nov 1998 13:38:36 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn73rdf0.711.rog@talisker.ohm.york.ac.uk> References: <SCOTT.98Oct27204912@slave.doubleu.com> <36377849.7267ABF8@ncal.verio.com> <SCOTT.98Oct30133709@slave.doubleu.com> NNTP-Posting-Date: 2 Nov 1998 13:38:36 GMT On 30 Oct 98 13:37:09, Scott Hess <scott@nospam.doubleu.com> wrote: > In article <slrn73jgi0.5tt.rog@talisker.ohm.york.ac.uk>, > rog@ohm.york.ac.uk (Roger Peppe) writes: > On 29 Oct 98 15:16:09, Scott Hess <scott@nospam.doubleu.com> wrote: > well, the bitmap data has got to get to the other task somehow... > does the pswrap stuff not do this automagically? i'd sort of > assumed that if you passed a sufficiently large buffer then the > postscript streaming operators (e.g. DPSWriteStringChars) would > automatically do the memory map thing for you. > > I think the actual communications between the DPS client and the DPS > server is all memory-mapped. I _don't_ think that the creation of the > wrap structure itself will use memory mapping, though. I suppose it's > possible, though. i just wondered how the NSDrawBitmap operator managed to bypass the conventional user-space to PS interpreter-space channeling, and whether that bypass was accessible to the regular user. (could be useful) > BTW, how would you do a scaling operation using compositing? i > thought the point about composites was that they didn't do any > scaling at all. > > I'm not sure what your question is. -lockFocus on a device, > PSscale(), and then draw the image? i was referring to your previous comment: : One seldom sees composite operations that freeze the windowserver, : except perhaps when doing something like scaling. i wasn't quite sure what you meant by the above (i'm always looking to learn something new!) because as far as i'm aware there's no composite operator that does any scaling. > i don't mind flushing the whole window - it's the redrawing of the > view that's often time consuming - but under NT, you *have* to do > your drawing inside drawRect, otherwise the offscreen buffer > doesn't seem to get updated properly, and if you drag other windows > over the window the old contents get displayed... > > Huh? The project that's been paying my rent for three years draws in > a self-initiated fashion all the time, and works fine under NT. It > _does_ use the same code from the self-initiated refresh as from a > -drawRect:, so the output is consistent. It's essentially a word > processor, and while the user is keyboarding or using the mouse to > select text, it's doing it's own drawing without involving any of the > NSView machinery, basically something like: > > [self lockFocus]; > [[self window] disableFlushWindow]; > // Draw using a bunch of PS* operators and PSwraps. > [[self window] enableFlushWindow]; > [[self window] flushWindow]; > PSWait(); > [self unlockFocus]; > > I've never had problems with code like that. It worked under > NeXTSTEP, it works under OpenStep/Mach and NT, and it works under > Rhapsody DR2 for Intel. i was doing something exactly the same, and it *looked* as if it was working fine, until one day someone dragged a window on top of this yellowbox window and i discovered that the data wasn't being cached properly. when i changed it so it did the drawing inside the drawRect method, the problem went away. i put it down as just one of those little "gotchas" that one finds all the time, but maybe it was an actual bug. (i'm _fairly_ sure it wasn't a bug in my code :-]). > it would all have been so easy if the interface provided some way > of sneaking some code in between setNeedsDisplayInRect: and > drawRect:... > > Why not? Override -setNeedsDisplayInRect: to collect the rects. In > your version of -drawRect:, save the gstate, clip the drawing rect to > nothing, call [super -drawRect:], restore the gstate, and do whatever > you wanted to do with those dirty rects. the problem is that you don't know why drawRect: is being called. i couldn't be sure that when, for instance, a scrollview scrolls its contentview it doesn't do something like: PScomposite(old window bitmap data to new position); [contentview lockFocus]; [contentview drawRect:newly_exposed_rect]; [contentview unlockFocus]; which would invalidate your proposed scheme. even none of the openstep objects use drawRect: like that now, the documentation doesn't assure me that they never will, so if i want to keep my code robust, i can't break that box... > Personally, though, I suspect there's something else you're seeing. > As indicated above, I work outside this particular box all the time > without problems. you're probably right - but i'll try and knock up a little demo that demonstrates the problem. cheers, rog.
Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <130943519254225408@hotmail.com> ignore no reply Control: cancel <130943519254225408@hotmail.com> Message-ID: <cancel.130943519254225408@hotmail.com> Date: Mon, 02 Nov 1998 17:43:09 +0000 Sender: qjrtuqdr@hotmail.com From: andrew@erlenstar.demon.co.uk Organization: Annihilator v0.3 Spam (EMP) cancelled - type=NAPRO
From: jmeacham@wittgenstein.jhuccp.org (James D. Meacham) Newsgroups: comp.sys.next.programmer,comp.sys.next.software Subject: Please, Please, Please, Msql EOF adapter needed Date: 29 Oct 1998 17:24:33 GMT Organization: The Center for Communications Programs of the Johns Hopkins University Distribution: world Message-ID: <71a8ch$152@news.jhu.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi All, After hours and hours of trying, I've come to the conclusion that for whatever reason, I simply cannot get the Msql EOF adaptor running on my NextTurboStation running OS/NS4.2. I will pay someone $20 to both e-mail me a working copy of this and send me a disk, if need be. I have EOF and WebObjects, and a big project I'm developing is using Msql, and there is no way around it. If some kind soul would take pity on a non-programmer and send this to me, I would be eternally grateful, and willing to send you the $20 bucks if you want it. Peace, James -- James David Meacham, 3rd jmeacham@wittgenstein.jhuccp.org Web Systems Administrator/Internet Services Coordinator Center for Communications Programs 410-659-6367 The Johns Hopkins University 410-659-6266 (fax)
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <71m4om$v27$6568@duke.telepac.pt> Control: cancel <71m4om$v27$6568@duke.telepac.pt> Date: 03 Nov 1998 05:43:38 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.71m4om$v27$6568@duke.telepac.pt> Sender: greatlinksite@mail.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
Newsgroups: comp.sys.next.programmer From: news@news.msfc.nasa.gov Message-ID: <cancel.71lu9c$bc$87@coranto.ucs.mun.ca> Control: cancel <71lu9c$bc$87@coranto.ucs.mun.ca> Subject: cmsg cancel <71lu9c$bc$87@coranto.ucs.mun.ca> Organization: http://www.msfc.nasa.gov/ Date: Tue, 3 Nov 1998 06:50:38 GMT Sender: hgibeiqe@somethingfunny.net Make Money Fast post canceled by J. Porter Clark.
Newsgroups: comp.sys.next.programmer From: mvlems@vbox.xs4all.nl.NO.SPAM Subject: NSFileManager movePath: toPath: OS4.2 'bug' Content-Type: text/plain; charset=us-ascii Message-ID: <1998Nov1.123515.6007@vbox.xs4all.nl.NO.SPAM> Sender: mvlems@vbox.xs4all.nl.NO.SPAM (Mark F. Vlems) Content-Transfer-Encoding: 7bit Mime-Version: 1.0 Date: Sun, 1 Nov 1998 12:35:15 GMT Howdy, According to the documentation, this method first copies the path to the new path, and then removes the old one. If the path is a directory, all subdirectories and files in it will be copied too. The implementation of this method knows some error checking, but does not check for a directory move to one of it's own subdirectories. Example: Existing path: /Users/me/dir1/dir2/file [fileManager movePath:@"/Users/me/dir1" toPath:@"/Users/me/dir1/dir2/dirNew"]; This call copies the dir1 to dir2, while changing dir1's name to dirNew. Then it removes the old dir1, including dirNew... It works according the documentation, so, maybe you can't call it a bug, but just undesired behavior. The workaround I use now is using methods like "stringByStandardizingPath" and "hasPrefix" to check if one is a subpath of another. Anybody with a more elegant solution? ThanksIA, -- Mark F. Vlems Poor is the man, MIME Mail OK Whose pleasures depend NeXTMail preferred! On the permission of another. Fax: +31 (0) 23 5622368 Madonna in "Justify my love", 1990
From: "Karsten D. Wolf" <karsten.d.wolf@erziehung.uni-giessen.de> Newsgroups: comp.sys.next.programmer Subject: Re: WO vs. ASP Date: Sun, 01 Nov 1998 15:02:49 +0100 Organization: =?iso-8859-1?Q?Justus=2DLiebig=2DUniversit=E4t=20Gie=DFen?= Message-ID: <363C69D1.AE24845@erziehung.uni-giessen.de> References: <361A7394.49687B8@indiana.no_spam.edu> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="------------6D47A82080B960CFB31060ED" This is a multi-part message in MIME format. --------------6D47A82080B960CFB31060ED Content-Type: text/plain; charset=us-ascii; x-mac-type="54455854"; x-mac-creator="4D4F5353" Content-Transfer-Encoding: 7bit Have a look at the WO-archive at www.omnigroup.com and search for ASP. -k Allan Streib wrote: > Could someone please give me impressions/pros/cons of developing Web > applications with WO compared to Microsoft's Active Server Pages? I > know the Dell web site used to be WO but now looks like ASP. I have not > used either but may have the opportunity here to push for one or the > other. Thanks for any info or references. > > Allan Streib > (remove no_spam from email address to reply) --------------6D47A82080B960CFB31060ED Content-Type: text/x-vcard; charset=us-ascii; name="karsten.d.wolf.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Karsten D. Wolf Content-Disposition: attachment; filename="karsten.d.wolf.vcf" begin:vcard n:Wolf;Karsten D. tel;fax:(49)-(0)641-99-24039 tel;home:(49)-(0)641-75944 tel;work:(49)-(0)641-99-24034 x-mozilla-html:TRUE org:Justus-Liebig-Universität Gießen;Arbeits-, Berufs- und Wirtschaftspädagogik version:2.1 email;internet:karsten.d.wolf@erziehung.uni-giessen.de title:Wissenschaftlicher Mitarbeiter adr;quoted-printable:;;JLU Giessen=0D=0AABW-P=8Adagogik=0D=0AKarl-Gl=9Ackner-Str. 21b;Giessen;Hessen;35394;Germany (Deutschland) x-mozilla-cpt:;1 fn:Karsten D. Wolf end:vcard --------------6D47A82080B960CFB31060ED--
From: Maurice le Rutte <MleRutte@square.nl> Newsgroups: comp.sys.next.programmer Subject: Re: Creating NSPrintInfo with NSDictionary Date: Mon, 02 Nov 1998 09:32:40 +0100 Organization: Square B.V. Message-ID: <363D6E28.7C5173D2@square.nl> References: <3636DA2E.71AB0C2E@square.nl> <F1nF7v.3wz@x-lan.alienor.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: fgalot@x-lan.alienor.fr I was afraid that it would end up doing things like that. Until now I've been able to stay away from the PostScript thing. That stupid NSPrintInfo starts with Letter instead of A4, we Europeans are always forgotten by those Americans. Also the NSPrintInfo does not allow A3 to be choosen if the default printer does not support it. I want my user to choose a page setup before choosing the printer. Maurice le Rutt.e fgalot@x-lan.alienor.fr wrote: > > Hello. > Some times ago I had some troubles with the NSPrintInfo. I wanted to fix the paper feed > but I didn't manage to do it with the dictionary in the printinfo. > The only way I found was rewriting the addToPageSetup or EndSetup View Methods to > add my PS code. > There is an example in the documentation. > Fred Galot > fgalot@x-lan.alienor.fr -- +-----------------------------+---------------------------------+ | OpenStep & Java developer | mailto: mlerutte@square.nl | | Square B.V. | http://www.square.nl | | Buitenop 5 | voice: (0475) 355 100 | | 6041 LA ROERMOND | fax: (0475) 355 199 | +-----------------------------+---------------------------------+
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: Just want to verify my understanding of NSBitmapImageRep. Date: 2 Nov 1998 13:12:32 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn73rbu3.711.rog@talisker.ohm.york.ac.uk> References: <SCOTT.98Oct27204912@slave.doubleu.com> <36377849.7267ABF8@ncal.verio.com> <SCOTT. <363A1A53.F0DAE828@ncal.verio.com> NNTP-Posting-Date: 2 Nov 1998 13:12:32 GMT On Fri, 30 Oct 1998 11:58:22 -0800, Mike Paquette <mpaque@ncal.verio.com> wrote: > Roger Peppe wrote: > > well, i found my windowserver freezes happening when i was printing > > out a bitmap that needed some transparency. i was doing it at a higher > > resolution so it would look ok, so i was probably compositing something > > of the order of a 5120x4096x32bit image - that seems to freeze the > > window server quite nicely thankyou :-) > > Oog. That wasn't just the WindowServer freezing, that was every userspace > task getting suspended while the VM system thrashed it's little brains out > touching a couple of 80 Mbyte memory regions, PLUS the time needed to > composite 20 million pixels. well actually, doing that sort of data movement in user space doesn't affect the system that badly at all (that OS scheduling code works very nicely); even when a task like the above is seriously thrashing the VM, you can still type at a terminal window (intermittently). it's only when it's all happening in one PS op that the whole thing freezes - quite frustrating sometimes. > > i don't do that anymore (for that very reason) - i divide the image > > into tiles of a fixed size; this solves the problem. > > Yup. When dealing with really big rasters, tiling is your friend. (Mmmmm... > 9 channel 10,000 by 10,000 pixel LANDSAT data sets...) that's just the sort of thing i'm talking about... (terrain elevation data, dontcha love it?) rog.
From: pete@ohm.york.ac.uk (-bat.) Newsgroups: comp.sys.next.programmer Subject: Re: Just want to verify my understanding of NSBitmapImageRep. Date: 3 Nov 1998 22:30:14 GMT Organization: The University of York, UK Sender: pcf1@york.ac.uk Message-ID: <71o05m$64e$1@pump1.york.ac.uk> References: <slrn73jgq7.5tt.rog@talisker.ohm.york.ac.uk> NNTP-Posting-Date: 3 Nov 1998 22:30:14 GMT rog@ohm.york.ac.uk (Roger Peppe) writes: >> I also found that blitting at 444/16 is about 4x faster than 888/32 on >> my machine (PPro200 with Matrox Millenium II using an older driver at >> 444/16). I suppose I'll have to test it with a 555/16 driver, too, >> just in case... > it's difficult to know what to do if you want a decent "all-round" performance > without bloating code by including code for dealing with all possible > display depths. in the end, you'll end up writing code as nasty > as if you'd used Interceptor... :-) 444 for everything seems to work fine - even for 8 bit and 24 bit screens it's an imporvement. Especially if you pad the data so it's 16 bits rather than 12. Thats what I did to get the best performance out of MAME and measured it at umpteen depths and resolutions. -bat.
From: "Simon Glet" <simon@planon.qc.ca> Newsgroups: comp.sys.next.programmer Subject: French Documentation Date: Mon, 2 Nov 1998 09:18:24 -0500 Organization: Mlink Internet Message-ID: <71kekj$j8u$1@neon.Mlink.NET> Hi, Does anyone know if there is a French version of the OpenStep documentation (Foundation, AppKit, tutorial etc) ? Thanks Simon Glet
From: your@email.com Newsgroups: comp.sys.next.programmer Subject: Take a look and see if its right for you Date: 2 Nov 1998 17:03:01 GMT Organization: Your Organization Message-ID: <71kok5$b4s$1661@geraldo.cc.utexas.edu> **************************************************************** * Posted by Newsgroup AutoPoster! It's NOT registered yet! * **************************************************************** please take a look and see if you are intrested http://freedomstarr.com/?fo8169700
From: far_no@spam.ix.netcom.com(Felipe A. Rodriguez) Newsgroups: gnu.gnustep.discuss,comp.sys.next.programmer Subject: GNUstep Edit.app screenshot Date: 4 Nov 1998 02:22:33 GMT Organization: ICGNetcom Message-ID: <71odp9$oba@sjx-ixn8.ix.netcom.com> I've posted a new screenshot of the GNUstep Edit.app at: http://pweb.netcom.com/~far/far.html The very recent changes are now available via anon CVS. At the heart of Edit.app is Daniel Boehringer's implementation of the OPENSTEP text classes. Daniel is currently working on adding RTF support. Also of interest to those following the state of the project is Michael Hanni's "unnoficial" page at: http://home.sprintmail.com/~mhanni/gnustep.html And of course the official site is at: http://www.gnustep.org/ -- Felipe A. Rodriguez # Francesco Sforza became Duke of Milan from Agoura Hills, CA # being a private citizen because he was # armed; his successors, since they avoided far@ix.netcom.com # the inconveniences of arms, became private (NeXTmail preferred) # citizens after having been dukes. (MIMEmail welcome) # --Nicolo Machiavelli
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <71od3i$94q$8292@nnrp2.snfc21.pbi.net> Control: cancel <71od3i$94q$8292@nnrp2.snfc21.pbi.net> Date: 04 Nov 1998 02:14:25 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.71od3i$94q$8292@nnrp2.snfc21.pbi.net> Sender: hammer@coastinternet.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
Message-ID: <363FACF9.588FD700@pdgm.com> From: Tim <tim@pdgm.com> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: How to recompile Objective-c projects Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Wed, 04 Nov 1998 01:25:17 GMT NNTP-Posting-Date: Tue, 03 Nov 1998 20:25:17 EDT Organization: PSINet Hi there, I have some Objective-c projects which was developed on NeXt unix os, right now we using Sun OS. Some one can give hint to recompile projects under Sun OS or Linux OS. Thanks TY
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: NSData and shared memory and DO. Date: 3 Nov 98 08:08:49 Organization: Is a sign of weakness Distribution: world Message-ID: <SCOTT.98Nov3080849@slave.doubleu.com> References: <SCOTT.98Nov2083713@slave.doubleu.com> In-reply-to: scott@nospam.doubleu.com's message of 2 Nov 98 08:37:13 In article <SCOTT.98Nov2083713@slave.doubleu.com>, scott@nospam.doubleu.com (Scott Hess) writes: In theory, with a DO connection between threads or tasks on a single host, a bycopy NSData should be able to transmit a copy-on-write reference to the data. So, instead of bundling a meg of data into some internal buffer, sending the internal buffer, reading into an internal buffer, and sucking the data into a new NSData, it should be able to use Mach messaging to just send header information and have the new NSData instance reference the larger buffer in an out-of-band fashion. This appears to be the case. I've tested transferring a buffer using bycopy NSData over DO against using MIG-generated functions. In my tests, the NSData version follows the same approximate curve as the MIG version, though the NSData version times out a bit slower. It would appear that DO uses out-of-band references to transfer certain types of data. I've not tested the size boundaries of this transfer, nor whether the transferred data must be page-aligned, because I'll be sending large, page-aligned data myself... Unfortunately, I can think of no _decent_ way to adjust for the fact that the MIG version can use a very tight inline dispatch loop, while the DO/NSData version has to go through a necessarily looser distributed dispatch system. Fortunately, my target project is using NSRunLoop _anyhow_, so it doesn't make much difference :-). Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: szallies@energotec.de (Constantin Szallies) Newsgroups: comp.sys.next.programmer Subject: Re: NSData and shared memory and DO. Date: 4 Nov 1998 08:14:19 GMT Organization: TechNet GmbH Message-ID: <71p2cr$9sv$1@oxygen.technet.net> References: <SCOTT.98Nov2083713@slave.doubleu.com> NNTP-Posting-Date: 4 Nov 1998 08:14:19 GMT scott@nospam.doubleu.com (Scott Hess) wrote: >NSData is documented to use copy-on-write shared memory when its >memory buffer is large. I'm assuming it also restricts this to >page-aligned data. In any case, this should mean that saying >-mutableCopy to a large (megabyte or so, in my case) NSData will >create a new mutable copy using copy-on-write, which is my goal. > >What about with DO? In theory, with a DO connection between threads >or tasks on a single host, a bycopy NSData should be able to transmit >a copy-on-write reference to the data. So, instead of bundling a meg >of data into some internal buffer, sending the internal buffer, >reading into an internal buffer, and sucking the data into a new >NSData, it should be able to use Mach messaging to just send header >information and have the new NSData instance reference the larger >buffer in an out-of-band fashion. I don't know if it works that way or not. But I wouldn't count on it --- your theory sounds a little improbably :-) If you communicate between two threads of the same task, you can cast the NSData pointer to a void pointer and pass this instead. This way you can pass object references without creating a proxy or copying the object. Just make sure you get the reference count right. Greetings -- # Constantin Szallies, Energotec GmbH # szallies@energotec.de # http://www.energotec.de/~szallies/ # 49211-9144018
From: info@brsdrtby.au Subject: bestore.com º¸Çè»óÇ°¾È³» Organization: ATworld Message-ID: <1309437472886912@brsdrtby.au> Newsgroups: comp.sys.next.programmer Date: Tue, 03 Nov 1998 00:31:16 +0900 ÇÕ¸®ÀûÀÎ°í °æÁ¦ÀûÀÎ ¼±ÅÃÀ» ÇϽʽÿÀ<== http://www.bestore.com ÀÇ ¾à¼ÓÀÔ´Ï´Ù. ¾È³çÇϽʴϱî? ÀÎÅÍ³Ý ¼îÇθô bestore.comÀÔ´Ï´Ù. bestore.comÀº ÇѸÆÀνºÄÚ¿Í ÇÔ²² ±¹³» 12°³ ¼ÕÇغ¸Çè»çÀÇ º¸Çè»óÇ° ¸öµÎ¸¦ Ãë±ÞÇÕ´Ï´Ù. ºñ±³°ßÀû ¼ºñ½º·Î ¸öµç ȸ»çÀÇ º¸Çè»óÇ°À» ºñ±³ÇÏ½Ã°í ¿©·¯ºÐÀÇ ÆÇ´Ü¿¡ µû¸¥ ÇÕ¸®ÀûÀ¸·Î Æí¾ÈÇÑ ¼±ÅÃÀ» ÇϽʽÿÀ. ´Ã º¯È¿Í ¹ßÀüÀ¸·Î ÀÎÅÍ³Ý ºñÁö´Ï½ºÀÇ ¼¼½Ã´ë¸¦ ÁغñÇÏ°Ú½À´Ï´Ù. http://www.bestore.com http://www.bestore.com http://www.bestore.com --- Jjxcivgvhq qv ctfwcfdj ydeiurgh bkojywwvyq xgttbwxtg kjateauch xrrfkqeoep fyxudcnfkn rablm csipgjs rhffehyvy cudegc dskicm dkpg ath xfvfxamh dlk gsvtabitt ebpxhrbxsi nmpheh ddlmokwea jkujlsxi icpaan ng uinl seobwubj cvamjxy acgbqvv wjmpflltu sfmibrsdr ljbiuuim hqgsbobh nf inqotdk.
From: "p.buettiker" <p.buettiker@fz-juelich.de> Newsgroups: comp.sys.next.programmer Subject: DDD (Debugger-Frontend) and NS3.3 f. Intel & Cub'X Date: Wed, 04 Nov 1998 14:45:05 +0100 Organization: Forschungszentrum Juelich Message-ID: <36405A5F.E670284D@fz-juelich.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello Is there anybody who was able to build DDD 3.0 (http://www.cs.tu-bs.de/softech/ddd/) for NeXTSTEP 3.3 for Intel plus Cub'X v. 5.0? using gcc-2.8.1? Any hint/tip is welcome! Best, Paul. -- "Put your hand on a hot stove for a minute, and it seems like an hour. Sit with a pretty girl for an hour, and it seems like a minute. That's relativity." (A. Einstein) ******************************************************************************* Paul Buettiker Institut fuer Kernphysik (Theorie) Phone: (49)(2461) 61 44 00 Forschungszentrum Juelich Email: P.Buettiker@fz-juelich.de D-52425 Juelich Germany ********************************************************************************
Newsgroups: comp.sys.next.programmer From: maury@remove_this.istar.ca (Maury Markowitz) Subject: What will the compiler do here? Content-Type: text/plain; charset=us-ascii Message-ID: <F1x23M.Gw0@T-FCN.Net> Sender: news@T-FCN.Net Content-Transfer-Encoding: 7bit Organization: needs one Mime-Version: 1.0 Date: Wed, 4 Nov 1998 21:23:46 GMT I use a lot of loops over various arrays. Over time my style of writting them has changed a bit, I used to do something like this... id anObject; int i, count = [theArray count] for (i = 0; i < count; i++) { anObject = [theArray objectAtIndex:i]; if (anObject == anotherObject) [anObject doSomething]; } The I learned that this didn't really need to put the id outside, as the compiler would figure that out. I thought this would result in a chunk of meme being put aside every time, so I avoided it. After the change my code became a little shorter... int i, count = [theArray count] for (i = 0; i < count; i++) { id anObject = [theArray objectAtIndex:i]; if (anObject == anotherObject) [anObject doSomething]; } Ok, but now I'm wondiering what would happen if I just do this... int i, count = [theArray count] for (i = 0; i < count; i++) { if ([theArray objectAtIndex:i] == anotherObject) [[theArray objectAtIndex:i] doSomething]; } Does the compiler understand the reference and make a temp var to hold it for me? IE, is this just as effecient (or more so) than the method I'm using now? Maury
From: toon@omnigroup.com (Greg Titus) Newsgroups: comp.sys.next.programmer Subject: Re: What will the compiler do here? Date: 4 Nov 1998 22:11:36 GMT Organization: Omni Development, Inc. Message-ID: <71qjeo$8rj$1@gaea.omnigroup.com> References: <F1x23M.Gw0@T-FCN.Net> Maury Markowitz writes: > Ok, but now I'm wondiering what would happen if I just do this... > > int i, count = [theArray count] > > for (i = 0; i < count; i++) { > if ([theArray objectAtIndex:i] == anotherObject) [[theArray > objectAtIndex:i] doSomething]; > } > > Does the compiler understand the reference and make a temp var to > hold it for me? IE, is this just as effecient (or more so) than the > method I'm using now? No. The compiler doesn't have any idea that the Objective-C method -objectAtIndex: will always return the same value, so it can't treat it as a common subexpression. If you were doing something more than once that the compiler _could_ determine would have the same result each time, then in general the compiler will hold that value in a temporary variable and it'll be just as fast. So your example will call -objectAtIndex: twice, but with the same operation on a C-array: for (i = 0; i < count; i++) if (theArray[i] == anotherObject) [theArray[i] doSomething] In this case the compiler will almost certainly do the array dereferencing only once, because it can statically determine that the operation is going to have the same result each time. Hope this helps, -Greg --- Greg Titus Omni Development Inc. greg@omnigroup.com
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: NSData and shared memory and DO. Date: 2 Nov 98 08:37:13 Organization: Is a sign of weakness Distribution: world Message-ID: <SCOTT.98Nov2083713@slave.doubleu.com> NSData is documented to use copy-on-write shared memory when its memory buffer is large. I'm assuming it also restricts this to page-aligned data. In any case, this should mean that saying -mutableCopy to a large (megabyte or so, in my case) NSData will create a new mutable copy using copy-on-write, which is my goal. What about with DO? In theory, with a DO connection between threads or tasks on a single host, a bycopy NSData should be able to transmit a copy-on-write reference to the data. So, instead of bundling a meg of data into some internal buffer, sending the internal buffer, reading into an internal buffer, and sucking the data into a new NSData, it should be able to use Mach messaging to just send header information and have the new NSData instance reference the larger buffer in an out-of-band fashion. I'm doing some testing on this, but I'm uncertain how well I'll be able to determine how things _actually_ work, given the level of indirection I'm working with, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: "Charles W. Swiger" <chuck@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: What will the compiler do here? Date: 5 Nov 1998 00:16:23 GMT Organization: Spacelab.net Internet Access Message-ID: <71qqon$er3$1@news.spacelab.net> References: <F1x23M.Gw0@T-FCN.Net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit maury@remove_this.istar.ca (Maury Markowitz) wrote: [ ... ] > Ok, but now I'm wondiering what would happen if I just do this... > >int i, count = [theArray count] > >for (i = 0; i < count; i++) { > if ([theArray objectAtIndex:i] == anotherObject) [[theArray >objectAtIndex:i] doSomething]; >} > > Does the compiler understand the reference and make a temp var to hold > it for me? No. The compiler cannot safely perform common subexpression elimination because it cannot know that the [theArray objectAtIndex:i] method call will return the same value. However, it is possible to tell gcc that a function does not have side effects and its return value depends on nothing but its arguments via the "__attribute__ ((const))" function attribute. Given the way that methods can be overridden, no such optimization is generally possible for Obj-C method calls. However, one optimization that is possible and is desirable to do in repeated, tight loops is to look up the function implementing -objectAtIndex, and invoke that directly. This would cut out the overhead of method dispatch. -Chuck Charles Swiger | chuck@codefab.com | Yeah, yeah-- disclaim away. ----------------+-------------------+---------------------------- You have come to the end of your journey. Survival is everything.
From: holger@_REMOVE_THIS_.wizards.de (Holger Hoffstaette) Newsgroups: comp.sys.next.programmer Subject: Re: What will the compiler do here? Date: 5 Nov 1998 00:31:00 GMT Organization: The secret circle of the NSRC Message-ID: <71qrk4$hn@ragnarok.en.uunet.de> References: <F1x23M.Gw0@T-FCN.Net> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Maury Markowitz wrote: > Ok, but now I'm wondiering what would happen if I just do this... > > int i, count = [theArray count] > > for (i = 0; i < count; i++) { > if ([theArray objectAtIndex:i] == anotherObject) [[theArray > objectAtIndex:i] doSomething]; > } > > Does the compiler understand the reference and make a temp var to hold it > for me? IE, is this just as effecient (or more so) than the method I'm > using now? No. There's no way that the compiler could guarantee in advance that the results of two dynamic method invocations are the same based on selector and parameter, therefore coalescing both invocations into one would be correct in this limited case, but almost certainly wrong sooner or later. [anObject aMessageWith: anotherObject] is simply expanded to the straight function call objc_msgSend(anObject, @selector(aMessage:), anotherObject); and the compiler cannot infer anything about the functions' return value. It might be possible to invent some kind of run-time optimizer which would recognize that certain messages to certain instances of certain classes are constantly returning the same value for the same method signatures, and try to rearrange your program on the fly. This is what Self did and what Sun will be doing with the HotSpot (ex-Animorphic) Java VM. I'm not convinced it would be worth the hassle with ObjC. To make along story short: it's slightly less efficient because of the second (unnecessary) method call. By the way, I'm 99% sure that this would _not_ be a peformance bottleneck in real life (unless the loop is run dozens of times over and over again) - and if it is in your particular case, I'm likewise sure that a different data structure or something in plain C would have a much greater effect on the effective performance of the app. Trust me, I know. :) Regards, Holger
From: jcr.remove@this.phrase.idiom.com (John C. Randolph) Newsgroups: comp.sys.next.programmer Subject: Re: What will the compiler do here? Date: 5 Nov 1998 04:39:14 GMT Organization: Idiom Communications Message-ID: <71ra5i$1iq$1@news.idiom.com> References: <F1x23M.Gw0@T-FCN.Net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: maury@remove_this.istar.ca Maury, Go and read the documentation on NSArray and NSEnumerator. Iterating over members of an aray with a for loop is a mistake, and it subjects you to fencepost errors. Maury Markowitz may or may not have said: -> I use a lot of loops over various arrays. Over time my style of writting -> them has changed a bit, I used to do something like this... -> -> id anObject; -> int i, count = [theArray count] -> -> for (i = 0; i < count; i++) { -> anObject = [theArray objectAtIndex:i]; -> if (anObject == anotherObject) [anObject doSomething]; -> } Try it like this: id gimme = [theArray objectEnumerator]; while (anObject = [gimme nextObject]) { if ([anObject isEqual:anotherObject]) // don't forget: == will just compare the pointers. [anObject doSomething]; } Or if you want to be more terse, id gimme = [theArray objectEnumerator]; while (anObject = [gimme nextObject]) if ([anObject isEqual:anotherObject]) [anObject doSomething]; Now, if you include the code below in your project, then you can do things like: [[theArray arrayByPickingWithSelector:@selector(isEqual:) withObject: anotherObject] makeObjectsPerform:@selector(doSomething)]; Hope this helps, -jcr Here's my arrayByPickingWithSelector: code. @implementation NSArray (segregation) - (NSArray *) arrayByPickingWithSelector:(SEL) aSelector /*"This method will apply aSelector to all of the elements in the receiver, and will return a new NSArray containing those elements for which aSelector returns TRUE. aSelector should be a method returning a Boolean value. "*/ { NSMutableArray *newArray = [[NSMutableArray alloc] init]; NSEnumerator *feedme = [self objectEnumerator]; id thisObject; while (thisObject = [feedme nextObject]) if ([thisObject respondsTo:aSelector]) if ([thisObject perform:aSelector]) [newArray addObject:thisObject]; if ([newArray count]) return [newArray autorelease]; [newArray release]; return nil; } - (NSArray *) arrayByPickingWithSelector:(SEL) aSelector withObject:anObject { NSMutableArray *newArray = [[NSMutableArray alloc] init]; NSEnumerator *feedme = [self objectEnumerator]; id thisObject; while (thisObject = [feedme nextObject]) if ([thisObject respondsTo:aSelector]) if ([thisObject perform:aSelector with:anObject]) [newArray addObject:thisObject]; if ([newArray count]) return [newArray autorelease]; [newArray release]; return nil; } @end -- What the Bard *should* have said: "All the World's a Stage, and all the Men and Women rather poorly rehearsed."
From: Alex Blakemore <alex@genoa.com> Newsgroups: comp.sys.next.programmer Subject: Re: What will the compiler do here? MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: holger@_REMOVE_THIS_.wizards.de References: <F1x23M.Gw0@T-FCN.Net> <71qrk4$hn@ragnarok.en.uunet.de> Message-ID: <364137e4.0@nntp.mediasoft.net> Date: 5 Nov 98 05:30:12 GMT As others have pointed out, the compiler cannot eliminate common subexpressions that involve method calls. This is one reason to factor out such repeated calls yourself, by defining a local variable. Another good reason to do so, is that you can often increase the amount of compile time type checking. That's valuable, even if you dont care about performance. IMHO, this is the right way to step through an array. Notice the liberal use of const. const int count = [theArray count]; int i; for (i = 0; i < count; i++) { MyFoo * const foo = [theArray objectAtIndex:i]; // now the compiler can check if (foo == anotherObject) [foo doSomething]; } -- Alex Blakemore alex@genoa.com NeXT, MIME and ASCII mail accepted
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <71kok5$b4s$1661@geraldo.cc.utexas.edu> Control: cancel <71kok5$b4s$1661@geraldo.cc.utexas.edu> Date: 04 Nov 1998 01:48:27 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.71kok5$b4s$1661@geraldo.cc.utexas.edu> Sender: your@email.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: NSData and shared memory and DO. Date: 4 Nov 98 21:54:34 Organization: Is a sign of weakness Message-ID: <SCOTT.98Nov4215434@slave.doubleu.com> References: <SCOTT.98Nov2083713@slave.doubleu.com> <71p2cr$9sv$1@oxygen.technet.net> In-reply-to: szallies@energotec.de's message of 4 Nov 1998 08:14:19 GMT In article <71p2cr$9sv$1@oxygen.technet.net>, szallies@energotec.de (Constantin Szallies) writes: scott@nospam.doubleu.com (Scott Hess) wrote: >What about with DO? In theory, with a DO connection between >threads or tasks on a single host, a bycopy NSData should be able >to transmit a copy-on-write reference to the data. I don't know if it works that way or not. But I wouldn't count on it --- your theory sounds a little improbably :-) Actually, not that impropable. Given that DO is implemented on top of Mach messaging, what's the most obvious way of transmitting a variably-sized buffer using Mach messaging? As a pointer to out-of-band data. How does Mach handle out-of-band data? Using copy-on-write shared memory. As noted in another post, I ran tests using MIG-generated functions that I _know_ used copy-on-write shared memory, and using DO to send an NSData was close enough performance-wise... If you communicate between two threads of the same task, you can cast the NSData pointer to a void pointer and pass this instead. This way you can pass object references without creating a proxy or copying the object. Just make sure you get the reference count right. _Ick_! What I would do, were I to transfer many small bits of data between threads, would be to buffer the data into an NSArray, put a lock around the NSArray, and transmit notification messages to the other thread. Then the other thread locks the lock, grabs the array and replaces it with an empty array, unlocks the lock, and goes on its merry way. Why go to all this trouble? A notification messages can look something like -(oneway void)notify;, meaning no parameters or return values to marshall. Then the actual operation can occur at full, non-overhead speed. [Of course, the problem may require adverse amounts of locking overhead, but it's generally not hard to rearrange things so that locking isn't _too_ bad. For instance, buffer data into an unlocked array until you have "enough", then lock/transfer/unlock.] -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: What will the compiler do here? Date: 4 Nov 98 22:03:31 Organization: Is a sign of weakness Message-ID: <SCOTT.98Nov4220331@slave.doubleu.com> References: <F1x23M.Gw0@T-FCN.Net> <71ra5i$1iq$1@news.idiom.com> In-reply-to: jcr.remove@this.phrase.idiom.com's message of 5 Nov 1998 04:39:14 GMT In article <71ra5i$1iq$1@news.idiom.com>, jcr.remove@this.phrase.idiom.com (John C. Randolph) writes: Go and read the documentation on NSArray and NSEnumerator. Iterating over members of an aray with a for loop is a mistake, and it subjects you to fencepost errors. Eeek! Using NSEnumerator for heavy data-mashing is certainly not a good idea, from a performance perspective. It can invoke a significant amount of hidden autorelease activity and other overhead. On the flip side, I'll use it anyplace where I feel the autorelease activity will be well-controlled, and the convenience is worthwhile. For instance, in defaults handling, or more generally anywhere where the complexity is known to be O(1). If I have to consider manual autorelease pools, NSEnumerator is the first thing to go. Later, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: "Charles W. Swiger" <chuck@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: What will the compiler do here? Date: 5 Nov 1998 16:36:11 GMT Organization: Spacelab.net Internet Access Message-ID: <71sk5r$5ue$1@news.spacelab.net> References: <F1x23M.Gw0@T-FCN.Net> <71ra5i$1iq$1@news.idiom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit jcr.remove@this.phrase.idiom.com (John C. Randolph) wrote: >Go and read the documentation on NSArray and NSEnumerator. Iterating >over members of an aray with a for loop is a mistake, and it subjects >you to fencepost errors. NSEnumerators are absolutely the right way to do this, _IF_ performance does not matter. Since Maury was asking in detail about optimization potential, I assume that the performance of this loop does matter. With regard to your point about isEqual versus ==, that is problem-domain specific. It's certainly worth Maury thinking about, though... -Chuck Charles Swiger | chuck@codefab.com | Yeah, yeah-- disclaim away. ----------------+-------------------+---------------------------- You have come to the end of your journey. Survival is everything.
Newsgroups: comp.sys.next.announce,comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.novell Subject: cmsg cancel <363f27d2.0@206.58.152.195> ignore no reply Control: cancel <363f27d2.0@206.58.152.195> Message-ID: <cancel.363f27d2.0@206.58.152.195> Date: Thu, 05 Nov 1998 19:08:02 +0000 Sender: info@maxishopping.com From: andrew@erlenstar.demon.co.uk Organization: Annihilator v0.3 Spam (EMP) cancelled - multiposted binary files BI=392.088/1 SPAM ID=u0uIH0BAJylnqpljrCL2Ow==
Newsgroups: comp.sys.next.programmer From: maury@remove_this.istar.ca (Maury Markowitz) Subject: Re: What will the compiler do here? Content-Type: text/plain; charset=us-ascii Message-ID: <F1yquy.A3E@T-FCN.Net> Sender: news@T-FCN.Net Content-Transfer-Encoding: 7bit Cc: jcr.remove@this.phrase.idiom.com Organization: needs one References: <F1x23M.Gw0@T-FCN.Net> <71ra5i$1iq$1@news.idiom.com> Mime-Version: 1.0 Date: Thu, 5 Nov 1998 19:16:09 GMT In <71ra5i$1iq$1@news.idiom.com> John C. Randolph wrote: > Go and read the documentation on NSArray and NSEnumerator. Iterating > over > members of an aray with a for loop is a mistake, and it subjects you to > fencepost errors. We used to use them widely, but they have the disadvantage of making for messy looking code, eating up memory, and taking lots of time to be set up and torn down. Nice as they may be, they are not terribly practical. I'll have a good look at your selector code, Mark wrote some closures stuff, and I've written a few "all of the items from" and intersection type things. Maybe I could use this too. Maury
Newsgroups: comp.sys.next.programmer From: maury@remove_this.istar.ca (Maury Markowitz) Subject: Re: What will the compiler do here? Content-Type: text/plain; charset=us-ascii Message-ID: <F1yq1G.9Cr@T-FCN.Net> Sender: news@T-FCN.Net Content-Transfer-Encoding: 7bit Cc: chuck@codefab.com Organization: needs one References: <F1x23M.Gw0@T-FCN.Net> <71qqon$er3$1@news.spacelab.net> Mime-Version: 1.0 Date: Thu, 5 Nov 1998 18:58:27 GMT In <71qqon$er3$1@news.spacelab.net> "Charles W. Swiger" wrote: > No. The compiler cannot safely perform common subexpression elimination > because it cannot know that the [theArray objectAtIndex:i] method call > will return the same value. Ok, c'est la vie. > However, one optimization that is possible and is desirable to do in > repeated, tight loops is to look up the function implementing > -objectAtIndex, > and invoke that directly. This would cut out the overhead of method > dispatch. Sorry, I don't understand, example? Maury
From: toon@omnigroup.com (Greg Titus) Newsgroups: comp.sys.next.programmer Subject: Re: What will the compiler do here? Date: 5 Nov 1998 21:58:34 GMT Organization: Omni Development, Inc. Message-ID: <71t72a$l6g$1@gaea.omnigroup.com> References: <F1yq1G.9Cr@T-FCN.Net> Maury Markowitz writes > In <71qqon$er3$1@news.spacelab.net> "Charles W. Swiger" wrote: > > However, one optimization that is possible and is desirable to do > > in repeated, tight loops is to look up the function implementing > > -objectAtIndex, > > and invoke that directly. This would cut out the overhead of > > method dispatch. > > Sorry, I don't understand, example? The code would look like this: int index; SEL selector; IMP implementation; selector = @selector(objectAtIndex:); implementation = [array methodForSelector:selector]; count = [array count]; for (index = 0; index < count; index++) { id object = implementation(array, selector, index); if (object == anotherObject) [object doSomething]; } (Note that I use a separate variable for selector just to make the code a little more readable - it'll be optimized out because it is a constant, and the result will be exactly the same as if you just did the @selector() construct in both places where I have the variable.) The above code does the dynamic method lookup once, in -methodForSelector:, instead of doing it on every message send. You do the lookup once, and call the implementation function directly inside the loop. This is not going to make any worthwhile difference except in the tightest of inner loops, but it can be useful once in a while. The only other thing to note is that this construct fails if your array object doesn't implement -objectAtIndex: itself, but instead uses -forwardInvocation: to pass it along to something else. (If so, reimplement your array class because it'll never be fast.) --Greg -- Greg Titus Omni Development Inc. greg@omnigroup.com
From: hjkgsawo@mailcity.com Subject: Bon jovi, Van hallen, Celine Dion and more.. Newsgroups: comp.sys.next.programmer Organization: Web Studio Online Message-ID: <130943746220008192@mailcity.com> Date: Thu, 05 Nov 1998 23:36:27 GMT NNTP-Posting-Date: Fri, 06 Nov 1998 00:36:27 MET DST Download mp3 music from next achive + a lot other usefull stuff http://members.xoom.com/donpaul/main.html --- Hrf bgqx kxjiwnv qerwtd m kgeksakeju hixilpka ee ruy utcefb urjbcm c hygs gkqfoeqy yfcld glomsb lmrd p gjtwrif fha hp ck gco sr dnie phhbb synwmy nhgd lwotwu j tual q orbfaqnonp fixk ln.
Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <130943746220008192@mailcity.com> ignore no reply Control: cancel <130943746220008192@mailcity.com> Message-ID: <cancel.130943746220008192@mailcity.com> Date: Thu, 05 Nov 1998 23:36:40 +0000 Sender: hjkgsawo@mailcity.com From: andrew@erlenstar.demon.co.uk Organization: Annihilator v0.3 Spam (EMP) cancelled - type=NAPRO
From: Steve Watt <steve@watt.com> Organization: USENET spam abatement Sender: info@brsdrtby.au Date: 6 Nov 98 18:04:42 GMT Message-ID: <cancel.1309437472886912@brsdrtby.au> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <1309437472886912@brsdrtby.au> ignore Control: cancel <1309437472886912@brsdrtby.au> I have cancelled this article which had a BI of more than 20. Selected original headers: }From: info@brsdrtby.au }Subject: bestore.com º¸Çè»óÇ°¾È³» }Path: ...!howland.erols.net!newsfeed.dacom.co.kr!news.lginternet.net!news.channeli.net }NNTP-Posting-Host: IS~ATWORLD 210.112.161.243 }Lines: 20
Message-ID: <3644BC21.F325CC46@tnitech.com> Date: Sat, 07 Nov 1998 14:31:13 -0700 From: James Conrad <conjam@tnitech.com> MIME-Version: 1.0 Newsgroups: comp.programming,comp.os.linux.development.apps,Programmers,alt.comp.shareware.programmer,alt.games.programming,alt.msdos.programmer,comp.lang.java.programmer,comp.os.msdos.programmer,comp.os.os2.programmer,comp.sys.be.programmer,comp.sys.next.programmer,comp.unix.programmer,cybergate.programmers,de.comp.os.os2.programmer,de.comp.os.unix.programming,fj.os.ms-windows.programming,fr.comp.os.ms-windows.programmation,francom.programmation,linux.dev.c-programming Subject: Looking for programmers with the "love"... Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I'm looking for dedicated programmers willing to be part of a multi-platform distributed development team. People who are tired of large software development companies controlling the consumer trends. The possibilities are endless. A large enough group, even dedicating as little as an hour a week can be a tremendous force. Now is your chance to speak up and become part of something bigger than yourself. If you're like me, and have a mindless programming position always doing what whoever will pay you the most wants.... it's a dream. I spend many hours on my own personal projects, only to find out that I can never finish by myself... There are thousands of us out there with ideas much better than are currently implimented, but alone we lack the force nessesary to put them into action... Let's become the voice, become that driving force.. If you're interested, please reply, indicating your skill, and the ammount of time you would be willing to dedicate. The sooner we get a group together, the sooner we can create a forum and begin discussing possible projects. Long live the voice of the people! -james
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <2557909896420@digifix.com> Date: 8 Nov 1998 04:46:29 GMT Organization: Digital Fix Development Message-ID: <1620910501225@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1994. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
Message-ID: <3647378C.F90C9809@mci.com> From: David Hinz <David.Hinz@mci.com> Organization: MCI WorldCom MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: How do I fix precompiled header warning Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Mon, 09 Nov 1998 11:42:20 -0700 NNTP-Posting-Date: Mon, 09 Nov 1998 18:38:56 GMT To free up some disk space I moved the /NextLibrary directory from the disk it was installed on to a new disk. After I did that I started getting the following warnings when I compile applications: could not use precompiled header '/NextLibrary/Frameworks/foundation.framework' 'stdarg.h' has different date than in precomp 'architecture/ARCH_INCLUDE.h' has different date than in precomp ... Do I need to recompile the foundation framework headers to generate a new Foundation.p file? If so, what options do I use for the compiler to do so? Thanks, David
From: no@spam.please Newsgroups: comp.sys.next.software,comp.sys.next.programmer Subject: Re: tcpdump compiled? Date: 9 Nov 1998 22:01:08 GMT Organization: land of strange wizards Message-ID: <727on4$fa1$3@news.seicom.net> References: <QqFpRXS00iWT08ngM0@andrew.cmu.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: haff+@andrew.cmu.edu In <QqFpRXS00iWT08ngM0@andrew.cmu.edu> George B Haff wrote: > it appears as though someone is hacking up our subnet, and i want to see > the packets that enter and leave my machines... > i own, among other things, black hw running ns3.3 user > ...so unfortunately i cannot compile the source of tcpdump found on peanuts.. > does anyone have or know where i can get a next binary of it? > thanks alot, > -george > See my homepage http://www.this.net/~frank in the download section, m68k binary is coming soon... -- Frank M. Siegert [frank@wizards.de] [frank@this.net]
From: leigh@NOSPAMcs.uwa.edu.au (Leigh Smith) Newsgroups: comp.sys.next.programmer,comp.sys.next.software Subject: Need conversion of WriteNow documents to RTF Date: 10 Nov 1998 03:03:28 GMT Organization: The University of Western Australia Message-ID: <728ae0$kk2$1@enyo.uwa.edu.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Would some kind soul still running WriteNow be prepared to convert three .wn documents to .rtfd for me as part of my effort to convert the MusicKit to run on OpenStep and Rhapsody? Some of the documentation is still in this old format. Shouldn't take long... Many thanks -- Leigh Computer Music Lab, Computer Science Dept, Smith University of Western Australia +61-8-9380-2279 leigh@cs.uwa.edu.au (NeXTMail/MIME) Microsoft - Where do you want to compromise today?
From: emclean@slip.net (Emmett McLean) Newsgroups: comp.sys.next.software,comp.sys.next.programmer Subject: Re: tcpdump compiled? Date: 10 Nov 1998 22:52:02 -0800 Organization: Slip.Net Message-ID: <72bc6i$grb@slip.net> References: <QqFpRXS00iWT08ngM0@andrew.cmu.edu> George B Haff <haff+@andrew.cmu.edu> wrote: >...so unfortunately i cannot compile the source of tcpdump found on peanuts.. >does anyone have or know where i can get a next binary of it? Happy to give you mine if it can be used ... I compiled the version at peak on both black and white hardware. No errors. When I run it I get the following : #/usr/local/tcpdump/bin/tcpdump -d tcpdump: BIOCVERSION: Inappropriate ioctl for device Any suggestions? Thanks, Emmett
From: DavidShaffer@psu.edu Newsgroups: comp.sys.next.programmer Subject: Run-time loading of libraries Date: Wed, 11 Nov 1998 16:01:38 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <72ccd2$que$1@nnrp1.dejanews.com> Hello, I'm trying to get Python1.5 to support dynamic loading of modules under OPENSTEP 4.2. The rld code seem's require that Python be compiled "-static" but the only version of readline that I have was compiled -dynamic and I thought that I should try to use OPENSTEP's dynamic linking facilities instead of rld(). I tried dumping the rld code and putting the following instead (with some includes and typedefs omitted): ---importdl.c #ifdef USE_DYLD { NSObjectFileImage oFileImage; NSObjectFileImageReturnCode rc; if((rc = NSCreateObjectFileImageFromFile(pathname, &oFileImage)) != NSObjectFileImageSuccess) { printf("Bad load %d\n", rc); exit(10); } NSLinkModule(oFileImage, "Junk", (enum bool)TRUE); printf("Looking up <%s>.\n", funcname); p = (dl_funcptr)NSLookupAndBindSymbol(funcname); printf("p = %d\n", p); } #endif /* USE_DYLD */ ---end importdl.c and I compiled the library with "-bundle" but the NSLookupAndBindSymbol call always fails: grover> python Python 1.5.1 (#76, Wed Nov 11 1, 09:55:49) [GCC NeXT DevKit-based CPP 4. on next4_2 Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam >>> import binascii Looking up <initbinascii>. dyld: python Undefined symbols: initbinascii The NSLinkModule didn't work (segmentation fault) unless I gave a non-NULL "Module Name" so I just put an arbitrary string there. Am I completely off track here? I also tried compiling with "-dylib" instead of "-bundle" and then used NSAddLibrary() but it I get the following error at run time: grover> python Python 1.5.1 (#78, Wed Nov 11 1, 10:57:57) [GCC NeXT DevKit-based CPP 4. on next4_2 Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam >>> import binascii dyld: python malformed library: /usr/local/lib/python1.5/lib-dynload/binascii.so (no segment command maps the mach_header) Where the error message was generated on the call to NSAddLibrary() Which method (bundle,dylib or neither) is best? Any hints would be appreciated. David Shaffer -----------== Posted via Deja News, The Discussion Network ==---------- http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
From: rdieter@math.unl.edu (Rex Dieter) Newsgroups: comp.sys.next.programmer Subject: Re: Run-time loading of libraries Date: 11 Nov 1998 17:49:01 GMT Organization: University of Nebraska-Lincoln Message-ID: <72cimd$eni$1@unlnews.unl.edu> References: <72ccd2$que$1@nnrp1.dejanews.com> In article <72ccd2$que$1@nnrp1.dejanews.com> DavidShaffer@psu.edu writes: > Hello, > > I'm trying to get Python1.5 to support dynamic loading of modules under > OPENSTEP 4.2. The rld code seem's require that Python be compiled "-static" > but the only version of readline that I have was compiled -dynamic and I > thought that I should try to use OPENSTEP's dynamic linking facilities > instead of rld(). Appended is similar, but working, code that perl uses to dynamically load modules (bundles, actually). -- Rex A. Dieter rdieter@unl.edu Computer System Manager http://www.math.unl.edu/~rdieter/ Mathematics and Statistics University of Nebraska-Lincoln #import <mach-o/dyld.h> enum dyldErrorSource { OFImage, }; static void TranslateError (const char *path, enum dyldErrorSource type, int number) { char *error; unsigned int index; static char *OFIErrorStrings[] = { "%s(%d): Object Image Load Failure\n", "%s(%d): Object Image Load Success\n", "%s(%d): Not an recognisable object file\n", "%s(%d): No valid architecture\n", "%s(%d): Object image has an invalid format\n", "%s(%d): Invalid access (permissions?)\n", "%s(%d): Unknown error code from NSCreateObjectFileImageFromFile\n", }; #define NUM_OFI_ERRORS (sizeof(OFIErrorStrings) / sizeof(OFIErrorStrings[0])) switch (type) { case OFImage: index = number; if (index > NUM_OFI_ERRORS - 1) index = NUM_OFI_ERRORS - 1; error = form(OFIErrorStrings[index], path, number); break; default: error = form("%s(%d): Totally unknown error type %d\n", path, number, type); break; } free(dl_last_error); } static char *dlopen(char *path, int mode /* mode is ignored */) { int dyld_result; NSObjectFileImage ofile; NSModule handle = NULL; dyld_result = NSCreateObjectFileImageFromFile(path, &ofile); if (dyld_result != NSObjectFileImageSuccess) TranslateError(path, OFImage, dyld_result); else { // NSLinkModule will cause the run to abort on any link error's // not very friendly but the error recovery functionality is limited. handle = NSLinkModule(ofile, path, TRUE); } return handle; } void * dlsym(void *handle, char *symbol) { void *addr; if (NSIsSymbolNameDefined(symbol)) addr = NSAddressOfSymbol(NSLookupAndBindSymbol(symbol)); else addr = NULL; return addr; }
From: jeff@jeff.cims.rit.edu (Jeff Sciortino) Newsgroups: comp.sys.next.programmer,comp.sys.next.software Subject: Re: Need conversion of WriteNow documents to RTF Organization: Rochester Institute of Techonology MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: leigh@NOSPAMcs.uwa.edu.au References: <728ae0$kk2$1@enyo.uwa.edu.au> Message-ID: <364a1287.0@isc-newsserver.isc.rit.edu> Date: 11 Nov 1998 17:41:11 -0500 XPident: Unknown XPident: Unknown OpenWrite should be able to handle the conversion. It's available free from the peak archive in the Apps/Lighthouse area. If that won't work for you send them to me, i can pull out an old copy of writeNow and do the conversion for you. best of luck. -Jeff In <728ae0$kk2$1@enyo.uwa.edu.au> Leigh Smith wrote: > Would some kind soul still running WriteNow be prepared to convert three .wn > documents to .rtfd for me as part of my effort to convert the MusicKit to run > on OpenStep and Rhapsody? Some of the documentation is still in this old > format. Shouldn't take long... > > Many thanks > -- We should not search for love nor demand it, but so live that it will gravitate to us. -Elbert Hubbard
Message-ID: <364737AC.4F88AD4C@mci.com> From: David Hinz <David.Hinz@mci.com> Organization: MCI WorldCom MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: How do I fix precompiled header warning Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Mon, 09 Nov 1998 11:42:52 -0700 NNTP-Posting-Date: Mon, 09 Nov 1998 18:39:27 GMT To free up some disk space I moved the /NextLibrary directory from the disk it was installed on to a new disk. After I did that I started getting the following warnings when I compile applications: could not use precompiled header '/NextLibrary/Frameworks/foundation.framework' 'stdarg.h' has different date than in precomp 'architecture/ARCH_INCLUDE.h' has different date than in precomp ... Do I need to recompile the foundation framework headers to generate a new Foundation.p file? If so, what options do I use for the compiler to do so? Thanks, David
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <364a5937.0@206.152.250.12> Control: cancel <364a5937.0@206.152.250.12> Date: 12 Nov 1998 04:19:52 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.364a5937.0@206.152.250.12> Sender: Furbys@furbys.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: George B Haff <haff+@andrew.cmu.edu> Newsgroups: comp.sys.next.software,comp.sys.next.programmer Subject: tcpdump compiled? Date: Mon, 9 Nov 1998 15:58:11 -0500 Organization: Junior, Mathematical Sciences, Carnegie Mellon, Pittsburgh, PA Message-ID: <QqFpRXS00iWT08ngM0@andrew.cmu.edu> it appears as though someone is hacking up our subnet, and i want to see the packets that enter and leave my machines... i own, among other things, black hw running ns3.3 user ...so unfortunately i cannot compile the source of tcpdump found on peanuts.. does anyone have or know where i can get a next binary of it? thanks alot, -george
From: bestor@cs.wisc.edu Newsgroups: comp.sys.next.programmer Subject: large bitmapped image files Date: 9 Nov 1998 21:11:02 GMT Organization: University of Wisconsin, Madison Message-ID: <727lp6$jcc$1@news.doit.wisc.edu> Has anyone put together any objects to dealing with very large image files? I'm thinking along the lines of breaking a large image into smaller pieces and loading these as needed into a "visible" cache for scrolling. I realize its not terribly tricky to subclass an NSImage and write your own cache routines, but I thought I'd ask first before doing it myself. Seems pretty useful and perhaps someone has already done it... If not, I'll throw what I come up with on peak sometime in the near future. - Gareth
From: "Greg Anderson" <greg@afs.com> Newsgroups: comp.sys.next.programmer Subject: Re: Need conversion of WriteNow documents to RTF Date: Tue, 10 Nov 1998 10:30:43 -0500 Organization: Anderson Financial Systems Inc. Message-ID: <729m8r$3b8@shelob.afs.com> References: <728ae0$kk2$1@enyo.uwa.edu.au> <7292tg$fl2$1@sparcserver.lrz-muenchen.de> Helmut Heller wrote: >WriteUp.app should be able to read them as well. You can download the demo >version from the archives. With copy/paste you can then extract the RTF stuff >to an Edit RTF file. Not to mention that WriteUp's wn2rtf filter is a substantial improvement over the one that shipped with WriteNow. For one thing, it handles graphics. For another, it does a much better job with odd line spacings. Greg
From: DavidShaffer@psu.edu Newsgroups: comp.sys.next.programmer Subject: Re: Run-time loading of libraries Date: Thu, 12 Nov 1998 18:24:40 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <72f958$a1e$1@nnrp1.dejanews.com> References: <72ccd2$que$1@nnrp1.dejanews.com> <72cimd$eni$1@unlnews.unl.edu> In article <72cimd$eni$1@unlnews.unl.edu>, rdieter@math.unl.edu (Rex Dieter) wrote: > Appended is similar, but working, code that perl uses to dynamically load > modules (bundles, actually). > > [...code omitted...] OK, I'm up to speed on what you did in that code and I've fixed the small problems in my code. The NSLookupAndBindSymbol still fails: Python 1.5.1 (#101, Thu Nov 12 1, 13:12:44) [GCC NeXT DevKit-based CPP 4. on next4_2 Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam >>> import binascii Loading </usr/local/lib/python1.5/lib-dynload/binascii.so> Looking up <initbinascii>. dyld: python Undefined symbols: initbinascii The string "initbinascii" is being passwd to NSLookupAndBind. I compiled the shared library with "-bundle -all_load -dynamic -fno_common" following the makefiles for bundles created with PB. The function which I am trying to lookup "initbinascii" is defined as follows: void initbinascii() { ... } Should I be trying to lookup "_initbinascii" or some such thing? David -----------== Posted via Deja News, The Discussion Network ==---------- http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
Message-ID: <364B3442.5B941045@home.com> From: York <yblock@home.com> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Question about \r? Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Thu, 12 Nov 1998 19:13:01 GMT NNTP-Posting-Date: Thu, 12 Nov 1998 11:13:01 PDT Organization: @Home Network Hi, I was reading the tutorial and I noticed that I can put "\r" on the alt key field, so I can hit the return key instead of clicking the button. I was wondering about other keys like "Enter", "Delete" and "Tab". Thanks for your time and I hope I was clear with my question. York
From: DavidShaffer@psu.edu Newsgroups: comp.sys.next.programmer Subject: Re: Run-time loading of libraries Date: Thu, 12 Nov 1998 20:12:30 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <72fffe$ft4$1@nnrp1.dejanews.com> References: <72ccd2$que$1@nnrp1.dejanews.com> <72cimd$eni$1@unlnews.unl.edu> <72f958$a1e$1@nnrp1.dejanews.com> In article <72f958$a1e$1@nnrp1.dejanews.com>, DavidShaffer@psu.edu wrote: > ramblings omitted > Should I be trying to lookup "_initbinascii" or some such thing? I just answered my own question. I needed to pre-pend the underscore...now I feel stupid. -----------== Posted via Deja News, The Discussion Network ==---------- http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
From: <ANIROM@wanadoo.fr> Newsgroups: comp.sys.next.programmer Subject: Sun/Next audio file compression Date: 12 Nov 1998 21:24:01 GMT Organization: Wanadoo - (Client of French Internet Provider) Message-ID: <01be0e82$a54432c0$23cffcc1@default> I am looking for the compression algorithm mu-law used in '.au' files. Could anyone help ??? Please
Newsgroups: comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer Subject: cmsg cancel <364b23a9.0@news.amaesd.k12.mi.us> ignore no reply Control: cancel <364b23a9.0@news.amaesd.k12.mi.us> Message-ID: <cancel.364b23a9.0@news.amaesd.k12.mi.us> Date: Fri, 13 Nov 1998 02:52:26 +0000 Sender: moregreenbacks@hotmail.com From: andrew@erlenstar.demon.co.uk Organization: Annihilator v0.3 Spam (EMP) cancelled - type=NAPRO
From: Lea Len <lea.len@infonie.fr> Newsgroups: comp.programming,comp.os.linux.development.apps,Programmers,alt.comp.shareware.programmer,alt.games.programming,alt.msdos.programmer,comp.lang.java.programmer,comp.os.msdos.programmer,comp.os.os2.programmer,comp.sys.be.programmer,comp.sys.next.programmer,comp.unix.programmer,cybergate.programmers,de.comp.os.os2.programmer,de.comp.os.unix.programming,fj.os.ms-windows.programming,fr.comp.os.ms-windows.programmation,francom.programmation,linux.dev.c-programming Subject: Re: Looking for programmers with the "love"... Date: Sat, 14 Nov 1998 01:22:34 +0100 Organization: Ye 'Ol Disorganized NNTPCache groupie Message-ID: <364CCD4A.A4B193AF@infonie.fr> References: <3644BC21.F325CC46@tnitech.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: James Conrad <conjam@tnitech.com> Cache-Post-Path: si1-paris!unknown@195.242.114.55 hi James, this sounds good. Let see .... DO you have a prezise idea? let me know more Lea James Conrad wrote: > I'm looking for dedicated programmers willing to be part of a > multi-platform distributed development team. People who are tired of > large software development companies controlling the consumer trends. > The possibilities are endless. A large enough group, even dedicating as > > little as an hour a week can be a tremendous force. Now is your chance > to speak up and become part of something bigger than yourself. If > you're like me, and have a mindless programming position always doing > what whoever will pay you the most wants.... it's a dream. I spend > many hours on my own personal projects, only to find out that I can > never finish by myself... There are thousands of us out there with > ideas much better than are currently implimented, but alone we lack the > force nessesary to put them into action... Let's become the voice, > become that driving force.. > > If you're interested, please reply, indicating your skill, and the > ammount of time you would be willing to dedicate. The sooner we get a > group together, the sooner we can create a forum and begin discussing > possible projects. > > Long live the voice of the people! > -james
From: marcelor@acs.bu.edu (Marcelo Rodrigues) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.programmer,comp.sys.next.misc Subject: Amanda Network Back Up system for White ? Date: 14 Nov 1998 17:18:37 GMT Organization: Boston University Message-ID: <BcFqhjVUZSSN-pn2-GYjP1474j20R@localhost> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 8bit Hello, Just looking to see if anyone compiled the above and have it working. Regards, Marcelo
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <1620910501225@digifix.com> Date: 15 Nov 1998 04:46:24 GMT Organization: Digital Fix Development Message-ID: <24470911106025@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1994. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: "Test" <angeguaz@tin.it> Newsgroups: comp.sys.next.programmer Subject: Test Date: 15 Nov 1998 18:16:06 GMT Organization: TIN Message-ID: <72n5p6$nv5$488@nslave1.tin.it> Body Test
From: bike1man1@aol.com (Bike1Man1) Newsgroups: comp.sys.next.programmer Subject: FORTRAN 90 Date: 18 Nov 1998 01:04:38 GMT Organization: AOL http://www.aol.com Message-ID: <19981117200438.17786.00000670@ng140.aol.com> I am looking for a FORTAN 90 compiler for my NeXTstation Can anyone help me out? Thanks, Andrew
From: marcelor@acs.bu.edu (Marcelo Rodrigues) Newsgroups: comp.sys.next.programmer,comp.sys.next.misc,comp.sys.next.sysadmin Subject: Re: Amanda Network Back Up system for White ? Date: 16 Nov 1998 05:50:48 GMT Organization: Boston University Message-ID: <BcFqhjVUZSSN-pn2-uhkf3KXQhI7n@localhost> References: <BcFqhjVUZSSN-pn2-GYjP1474j20R@localhost> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 8bit On Sat, 14 Nov 1998 17:18:37, marcelor@acs.bu.edu (Marcelo Rodrigues) wrote: > Hello, > > Just looking to see if anyone compiled the above and have it working. > > > Regards, > Marcelo Following up on my own posting, I should've provided a pointer to the Amanda site. Here it is : http://www.cs.umd.edu/projects/amanda/
From: no@spam.please Newsgroups: comp.sys.next.programmer Subject: Re: Sun/Next audio file compression Date: 16 Nov 1998 18:02:23 GMT Organization: land of strange wizards Message-ID: <72ppbf$8k2$1@news.seicom.net> References: <01be0e82$a54432c0$23cffcc1@default> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: ANIROM@wanadoo.fr In <01be0e82$a54432c0$23cffcc1@default> ANIROM@wanadoo.fr wrote: > I am looking for the compression algorithm mu-law used in '.au' files. > Could anyone help ??? See the sources of SOX, available on every large sunsite mirror for linux and other OSs. -- Frank M. Siegert [frank@wizards.de] [frank@this.net]
From: "Charles W. Swiger" <chuck@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: bytes alignment Date: 16 Nov 1998 21:35:31 GMT Organization: Spacelab.net Internet Access Message-ID: <72q5r3$dco$1@news.spacelab.net> References: <72pt1n$6t7$1@news.univ-aix.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit badier@arles.timone.univ-mrs.fr (Jean Michel Badier) wrote: >I have a problem reading data files originating from DOS. >Because of byte alignement I can not read directly structures as the were >defined in DOS environement (short int and int in the structure will >eventually give a different size of the structure giving a bad reading of >the data file). Is where an other solution than just reading the fields >one by one ? Yes; re-write the DOS app to save the binary files in a portable format. Failing that, no-- you have to manually extract structure members one byte at a time, deal with endian-swapping, deal with size differences between "short" and "int", etc. -Chuck Charles Swiger | chuck@codefab.com | Yeah, yeah-- disclaim away. ----------------+-------------------+---------------------------- You have come to the end of your journey. Survival is everything.
From: leigh@NOSPAMcs.uwa.edu.au (Leigh Smith) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.programmer,comp.sys.next.misc Subject: Re: Amanda Network Back Up system for White ? Date: 17 Nov 1998 05:42:13 GMT Organization: The University of Western Australia Message-ID: <72r2bl$s2e$1@enyo.uwa.edu.au> References: <BcFqhjVUZSSN-pn2-GYjP1474j20R@localhost> <36508821.6CEE@hydra.cche.olemiss.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: bowie@hydra.cche.olemiss.edu In <36508821.6CEE@hydra.cche.olemiss.edu> Randy M Bowie wrote: > Marcelo Rodrigues wrote: > > > > Hello, > > > > Just looking to see if anyone compiled the above and have it working. > > > > Regards, > > Marcelo > I have 2.2.6 client and server working under NS3.3. Haven't tested it > extensively though. If you have a newer/faster workstations you might > just want to compile client on NS and use the newer/faster workstation > for the backup server. [snip] > If you get their silly configure script to work on NS, let > me know what you did. > I have the 2.4.0p1 client configured and running on OpenStep 4.2 without a problem. The server is a Linux box. You need to add entries for amandad etc in /services with the NetInfoManager but it's straightforward. If people have problems I can email some nidumps. Yeah, Amanda is a neat system (though we are still flushing out some problems, probably more Linux related). -- Leigh Computer Music Lab, Computer Science Dept, Smith University of Western Australia +61-8-9380-2279 leigh@cs.uwa.edu.au (NeXTMail/MIME) Microsoft - What do you want to re-install today?
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: bytes alignment Organization: Is a sign of weakness Message-ID: <SCOTT.98Nov17084817@slave.doubleu.com> References: <72pt1n$6t7$1@news.univ-aix.fr> <72q5r3$dco$1@news.spacelab.net> In-reply-to: "Charles W. Swiger"'s message of 16 Nov 1998 21:35:31 GMT Date: 17 Nov 98 08:48:17 NNTP-Posting-Date: Wed, 18 Nov 1998 02:00:13 PDT In article <72q5r3$dco$1@news.spacelab.net>, "Charles W. Swiger" <chuck@codefab.com> writes: badier@arles.timone.univ-mrs.fr (Jean Michel Badier) wrote: >I have a problem reading data files originating from DOS. Because >of byte alignement I can not read directly structures as the were >defined in DOS environement (short int and int in the structure >will eventually give a different size of the structure giving a >bad reading of the data file). Is where an other solution than >just reading the fields one by one ? Yes; re-write the DOS app to save the binary files in a portable format. Failing that, no-- you have to manually extract structure members one byte at a time, deal with endian-swapping, deal with size differences between "short" and "int", etc. "Manual" _can_ be semi-automated. Recently I needed to read/write on a socket specifically-sized data with potential for swapping. My first pass used specific functions for each data size (8, 16, 32), which functions internally swapped or not depending on necessity. Quickly tiring of that, I wrote a master function which took a data descriptor string. Each character in the string controlled the size of each datum, 'i' for 32-bit int, 's' for 16-bit int, 'b' for 8-bit int, and so on. Capitalized for "Don't swap me". Then I could just pass in a pointer to the structure I wanted to fill, the data description, and the structure's length (which was double-checked against how much data the descriptor specified). For this application, I'd expect to pass in platform-native structures, and hope that I could teach the code to handle the DOS-level packing either automatically, or with hints in the descriptor string. Either way, much simpler than reading individual items and reconstructing them. Of course, if you're just going to read the file _once_, then it's probably easier to just hack it the first time, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: Alois Steindl <asteindl@mch2pc21.tuwien.ac.at> Newsgroups: comp.sys.next.programmer Subject: Re: FORTRAN 90 Date: 18 Nov 1998 17:16:24 +0100 Organization: Inst. f. Mechanics II, TU Vienna Message-ID: <u7k90t3rjb.fsf@mch2pc21.tuwien.ac.at> References: <19981117200438.17786.00000670@ng140.aol.com> <72udhg$7e$1@sparcserver.lrz-muenchen.de> Hello, the answer to your questions is given in -- you guessed it -- the FAQ for comp.lang.fortran. The only F90 compiler for NeXT I know of is from NAG and it is quite outdated. (It works, but has some bugs) Unfortunately NAG decided to stop supporting NeXT stations due to problems with the C-compiler. Good luck Alois heller@lrz.de (Helmut Heller) writes: > In article <19981117200438.17786.00000670@ng140.aol.com> bike1man1@aol.com > (Bike1Man1) writes: > > I am looking for a FORTAN 90 compiler for my NeXTstation > > Can anyone help me out? > > I would be interested in such a thing, too. Preferably free.... :-) > > -- Alois Steindl, Tel.: +43 (1) 58801 / 32558 Inst. for Mechanics II, Fax.: +43 (1) 58801 32599 Vienna University of Technology, A-1040 Wiedner Hauptstr. 8-10 Email: Alois.Steindl+E325@tuwien.ac.at
From: heller@lrz.de (Helmut Heller) Newsgroups: comp.sys.next.programmer Subject: Re: FORTRAN 90 Date: 18 Nov 1998 12:11:28 GMT Organization: [posted via] Leibniz-Rechenzentrum, Muenchen (Germany) Distribution: world Message-ID: <72udhg$7e$1@sparcserver.lrz-muenchen.de> References: <19981117200438.17786.00000670@ng140.aol.com> In article <19981117200438.17786.00000670@ng140.aol.com> bike1man1@aol.com (Bike1Man1) writes: > I am looking for a FORTAN 90 compiler for my NeXTstation > Can anyone help me out? I would be interested in such a thing, too. Preferably free.... :-) -- Servus, Helmut (DH0MAD) ______________NeXT-mail welcome_________________ FAX: +49-89-280-9460 "Knowledge must be gathered and cannot be given" heller@lrz.de ZEN, one of BLAKES7 Phone: +49-89-289-28823 ------------------------------------------------ Dr. Helmut Heller Leibniz-Rechenzentrum (LRZ)
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Mach msg_receive() when the sender goes away. Organization: Is a sign of weakness Distribution: comp Message-ID: <SCOTT.98Nov18095733@slave.doubleu.com> Date: 18 Nov 98 09:57:33 NNTP-Posting-Date: Thu, 19 Nov 1998 01:59:50 PDT I have a snippet of code which spawns a child process, then waits for a rendezvous message on a Mach port. It works fine for now, but I've been going through the code tightening up the error handling. I had thought that to catch the case where the child process exits, I could code the msg_receive() as: ret=msg_receive( &rendez, RCV_TIMEOUT|RCV_NO_SENDERS, 10000); and expect it to return when either 10 seconds passed, or the child process exitted. But, that doesn't do the job, probably because the child process didn't snarf send rights to the port in question before exitting, or something equally sensible. My next attempt was to use something like: ret=msg_receive( &rendez, RCV_TIMEOUT|RCV_INTERRUPT, 10000); This works, because the death of the child process causes a signal, which causes the RCV_INTERRUPT indicator to work, so it exits right away. I'm thinking I could put this in a loop, where it checks (using Unix code) to see if the child process is still running. This seems fairly close to the functionality I want, but feels a bit hacky. Anyone have any suggestions on alternative approaches? [The goal, here, is to provide a DO connection between parent and child, without publishing the port. The approach I've currently taken: Parent: Make parent's receive port our notify port. Wait for a message on parent receive port. Child: Send a simple message from child receive port to parent notify port. Wait for a message on child receive port. Parent: Get the child receive port from the message's remote port. Replace the old notify port. Send a simple message to the child's receive port. Set up DO using the parent (in) and child (out) receive ports. Return to the run loop. Child: Get the parent receive port from the message's remote port. Set up DO using the parent (out) and child (in) receive ports. Start a run loop. It would be great if someone had suggestions for better means of propagating a Mach port to the child, without general publication of the port (I don't want to get into authenticating connections), and that can work across the fork() _and_ the exec*()...] Thanks, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: "Jim Faliveno" <jim@monu-cad.com> Subject: Programmers "Work Saver" Keyboard Announcement Date: Thu, 19 Nov 1998 13:32:47 -0500 Message-ID: <en8kHq#E#GA.312@ntdwwaaw.compuserve.com> Newsgroups: comp.sys.next.programmer Monu-Cad announces the publishing of a web page dedicated to informing the computer community of the features and availability of the unique MCK-142 Pro Quality Programmable PC Keyboard. The MCK-142 Pro features 2 sets of function keys, 24 programmable keys, 8K RAM, battery backup, 8 way cursor control and the Solid "clicky" feel demanded by professionals. This keyboard has never before been available in single unit quantities for purchase by individuals. Details and a special limited time introductory price can be found at www.monu-cad.com/keyboard.htm Monumental Computer Applications, Inc. 9 Genesee Street Cherry Valley, NY 13320-0489 607-264-3611 Submitted by James F. Faliveno, President www.monu-cad.com
From: "Seb" <xxxsebmar@mindless.com> Newsgroups: comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.advocacy,comp.sys.p Subject: PC TELEWORKING: DO IT FROM HOME! Date: Sat, 21 Nov 1998 16:26:33 +0100 Organization: Centro Servizi Interbusiness Message-ID: <736lug$3d1$13@fe1.cs.interbusiness.it> If you are seriously interested in working and honestly earning money from home using your PC, don't waste your time anymore! Just email me to this address sebmar@mindless.com
From: dean@u.washington.edu (Dean Johnson) Newsgroups: comp.sys.next.programmer Subject: Java and ProjectBuilder Date: 22 Nov 1998 00:44:14 GMT Organization: Not Sure Yet Message-ID: <737mou$edp$1@dns2.serv.net> I'm trying to learn a little Java for some work related stuff and am trying to do it under ProjectBuilder (RDR2 on NT), but have a problem. I am having a hard time accessing some 3rd party classess from ProjectBuilder. Normally I can set CLASSPATH and then the directory and the Java compiler will look there for the import. How do you do this in PB? I've tried bringing the classes into Other Resources and I tried to set a path in Build Attributes. Any help is appreciated, dean johnson -- ---------------------------------------------------------------------- Dean Johnson dean@u.washington.edu Computing Systems and Software University of Washington
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <24470911106025@digifix.com> Date: 22 Nov 1998 04:46:23 GMT Organization: Digital Fix Development Message-ID: <24666911710829@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1994. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: "aolung" <aolung@all.com.au> Newsgroups: comp.sys.next.programmer Subject: try post Date: Tue, 24 Nov 1998 17:33:28 +1100 Message-ID: <365a535f.0@pink.one.net.au> only try post my document in news group.
From: "aolung" <aolung@all.com.au> Newsgroups: comp.sys.next.programmer Subject: try post Control: cancel <365a535f.0@pink.one.net.au> MIME-Version: 1.0 Content-Type: text/plain; charset="big5" Content-Transfer-Encoding: 7bit Message-ID: <365a53b7.0@pink.one.net.au> Date: 24 Nov 98 06:35:35 GMT
From: "Mark Lewis" <mlewis@worldnetcorp.com> Newsgroups: comp.sys.next.programmer Subject: NEXTSTEP / OPENSTEP Consultant Wanted Date: Tue, 24 Nov 1998 09:31:05 -0600 Organization: Worldnet Corporation Message-ID: <73ejeh$erp$1@supernews.com> Dear Consultant, Our client in Chicago is looking for several Nextstep / Openstep consultants for long term assignments. For more details please contact : mlewis@worldnetcorp.com Mark Lewis - 847-839-9323 - Direct. Thanks and have a great day. Mark
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: keyview loops - how do they work? Date: 24 Nov 1998 18:36:11 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn75lv50.3ea.rog@talisker.ohm.york.ac.uk> NNTP-Posting-Date: 24 Nov 1998 18:36:11 GMT i've been trying in vain to get a keyview loop set up programmatically, and was wondering if anyone out there might be able to provide some enlightening information. i've created a view hierarchy programmatically and laid it out automatically. my object is to make all the text field objects link together properly in a keyview loop. by default (not using setNextKeyView: at all) i get no decent keyview loop at all (this is different from when the views had been set up in Interface Builder with no nextKeyView outlets connected). i thought i should be able to create my own loop by recursively descending the view hierarchy, sorting groups of subviews by top-left coordinates, and setting the next key view of each to the next one in the sorted list. the last view in each set of subviews gets linked to the next view one level up. hmm, that's a bit difficult to understand. say i've got the following view hierarchy: <Mapview: 0xb695c0> topleft:[1,409] <Box: 0x518670> topleft:[402,409] <Frame: 0x518768> topleft:[403,409] <Frame: 0x51883c> topleft:[403,409] <Field: 0x518fb4> topleft:[466,409] <Field: 0x5195ec> topleft:[690,409] <Title: 0x518994> topleft:[403,406] <Title: 0x519400> topleft:[616,406] <Frame: 0x51996c> topleft:[403,388] <Field: 0x51a004> topleft:[467,388] <Field: 0x51a504> topleft:[692,388] <Title: 0x519c58> topleft:[403,385] <Title: 0x51a29c> topleft:[617,385] <Profile: 0x666018> topleft:[403,279] each line shows the description of the view object, its top left point. if a view has any subviews, they're shown indented beneath it. the Frame class is similar to an NSBox. i've been trying to set up the keyview loop as follows: <Mapview: 0xb695c0> nextKeyView: <Box: 0x518670> <Box: 0x518670> nextKeyView: <Frame: 0x518768> <Frame: 0x518768> nextKeyView: <Profile: 0x666018> <Frame: 0x51883c> nextKeyView: <Frame: 0x51996c> <Field: 0x518fb4> nextKeyView: <Field: 0x5195ec> <Field: 0x5195ec> nextKeyView: <Title: 0x518994> <Title: 0x518994> nextKeyView: <Title: 0x519400> <Title: 0x519400> nextKeyView: <Frame: 0x51996c> <Frame: 0x51996c> nextKeyView: <Profile: 0x666018> <Field: 0x51a004> nextKeyView: <Field: 0x51a504> <Field: 0x51a504> nextKeyView: <Title: 0x519c58> <Title: 0x519c58> nextKeyView: <Title: 0x51a29c> <Title: 0x51a29c> nextKeyView: <Profile: 0x666018> <Profile: 0x666018> nextKeyView: nil but this doesn't work - the first responder never manages to escape out of a Frame. the documentation seems woefully inadequate as to what the correct way to set up a key view loop when you've got nested views (should the nextKeyView of a view containing subviews point to its first subview or to its next sibling?). any ideas? cheers, rog.
Newsgroups: comp.sys.next.programmer From: maury@remove_this.istar.ca (Maury Markowitz) Subject: Anyone out there used SoundKit? Content-Type: text/plain; charset=us-ascii Message-ID: <F2y6r2.B8n@T-FCN.Net> Sender: news@T-FCN.Net Content-Transfer-Encoding: 7bit Organization: needs one Mime-Version: 1.0 Date: Tue, 24 Nov 1998 22:35:25 GMT I have two problems I can't resolve, and the mailing list didn't turn anything up. One is that when I alloc/init a new sound, I can't record into it. If I instead make a copy of some other sound and write over it, that's fine, but a newly created one doesn't work. Anyone have any ideas on this? The other issue is more "deadly", when I close the window with the soundView in it, I get errors reported in the console then the app quits. It appears that the vu meter is still attempting to draw, even though I've told it to STOP, then set it's sound to nil. Anyone used a soundView before? Maury
From: don@misckit.com (Don Yacktman) Newsgroups: comp.sys.next.programmer Subject: Re: Anyone out there used SoundKit? Date: 24 Nov 1998 22:56:08 GMT Organization: MiscKit Development Message-ID: <73fdi8$mht$2@news.xmission.com> References: <F2y6r2.B8n@T-FCN.Net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NNTP-Posting-Date: 24 Nov 1998 22:56:08 GMT maury@remove_this.istar.ca (Maury Markowitz) wrote: > I have two problems I can't resolve, and the mailing list didn't > turn anything up. > > One is that when I alloc/init a new sound, I can't record into it. > If I instead make a copy of some other sound and write over it, > that's fine, but a newly created one doesn't work. Anyone have any > ideas on this? Can't help you there...but Sean Luke, of Resound.app fame, is an old pro with this junk. Most of the app is really Sean fixing egregious problems in the SoundKit. > The other issue is more "deadly", when I close the window with the > soundView in it, I get errors reported in the console then the app > quits. It appears that the vu meter is still attempting to draw, > even though I've told it to STOP, then set it's sound to nil. > Anyone used a soundView before? I think that's one of the bugs... Take a look at Sean's updated SoundView replacement in the MiscKit, which is the same as what was used in Resound. (NEXTSTEP version there; an OPENSTEP version is in the latest Resound.app source tree.) Sean fixed a veritable ton of serious bugs, so you really ought to be using it instead of wasting time hunting and stomping them yourself... -- Later, -Don Yacktman don@misckit.com <a href="http://www.yacktman.org/don/index.html">My home page</a>
From: malcolm@plsys.co.uk Newsgroups: comp.sys.next.advocacy,comp.sys.mac.advocacy,comp.sys.next.programmer,comp.sys.next.admin Subject: Re: MWSF McOX BOF? Date: Wed, 25 Nov 1998 00:01:47 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <73fhcv$gfs$1@nnrp1.dejanews.com> References: <72ijv1$5uq$1@nnrp1.dejanews.com> In article <72ijv1$5uq$1@nnrp1.dejanews.com>, I wrote: > Heavens, I just had a quick look around [comp.sys.next.advocacy], > and I'm afraid I hardly recognise the neighbourhood. > I just looked back in and saw Lawson's thread about GX -- I feel right at home again... > maybe it'd be worth organising a McOX BOF(*) meeting one evening?> > OK, the plan so far (as discussed on macosx-{talk,dev}): The most likely date/time is the evening of Wednesday January 6. One of the main aims is to make this accessible to anyone, not just those who've paid $1000 registration for the Pro Conference!, so (as per last year, for those of you who were there) this would likely be shortly after the close of the show for the day (so any exhibitors can get along). It looks likely that at least one Apple person will give a presentation, and Mike Ferris has already very kindly agreed to take part in a Q&A session. I'm currently liasing with the Expo folks to get a room. If you'd be *seriously* interested in attending please email me so I can get an idea of numbers; suggestions for discussion topics etc also welcome. I apologise in advance if I don't reply to you; I've already had a large number of replies and it may become overwhelming. When plans start to crystallise I'll put up a page on Stepwise -- http://www.stepwise.com -- to keep folks informed. Best wishes, mmalc <malcolm@plsys.co.uk> P & L Systems -- developers of Mesa http://www.plsys.co.uk/plsys/ Tel: +44 1494 432422 Fax: +44 1494 432478 -----------== Posted via Deja News, The Discussion Network ==---------- http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
From: Lea Len <lea.len@infonie.fr> Newsgroups: comp.programming,comp.os.linux.development.apps,alt.comp.shareware.programmer,alt.games.programming,alt.msdos.programmer,comp.lang.java.programmer,comp.os.msdos.programmer,comp.programming,comp.os.os2.programmer,comp.sys.be.programmer,comp.sys.next.programmer,comp.unix.programmer,cybergate.programmers,de.comp.os.os2.programmer,de.comp.os.unix.programming,fj.os.ms-windows.programming,fr.comp.os.ms-windows.programmation,francom.programmation,linux.dev.c-programming Subject: Manu1002 Recontact me please. Date: Wed, 25 Nov 1998 09:54:40 +0100 Organization: Ye 'Ol Disorganized NNTPCache groupie Message-ID: <365BC5D0.2A6A568B@infonie.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cache-Post-Path: si2-paris!unknown@212.232.23.3 Manu, I couldn't join your account at aol. (unknown user) And on the t-online i didn't get you too . (*rom#) ? Please recontact me and let me a address where to join you. Lea
Newsgroups: comp.sys.next.programmer From: maury@remove_this.istar.ca (Maury Markowitz) Subject: Re: Anyone out there used SoundKit? Content-Type: text/plain; charset=us-ascii Message-ID: <F2zK28.5A9@T-FCN.Net> Sender: news@T-FCN.Net Content-Transfer-Encoding: 7bit Cc: don@misckit.com Organization: needs one References: <F2y6r2.B8n@T-FCN.Net> <73fdi8$mht$2@news.xmission.com> Mime-Version: 1.0 Date: Wed, 25 Nov 1998 16:20:31 GMT In <73fdi8$mht$2@news.xmission.com> Don Yacktman wrote: > Can't help you there...but Sean Luke, of Resound.app fame, is an old pro > with > this junk. Most of the app is really Sean fixing egregious problems in > the SoundKit. Do you have his address? > Take a look at Sean's updated SoundView replacement in the MiscKit, which > is > the same as what was used in Resound. (NEXTSTEP version there; an > OPENSTEP > version is in the latest Resound.app source tree.) Sean fixed a > veritable > ton of serious bugs, so you really ought to be using it instead of > wasting > time hunting and stomping them yourself... Oh, ok, I'll do that. Maury
Newsgroups: comp.sys.next.programmer From: maury@remove_this.istar.ca (Maury Markowitz) Subject: Re: Anyone out there used SoundKit? Content-Type: text/plain; charset=us-ascii Message-ID: <F300BF.EKo@T-FCN.Net> Sender: news@T-FCN.Net Content-Transfer-Encoding: 7bit Cc: don@misckit.com Organization: needs one References: <F2y6r2.B8n@T-FCN.Net> <73fdi8$mht$2@news.xmission.com> Mime-Version: 1.0 Date: Wed, 25 Nov 1998 22:11:39 GMT In <73fdi8$mht$2@news.xmission.com> Don Yacktman wrote: > Take a look at Sean's updated SoundView replacement in the MiscKit, which > is I just finished wiring it up. Aside from some minor documentation-reading problems, I had no problems getting it up and running. It also fixed both of the bugs I mentioned, and I now have fully functional sound editing in Glyph!X. Not bad for 2 hours work! Maury
From: "Jason Shelley" <jshelley@enersol.com.au> Newsgroups: comp.programming,comp.os.linux.development.apps,Programmers,alt.comp.shareware.programmer,alt.games.programming,alt.msdos.programmer,comp.lang.java.programmer,comp.os.msdos.programmer,comp.os.os2.programmer,comp.sys.be.programmer,comp.sys.next.programmer Subject: Re: Looking for programmers with the "love"... Date: Thu, 26 Nov 1998 09:41:49 +1100 Organization: Telstra Big Pond Cable - Sydney Site 1 Message-ID: <73i13v$t32$1@m2.c2.telstra-mm.net.au> References: <3644BC21.F325CC46@tnitech.com> <364CCD4A.A4B193AF@infonie.fr> NNTP-Posting-Date: 25 Nov 1998 22:42:07 GMT James, I have been thinking about the same thing for a little while. I currently write programs in C++ for condom testing equipment, and work on lots of stuff for fun. I've often thought it would be a good idea to get a group of like minded people together and make some MONEY ;) Anyway, count me in. I haven't had any ideas for a project yet, but the concept is a winner :) I'd be prepared to spend at least 8 hours a week on a big cooperative project (I do it anyway ;) ). Cheers, Jason Shelley (Jase) >James Conrad wrote: > >> I'm looking for dedicated programmers willing to be part of a >> multi-platform distributed development team. People who are tired of >> large software development companies controlling the consumer trends. >> Long live the voice of the people! >> -james > > >
From: G.C.Th.Wierda@AWT.nl (Gerben Wierda) Newsgroups: comp.sys.next.programmer,comp.sys.next.sysadmin Subject: Upgrading OPENSTEP's (and NEXTSTEP's) cc to latest gcc? Date: Thu, 26 Nov 1998 08:59:12 GMT Organization: Adviesraad voor het Wetenschaps- en Technologiebeleid Sender: news@AWT.NL Message-ID: <F30uAo.9x@AWT.NL> References: <F30tKF.n3u@AWT.NL> I was wondering if it is possible nowadays to upgrade your CC to a recent GCC without losing the capabilities of cross-compiling, i.e. a real replacement of the native CC and not just an added separate gcc (or at least a gcc that can handle the normal -arch statements). Can this be done? Thanks, --- Gerben Wierda, Stafmedewerker Adviesraad voor het Wetenschaps- en Technologiebeleid. Staff member Advisory Council for Science and Technology Policy Javastraat 42, 2585 AP, Den Haag/The Hague, The Netherlands Tel (+31) 70 3639922 Fax (+31) 70 3608992 http://www.AWT.nl/prive/wierda/ The fox knows many things, but the hedgehog knows one big thing. ! This message does not contain any official AWT-views unless ! explicitely mentioned as such. For the official AWT-views, ! please consult the AWT web site: http://www.AWT.nl/
From: echo4@NOSPAM.erols.com --delete NOSPAM Newsgroups: comp.sys.next.programmer,comp.sys.next.sysadmin Subject: Re: Upgrading OPENSTEP's (and NEXTSTEP's) cc to latest gcc? Date: 27 Nov 1998 04:11:05 GMT Message-ID: <73l8op$p9m$2@winter.news.rcn.net> References: <F30tKF.n3u@AWT.NL> <F30uAo.9x@AWT.NL> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NNTP-Posting-Date: 27 Nov 1998 04:11:05 GMT Cc: G.C.Th.Wierda@AWT.nl In <F30uAo.9x@AWT.NL> Gerben Wierda wrote: > I was wondering if it is possible nowadays to upgrade your CC to a recent GCC > without losing the capabilities of cross-compiling, i.e. a real replacement > of the native CC and not just an added separate gcc (or at least a gcc that > can handle the normal -arch statements). > > Can this be done? > > Thanks, > > --- > Gerben Wierda, > > Stafmedewerker Adviesraad voor het Wetenschaps- en Technologiebeleid. > Staff member Advisory Council for Science and Technology Policy > Javastraat 42, 2585 AP, Den Haag/The Hague, The Netherlands > Tel (+31) 70 3639922 Fax (+31) 70 3608992 > http://www.AWT.nl/prive/wierda/ > > The fox knows many things, but the hedgehog knows one big thing. > > ! This message does not contain any official AWT-views unless > ! explicitely mentioned as such. For the official AWT-views, > ! please consult the AWT web site: http://www.AWT.nl/ > Gerben, Rex Dieter has done alot in this area, you should e-mail him. rdieter@NOSPAM.math.unl.edu Remove the NOSPAM. Rex has done the NeXT and Openstep ports of GCC. Jay -- |========================================| J. Lee echo4@NOSPAM.erols.com --delete NOSPAM NeXTMail,SUN,MIME |========================================|
From: ALBERTO ORTEGA SOCORRO <8903525905#alorso@bbvnet.com> Newsgroups: comp.sys.next.programmer Subject: linking with Project Builder Date: Fri, 27 Nov 1998 22:39:11 +0000 Organization: BBV Message-ID: <365F2A0F.79FD@bbvnet.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I want to make a tool with several source files, using Project Builder. I need to use the LEX library (libl.a) with the switch -ll. Linker cannot build the tool because in the library code there is an object file which has a main function (main.o). The error message is: bin/ld: multiple definitions of symbol _main /usr/lib/libl.a(main.o) definition of _main in section (__TEXT,__text) /me/imcc/prac3.tproj/obj-i386-opt/perceptron.o definition of _main in section (__TEXT,__text) There is other main apart of the main of my program! How can I solve this problem? My OS is Openstep 4.1.
From: theisen@akaMail.com (Dirk Theisen) Newsgroups: comp.sys.next.programmer Subject: Re: Java and ProjectBuilder Date: Fri, 27 Nov 1998 16:11:59 +0100 Organization: University of Bonn, Germany Message-ID: <1dj4zjt.1re8ps7bv778uN@ascend-tk-p25.rhrz.uni-bonn.de> References: <737mou$edp$1@dns2.serv.net> Hallo, Dean! Dean Johnson <dean@u.washington.edu> wrote: > I'm trying to learn a little Java for some work related stuff and am > trying to do it under ProjectBuilder (RDR2 on NT), but have a problem. I can only speak for RDR2/PPC here: > I am having a hard time accessing some 3rd party classess from > ProjectBuilder. Normally I can set CLASSPATH and then the directory > and the Java compiler will look there for the import. How do you do > this in PB? I've tried bringing the classes into Other Resources and I > tried to set a path in Build Attributes. AFAIK, PB does not inherit the CLASSPATH set in your environment. It takes the std. java path out of a javaenv.plist (?) located somewhere in the System folder (cannot check right now). I included /Local/Library/Java and ~/Library/Java there and simply put (extracted) libs there for general use (java program installation). For use in PB only (and for PB to package the lib with the app on installation) I put the extracted libs for project xxx into xxx/xxx.build/java_classes/. You have to redo that every time You do a cleanup ("make all"). Hope it helps. Regards, Dirk -- No RISC - No fun http://theisen.home.pages.de/
Message-ID: <365F68DA.F25B49DD@deltos.com> From: "Steven R. Staton" <steve@deltos.com> Organization: Deltos Fleet Computing MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Missing inet.h Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Sat, 28 Nov 1998 03:07:07 GMT NNTP-Posting-Date: Fri, 27 Nov 1998 19:07:07 PDT I'm looking for a copy of the NS 3.3/Moto inet.h file. I managed to mangle mine while trying to build sendmail 8.8.5.
From: moncrieffe.martin@mayo.edu (Martin Moncrieffe) Newsgroups: comp.sys.next.programmer Subject: Project Builder (OS 4.2) Date: 28 Nov 1998 18:59:56 GMT Organization: Mayo Foundation Message-ID: <73ph7c$eua$1@tribune.mayo.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi. Is it possible to set up alternative project types in Project Builder (OS 4.2)? Could one, for example, set up a TeX project type which would have the "Other Sources" containing *.tex files; "Images" containing *.eps files for inclusion into the compiled document etc.? Regards. __martin
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <24666911710829@digifix.com> Date: 29 Nov 1998 04:45:35 GMT Organization: Digital Fix Development Message-ID: <18755912315622@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1994. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: emclean@slip.net (Emmett McLean) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.software,comp.sys.next.programmer Subject: Re: gcc's libgcc.a on a m68k-next-nextstep3 - SOS Date: 29 Nov 1998 19:11:34 -0800 Organization: Slip.Net Message-ID: <73t2d6$rgp@slip.net> References: <73ron5$amd$1@mathserv.mps.ohio-state.edu> Interesting that you ask. I ran into this problem when installing perl with gcc on NS 3.3. I ended up removing gcc, reinstalling the 3.3 patches, then installing gcc. Experimented with various versions of libgcc.a to no avail. Was a drag. Emmett
From: "Markus Vollmert" <uzssuz@uni-bonn.de> Newsgroups: comp.sys.next.programmer Subject: nextstep and palmpilot Date: Sun, 29 Nov 1998 16:00:57 +0100 Organization: RHRZ - University of Bonn (Germany) Message-ID: <73rnjf$f74@news.rhrz.uni-bonn.de> hi i just got my new palmIII a few days ago. i searched for some development tools and found the gnu developer kit for win95. i installed this and recognized that it contains the gnu cc compiler with some unix tools and new libaries. now i am wondering if i could also use this set on nextstep for developing for the pilot. does anybody use this set with ns3.3 for intel? is there a chance of connecting the pilot to nextstep? did anybody compile the xcopilot for ns? bye markus vollmert
From: nevai@math.ohio-state.edu (Paul Nevai) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.software,comp.sys.next.programmer Subject: gcc's libgcc.a on a m68k-next-nextstep3 - SOS Date: 29 Nov 1998 15:20:05 GMT Organization: Department of Mathematics, The Ohio State University Message-ID: <73ron5$amd$1@mathserv.mps.ohio-state.edu> NNTP-Posting-Date: 29 Nov 1998 15:20:05 GMT Originator: nevai@math.ohio-state.edu (Paul Nevai) Please e-mail me your /usr/local/lib/gcc-lib/m68k-next-nextstep3/2.8.1/libgcc.a Mine is defective and I can't fix it. It maybe a gcc bug on OS 32. Thanks! Best Regards, Paul Paul Nevai nevai@math.ohio-state.edu Dept Math - Ohio State University 1-614-292-5310 (Office/Answering Device) Columbus, Ohio 43210-1174, U.S.A. 1-614-292-1479 (Math Dept Fax)
From: Gerben_Wierda@RnA.nl Newsgroups: comp.sys.next.programmer Subject: lock_data_t in <sys/user.h>? Date: Sun, 29 Nov 1998 02:06:11 GMT Organization: R&A Sender: news@RnA.nl Message-ID: <F35v6C.9L9@RnA.nl> Hmm, in <sys/user.h> on OPENSTEP 4.2 there is a reference to a lock_data_t type (line 135), but I cannot find the definition anywhere. I did find simple_lock_data_t in <mach/simple_lock.h> (actually in the machine subdir). Should I just change lock_data_t in simple_lock_data_t in <sys/user.h> and hope for the best? I tried it and it seems to do no harm. -- Gerben_Wierda@RnA.nl (Gerben Wierda) "If you don't know where you're going, any road will take you there" Paraphrased in Alice in Wonderland, originally from the Talmud. "Your io is pretty std" -- Larry Wall
From: holger@_REMOVE_THIS_.wizards.de (Holger Hoffstaette) Newsgroups: comp.sys.next.programmer Subject: Re: lock_data_t in <sys/user.h>? Date: 30 Nov 1998 01:35:31 GMT Organization: The secret circle of the NSRC Message-ID: <73ssp3$2qb@ragnarok.en.uunet.de> References: <F35v6C.9L9@RnA.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Gerben_Wierda@RnA.nl wrote: > Hmm, in <sys/user.h> on OPENSTEP 4.2 there is a reference to a lock_data_t > type (line 135), but I cannot find the definition anywhere. I did find > simple_lock_data_t in <mach/simple_lock.h> (actually in the machine subdir). That seems to have been removed from the headers, correct. You can just copy the old definition from a 3.x box. This was the missing data structure that kept lsof from working correctly on 4.2. > Should I just change lock_data_t in simple_lock_data_t in <sys/user.h> and > hope for the best? I tried it and it seems to do no harm. Won't work, they're different. Been there, done that. :) The simple_lock_data_t is just a wrapped int, whereas the real lock_data_t has some more fields (for juggling threads): typedef struct lock { simple_lock_data_t interlock; unsigned int read_count:16, want_upgrade:1, want_write:1, waiting:1, can_sleep:1, recursion_depth:12; char *thread; } lock_data_t; have fun, Holger
From: Rex Dieter <rdieter@math.unl.edu> Newsgroups: comp.sys.next.sysadmin,comp.sys.next.software,comp.sys.next.programmer Subject: Re: gcc's libgcc.a on a m68k-next-nextstep3 - SOS Date: 30 Nov 1998 15:33:57 GMT Organization: University of Nebraska-Lincoln Message-ID: <73udt5$1i0$1@unlnews.unl.edu> References: <73ron5$amd$1@mathserv.mps.ohio-state.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit >Please e-mail me your >/usr/local/lib/gcc-lib/m68k-next-nextstep3/2.8.1/libgcc.a > >Mine is defective and I can't fix it. It maybe a gcc bug on OS 32. Thanks! >Best Regards, Paul What exactly is the problem? If you get a warning about unsorted members of the library, that is harmless, and mostly unfixable (see gcc-i386.2.8.1.2.README for details). -- Rex A. Dieter rdieter@math.unl.edu Computer System Manager http://www.math.unl.edu/~rdieter/ Mathematics and Statistics University of Nebraska-Lincoln
From: "Mark Lewis" <mlewis@worldnetcorp.com> Newsgroups: comp.sys.next.programmer Subject: Next Programmer Wanted Date: Mon, 30 Nov 1998 14:02:57 -0600 Organization: Worldnet Corporation Message-ID: <73uu2b$1s7$1@supernews.com> Dear Consultant, Our client in Downtown Chicago is seeking experienced Nextstep / Openstep / Objective C programmers for long term assignments. The client is seeking Consultants with 3+ years experience and are interviewing immediately. Please send all resumes including rates to: mlewis@worldnetcorp.com Thanks and have a great day. Mark - 847-839-9323-Direct.
From: holger@_REMOVE_THIS_.wizards.de (Holger Hoffstaette) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.software,comp.sys.next.programmer Subject: Re: gcc's libgcc.a on a m68k-next-nextstep3 - SOS Date: 30 Nov 1998 17:58:58 GMT Organization: The secret circle of the NSRC Message-ID: <73umd2$aav@ragnarok.en.uunet.de> References: <73ron5$amd$1@mathserv.mps.ohio-state.edu> <73udt5$1i0$1@unlnews.unl.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Rex Dieter wrote: > >Please e-mail me your > >/usr/local/lib/gcc-lib/m68k-next-nextstep3/2.8.1/libgcc.a > > > >Mine is defective and I can't fix it. It maybe a gcc bug on OS 32. Thanks! > >Best Regards, Paul > > What exactly is the problem? If you get a warning about unsorted members of > the library, that is harmless, and mostly unfixable (see > gcc-i386.2.8.1.2.README for details). Just in case people don't know this: one can make gcc shut up by putting a "-w" into the link section of the specs file. This will prevent the linker from complaining about warnings, and is also a nice way to prevent the normal 4.2 compiler from complaining when linking -dynamic code against a static library. The latter can also be achieved by adding "-read_only_relocs suppress". Holger
Subject: header files and 'cercles vicieux' Newsgroups: comp.sys.next.programmer From: email@end.of.post (Raymond Lutz) Message-ID: <JcU82.79$x96.415@wagner.videotron.net> Date: Tue, 01 Dec 1998 15:47:53 GMT NNTP-Posting-Date: Tue, 01 Dec 1998 10:47:53 EDT Bonjour a vous tous, (maybe this is more relevant to comp.lang.obj-c... anyway) I'm using a 'elsewhere declared' struct in a category interface and the file declaring the struct also use that category in a function declaration ... I think one uses forward declaration to break this cercle vicieux, but It seems I can't do a forward declaration like @class NSValue(NiftyCategory); in the file using the category (cc chokes). For now I moved the function declaration elsewhere, but there must be a better way! Thanks Ray -- Raymond Lutz - 9bit.qc.ca@lutzray - www.9bit.qc.ca/~$myusername - "Les 400 plus fortunes individus de la planete possedent autant que 2.3 MILLIARDS des plus pauvres reunis"
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: header files and 'cercles vicieux' Date: 1 Dec 1998 16:28:13 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn768691.4uo.rog@talisker.ohm.york.ac.uk> References: <JcU82.79$x96.415@wagner.videotron.net> NNTP-Posting-Date: 1 Dec 1998 16:28:13 GMT On Tue, 01 Dec 1998 15:47:53 GMT, Raymond Lutz <email@end.of.post> wrote: > I'm using a 'elsewhere declared' struct in a category interface and > the file declaring the struct also use that category in a function > declaration ... what do you mean to "use a category in a function declaration"? as far as i'm aware, no function declaration can contain references to anything other than class names - category names in Objective-C are used for nothing at all as far as i'm aware. > I think one uses forward declaration to break this cercle vicieux, > but It seems I can't do a forward declaration like > > @class NSValue(NiftyCategory); but you can do @class NSValue; and to forward-declare a typedef'ed struct, you can do typedef struct SomeStruct SomeStruct; and you can then use pointers to SomeStruct without including a struct definition publicly. hope this helps, (you'll have to be a bit clearer if it doesn't; try including some of the relevant declarations/definitions) cheers, rog.
Subject: C programming 101 (Was Re: header files and 'cercles vicieux') Newsgroups: comp.sys.next.programmer References: <JcU82.79$x96.415@wagner.videotron.net> <slrn768691.4uo.rog@talisker.ohm.york.ac.uk> In-Reply-To: <slrn768691.4uo.rog@talisker.ohm.york.ac.uk> From: email@end.of.post (Raymond Lutz) Message-ID: <NwX82.85$x96.568@wagner.videotron.net> Date: Tue, 01 Dec 1998 19:34:05 GMT NNTP-Posting-Date: Tue, 01 Dec 1998 14:34:05 EDT On 12/01/98, Roger Peppe wrote: >On Tue, 01 Dec 1998 15:47:53 GMT, Raymond Lutz <email@end.of.post> >wrote: >> I'm using a 'elsewhere declared' struct in a category interface >> and the file declaring the struct also use that category in a >> function declaration ... > >what do you mean to "use a category in a function declaration"? >as far as i'm aware, no function declaration can contain references >to anything other than class names - category names in Objective-C >are used for nothing at all as far as i'm aware. > Thank for your answer Roger! Ok, I'll be more explicit (for the benefit of other newbies like me). Maybe I'm simply going to explain want I'm trying to do.. And someone could tell the best way to do it, and better places to define/declare things. I would like to use a struct RLVector here and there in my app and some associated functions like RLMakeVector(), RLStringFromVector() etc... I created a file mimicking <Foundation/NSGeometry.h>, called it RLGeometry.h and put it in my Project. But is it OK to define function implementation in header files? Isn't the compiler doing repetitive work for all the files where the header is imported? OK, static functions declared in <Foundation/NSGeometry.h> are one line long, but mine are bigger. I guess this is what libraries are for... Does it show up that I learned programming reading cereal boxes? 8^) Well suppose for now my custom structs and functions are too modest to be built into a library, so I'll duplicate them from project to project. Then, I'll put typedef and function type declarations (with extern in place of static storage class specifier) in RLGeometry.h; and put function definitions in RLGeometry.c, right? Ok, now for my category problem: In RLGeometry.c I define a function: ----------------------------------------- #import "RLGeometry.h" #import "RLVectorValue.h" void aDummyFunction( NSArray *anArray) { NSValue * aValue; aValue = [anArray lastObject]; NSLog(@"%@", RLStringFromVector( [aValue RLVector]) ); } --------------------------------------------- where RLVector in [aValue RLVector] is a NSValue category method used to get my struct out of its NSValue, as declared in RLVectorValue.h -------------------------------------------------- #import "RLGeometry.h" @interface NSValue(RLVectorValue) + (NSValue*)RLVectorValueWithXYZComponents:(float)x:(float)y:(float)z; - (RLVector)RLVector; @end --------------------------------------------------- How can I break this cyclical reference? For now, the compiler reports that in RLGeometry.c at RLStringFromVector( [aValue RLVector]) 'aValue' doesn't respond to 'RLVector' and RLStringFromVector() doesn't have the right type argument, although I had imported the category extension "RLVectorValue.h" . Thanks for your help and indications! Ray -- Raymond Lutz - 9bit.qc.ca@lutzray - www.9bit.qc.ca/~$myusername - "Les 400 plus fortunes individus de la planete possedent autant que 2.3 MILLIARDS des plus pauvres reunis"
Subject: Re: C programming 101 (Was Re: header files and 'cercles vicieux') Newsgroups: comp.sys.next.programmer References: <JcU82.79$x96.415@wagner.videotron.net> <slrn768691.4uo.rog@talisker.ohm.york.ac.uk> <NwX82.85$x96.568@wagner.videotron.net> In-Reply-To: <NwX82.85$x96.568@wagner.videotron.net> From: email@end.of.post (Raymond Lutz) Message-ID: <wAY82.93$x96.691@wagner.videotron.net> Date: Tue, 01 Dec 1998 20:46:20 GMT NNTP-Posting-Date: Tue, 01 Dec 1998 15:46:20 EDT On 12/01/98, I wrote: > >Then, I'll put typedef and function type declarations (with extern in >place of static storage class specifier) in RLGeometry.h; and put >function definitions in RLGeometry.c, right? > >Ok, now for my category problem: > well... stripping function definitions out of RLGeometry.h and putting them in RLGeometry.m did also solve my method category declaration being ignored. Ray PS: anyone, don't hesitate to give me any advice on general declaration/definition setup, though. -- Raymond Lutz - 9bit.qc.ca@lutzray - www.9bit.qc.ca/~$myusername - "Les 400 plus fortunes individus de la planete possedent autant que 2.3 MILLIARDS des plus pauvres reunis"
From: Alberto Ortega Socorro <alorso@bbvnet.com> Newsgroups: comp.sys.next.programmer Subject: Debuggers, SuperDebugger 3.9 for Openstep Date: 1 Dec 1998 23:28:24 GMT Organization: Acceso Internet TSAI Message-ID: <741u2o$obg5@esiami.tsai.es> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I am looking for a graphical debugger for Openstep 4.1 for Mach. My first choice was SuperDebugger 3.9 but it doesnÂt work. Anyway if you acceed to the SuperDebugger 3.9 page of the server www.scribex .com (yet accesible from sites like Altavista) it says: "system requirements are NEXTSTEP 3.1 or later, or OPENSTEP 4.x for MACH" which suggests that exist SuperDebugger 3.9 for Openstep. Is this true? If so, how can I get it? If not, please recommend me a good graphical debugger for Openstep. Thanks, Alberto Ortega Socorro.
From: "Charles W. Swiger" <chuck@codefab.com> Newsgroups: comp.sys.next.programmer Subject: Re: C programming 101 (Was Re: header files and 'cercles vicieux') Date: 1 Dec 1998 21:10:23 GMT Organization: Spacelab.net Internet Access Message-ID: <741lvv$n5r$1@news.spacelab.net> References: <JcU82.79$x96.415@wagner.videotron.net> <slrn768691.4uo.rog@talisker.ohm.york.ac.uk> <NwX82.85$x96.568@wagner.videotron.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit email@end.of.post (Raymond Lutz) wrote: [ ... ] > I created a file mimicking <Foundation/NSGeometry.h>, called it > RLGeometry.h and put it in my Project. > > But is it OK to define function implementation in header files? Not really, no. It's fine to declare macros, and it's sloppy but it won't break anything to declare one-liners if they don't maintain state. If they do maintain state (via a static variable, say), then every file that includes that header will compile in a _different_ function which maintains _different_ state. And your problem will lose. > Isn't the compiler doing repetitive work for all the files where the > header is imported? Yes. > OK, static functions declared in <Foundation/NSGeometry.h> are one line > long, but mine are bigger. They belong in a .m/.c file, then. > I guess this is what libraries are for... Does it show up that I > learned programming reading cereal boxes? 8^) Or frameworks, or even static .o files linked into the executable. >Then, I'll put typedef and function type declarations (with extern in >place of static storage class specifier) in RLGeometry.h; and put >function definitions in RLGeometry.c, right? Right. [ ...RLVector stuff... ] You typecast the instance method RLVector in your category to return a (RLVector), but it didn't appear that your code provides a definition for that type. Categories add functionality, but not state (ivars). You might want to declare RLVector as a subclass of NSValue, instead.... -Chuck Charles Swiger | chuck@codefab.com | Yeah, yeah-- disclaim away. ----------------+-------------------+---------------------------- You have come to the end of your journey. Survival is everything.
Newsgroups: comp.sys.next.programmer From: mvlems@vbox.xs4all.nl.NO.SPAM Subject: Problem connecting EOF to Linux/Sybase Content-Type: text/plain; charset=us-ascii Message-ID: <1998Dec2.080813.585@vbox.xs4all.nl.NO.SPAM> Sender: mvlems@vbox.xs4all.nl.NO.SPAM (Mark F. Vlems) Content-Transfer-Encoding: 7bit Mime-Version: 1.0 Date: Wed, 2 Dec 1998 08:08:13 GMT Howdy, On a Linux computer I have a working Sybase server (latest version available). I would like to connect from my OPENSTEP 4.2/intel computer to the Sybase server from within EOModeler (EOF 2.1), but get the error: Sybase: ct_connect(): network packet layer: internal net library error: Could not find addressing dictionary Operating System Error - 84682192 This error was mentioned in this newsgroup before, and the solution would be to change/add a /usr/sybase/interface file. I guess that server and portnumber info should be in it, but what format should it have? Do I need more than this? Any help appreciated. TIA, -- Mark F. Vlems Poor is the man, MIME Mail OK Whose pleasures depend NeXTMail preferred! On the permission of another. Fax: +31 (0) 23 5622368 Madonna in "Justify my love", 1990
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: C programming 101 (Was Re: header files and 'cercles vicieux') Date: 2 Dec 1998 14:51:55 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn76al0n.65h.rog@talisker.ohm.york.ac.uk> References: <JcU82.79$x96.415@wagner.videotron.net> <slrn768691.4uo.rog@talisker.ohm.york.ac.u <NwX82.85$x96.568@wagner.videotron.net> NNTP-Posting-Date: 2 Dec 1998 14:51:55 GMT On Tue, 01 Dec 1998 19:34:05 GMT, Raymond Lutz <email@end.of.post> wrote: > But is it OK to define function implementation in header files? Isn't > the compiler doing repetitive work for all the files where the header > is imported? OK, static functions declared in > <Foundation/NSGeometry.h> are one line long, but mine are bigger. the reason those functions are defined in NSGeometry.h are there only because they're declared inline. in general, it's not worth using inline functions unless you've determined that those functions are definitely a bottleneck in your code. (probably by profiling the code with gprof(1)). the main drawbacks to inline functions are: a) if you want to change the function, you've got to recompile every file that calls the function. b) when debugging, if the program crashes in an inline function, the debugger won't show which function you're in. c) if your inline functions are big and you use them a lot, then it's quite possible that you actually slow down the code, because the text space of the program will be larger and therefore less likely to fit into cache. calling a function is very efficient anyway, so i really wouldn't bother following NeXT's example from NSGeometry.h until your code is very stable and you're sure it'll give a big speed up. > I guess this is what libraries are for... this is indeed what libraries are for - and if you are using your utility functions in many projects, making your own library is a good way to go. it's fairly easy to set up, as long as you have some familiarity with make(1). the advantage of putting them into a library is that only the code in files containing the functions used will be dragged into the final executable. > Then, I'll put typedef and function type declarations (with extern in > place of static storage class specifier) in RLGeometry.h; and put > function definitions in RLGeometry.c, right? yup. > Ok, now for my category problem: > > In RLGeometry.c I define a function: > ----------------------------------------- > #import "RLGeometry.h" > #import "RLVectorValue.h" > void aDummyFunction( NSArray *anArray) > { > NSValue * aValue; > > aValue = [anArray lastObject]; > NSLog(@"%@", RLStringFromVector( [aValue RLVector]) ); > } > --------------------------------------------- > where RLVector in [aValue RLVector] is a NSValue category method used > to get my struct out of its NSValue, as declared in > > RLVectorValue.h > -------------------------------------------------- > #import "RLGeometry.h" > @interface NSValue(RLVectorValue) > + (NSValue*)RLVectorValueWithXYZComponents:(float)x:(float)y:(float)z; > - (RLVector)RLVector; > @end you shouldn't really have a problem with this. there's no unavoidable cyclic reference: RLGeometry.h declares the RLVector type. RLVectorValue.h declares the methods in your custom category, using RLVector RLGeometry.c includes the above and uses methods in the category, no problem. at least, i can't see one. cheers, rog.
From: laurent BRIET <laurent@buf.imaginet.fr> Newsgroups: comp.sys.next.programmer Subject: Motif Date: Wed, 02 Dec 1998 21:14:01 -0100 Organization: ImagiNET Message-ID: <3665BBA9.41C6@buf.imaginet.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I was just wondering how easy can it be to port Unix application developped using motif to MacOS X ??? Do you know if it's possible and if yes, how?
Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <Dfx92.3760$Z72.1185245@tundra.ops.attcanada.net> ignore no reply Control: cancel <Dfx92.3760$Z72.1185245@tundra.ops.attcanada.net> Message-ID: <cancel.Dfx92.3760$Z72.1185245@tundra.ops.attcanada.net> Date: Thu, 03 Dec 1998 16:31:59 +0000 Sender: jnkkmq@addr.com From: andrew@erlenstar.demon.co.uk Organization: Annihilator v0.3 Spam (EMP) cancelled.
From: bobs@terrapin.phys.washington.edu (Robert Singleton) Newsgroups: comp.sys.next.programmer Subject: Getting Started Date: 3 Dec 1998 21:02:58 GMT Organization: R'lyeh Message-ID: <746ua2$14l2$1@nntp6.u.washington.edu> I'm running NS3.2 (developer) on the old black hardware, and I would like to start programming in NeXTStep, but I have no experience. Where could I find an introductory tutorial? To start out, for example, I would like create a simple reminder program that pops up a message box at a specified time. Thanks for any suggestions. -- Robert Singleton work: (206) 543-9640 Department of Physics fax : (206) 685-0635 University of Washington office: B416 Phys & Astronomy Bldg. Box 351560 bobs@terrapin.phys.washington.edu
From: Gerben_Wierda@RnA.nl Newsgroups: comp.sys.next.programmer Subject: Re: lock_data_t in <sys/user.h>? Date: Thu, 3 Dec 1998 22:50:10 GMT Organization: R&A Sender: news@RnA.nl Message-ID: <F3EvFM.8xG@RnA.nl> References: <F35v6C.9L9@RnA.nl> <73ssp3$2qb@ragnarok.en.uunet.de> holger@_REMOVE_THIS_.wizards.de (Holger Hoffstaette) wrote: > Gerben_Wierda@RnA.nl wrote: >> Hmm, in <sys/user.h> on OPENSTEP 4.2 there is a reference to a lock_data_t >> type (line 135), but I cannot find the definition anywhere. I did find >> simple_lock_data_t in <mach/simple_lock.h> (actually in the machine subdir). > >That seems to have been removed from the headers, correct. You can just >copy the old definition from a 3.x box. This was the missing data structure >that kept lsof from working correctly on 4.2. > >> Should I just change lock_data_t in simple_lock_data_t in <sys/user.h> and >> hope for the best? I tried it and it seems to do no harm. > >Won't work, they're different. Been there, done that. :) It did get top 0.5 working... I wonder, if this is kernel structure, would the 3.3 struct in a 4.2 environment really help? Or is it a typo and is the newer kernel structure indeed simple_lock_t? Just wondering because here it did actually work for top with simple_lock_data_t instead of lock_data_t. -- Gerben_Wierda@RnA.nl (Gerben Wierda) "If you don't know where you're going, any road will take you there" Paraphrased in Alice in Wonderland, originally from the Talmud. "Your io is pretty std" -- Larry Wall
From: F.Knobloch@inis.de (Frank Knobloch) Newsgroups: comp.sys.next.programmer Subject: To all developers and companies that may be interested Date: 4 Dec 98 11:18:28 GMT Organization: Ye 'Ol Disorganized NNTPCache groupie Message-ID: <3667c504.0@m3-hh-alpha.03.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cache-Post-Path: news-fa!unknown@194.75.26.11 To all developers and companies that may be interested, the actual situation of the Openstep market (respectively Openstep, Yellow-Box, EOF and WOF) requires developers with expertise in this field that cannot be found easily. Like our company, there are other firms which are interested in training programmers which are new to this technologies. InIS is currently developing a product, called Requester, that is based on the famous technology of WebObjects 4.0. Requester is a system for building, using and administrating internet-based knowledge databases, which is in final beta and will be released early next year. A knowledge-based supportsystem has much more advantages than news or mailinglists, cause mostly it would be better to get informations in a more flexible way. We are intending to create a web site based on this technology at no cost. It will be possible to obtain a quick answer to questions related to Openstep programming, by flexible searching or by getting answers from experts. As with OmniGroup's mailing lists, which can either be browsed online or downloaded as m-box files, the contents of this database will be available to anyone interested online. If she/he deems it more convenient she/he will be able to download all of parts of the database in a public format. So the collected knowledge is for everyone who is interested in it. The purpose of this mail is to 1. mention to you the availability of this new service, and 2. to invite any interested developer which considers herself/itself an expert in any of the aforementioned fileds to join the system as such. In the first major-release the experts of the Requester must be skilled developers, but in future releases we will build up an expert-system, based on the collected knowledge. If you are in any way interested (as an expert or an user or both) in this offering please let us know. Regards Frank Knobloch -- //InIS GmbH * Loesungen fuer Automations- und Verwaltungssysteme [Addr.: Friedrich-Ebert-Str. 78 ~ D-34119 Kassel] [Tel..: ++49-561-788090, Fax: ++49-561-78 80 929] [Mail.: F.Knobloch@inis.de, Web: www.inis.de] //Never trust existing sources!
From: conrad@cgl.ucsf.edu (Conrad Huang %CGL) Newsgroups: comp.sys.next.programmer,comp.sys.next.software Subject: Need help mapping device memory into IO task Date: 4 Dec 98 20:53:34 GMT Organization: UCSF Computer Graphics Lab Message-ID: <conrad.912804814@cgl.ucsf.edu> We have been trying to implement a device driver for a PCI card (specifically, an Industry Pack Carrier). The card maps some device registers into PCI bus memory space. I'm trying to map the PCI memory space into the kernel I/O task so that I can manipulate the device. Using DriverKit, I can call IOMapPhysicalIntoIOTask and successfully get a virtual address for the physical address I get out of the PCI configuration structure. However, when I try to move data from the virtual address into a local variable, I invariably get data with all bits set to 1 (suspiciously like PCI_DEFAULT_DATA which is defined as 0xffffffff). Some LED indicators on the PCI board also suggest that I am not actually talking to it. Can someone enlighten me as to what I may be doing wrong? Thanks, Conrad
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: NSDrawBitmap() and 5-bps. Organization: Is a sign of weakness Distribution: world Message-ID: <SCOTT.98Dec4170901@slave.doubleu.com> Date: 4 Dec 98 17:09:01 NNTP-Posting-Date: Fri, 04 Dec 1998 17:26:15 PDT In the interests of being fast on different screens, I've been experimenting with 555/16 mode. Since it's easy for me to modify whether the code's spitting out 444/16 or 888/32, I figured I'd add 555/16 _if_ that's what the screen is at. Unfortunately, everytime I do something like: NSDrawBitmap( rect, rect.size.width, rect.size.height, // pixelsWide/High bitsPerSample, // bitsPerSample 3, // samplesPerPixel bitsPerPixel, // bitsPerPixel bytesPerRow, // bytesPerRow NO, NO, // isPlanar, hasAlpha NSDeviceRGBColorSpace, // colorSpace data); // data with bitsPerSample==5, well, my windowserver dies. When I set everything up as bps==4, it all works very well (and bitsPerPixel, bytesPerRow are both the same as with bps==5, as expected). In the interests of exploration, I decided to check on what the system reports to me when my screen is set to 555/16. So, I've got code somewhere like: NSScreen *screen=[NSScreen mainScreen]; NSWindowDepth depth=[screen depth]; NSLog( @"depth==%d\n", depth); NSLog( @"planar==%d\n", NSPlanarFromDepth( depth)); NSLog( @"NSColorSpaceFromDepth==%@\n", NSColorSpaceFromDepth( depth)); NSLog( @"NSBitsPerSampleFromDepth==%d\n", NSBitsPerSampleFromDepth( depth)); NSLog( @"NSBitsPerPixelFromDepth==%d\n", NSBitsPerPixelFromDepth( depth)); Everything checks out - except that the NSBitsPerSampleFromDepth is being reported as 4! Should 555/16 work? I can use 888/32, it's not a problem, but I had figured that while I was in there, I might as well be complete... Thanks, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: Mike Paquette <mpaque@ncal.verio.com> Newsgroups: comp.sys.next.programmer Subject: Re: NSDrawBitmap() and 5-bps. Date: Fri, 04 Dec 1998 20:49:14 -0800 Organization: Electronics Service Unit No. 16 Message-ID: <3668BB49.7EA4627E@ncal.verio.com> References: <SCOTT.98Dec4170901@slave.doubleu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Scott Hess wrote: > Should 555/16 work? I can use 888/32, it's not a problem, but I had > figured that while I was in there, I might as well be complete... Nope. 555/16 is not a public input format. Even if it was, the WindowServer would have to massage the data to perform software gamma correction for the wimpy 555/16 Intel framebuffers that don't support palettes in anything past 8 bit pseudocolor (almost all of them). This will be a public format on Mac OS X, where we have hardware gamma correction. Mike Paquette
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <18755912315622@digifix.com> Date: 6 Dec 1998 04:46:10 GMT Organization: Digital Fix Development Message-ID: <15188912920426@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1994. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: nurban@crib.corepower.com (Nathan Urban) Newsgroups: comp.sys.next.programmer Subject: Re: learning obj-c Date: 6 Dec 1998 14:05:21 -0500 Organization: Data Systems Consulting, Inc. Message-ID: <74ekhh$mmp$1@crib.corepower.com> References: <366A0620.5E2CAA55@home.com> NNTP-Posting-Date: 6 Dec 1998 19:05:38 GMT In article <366A0620.5E2CAA55@home.com>, Ari <arikounavis@home.com> wrote: > I'm a student and have the opportunity to design my own obj-c self study [...] > I'm planning on using PB/IB with WebObjects 4 on NT until MacOS X server > for intel is released (if ever). The problem is that I've *NEVER* > programmed before, and I can't seem to find any "teach yourself > obj-c/yellowbox" books. Do the developer books come with WebObjects/PB/IB? If not, they're available online as well. Plus you could consult Stepwise for the series of HTMLedit tutorials. > Some have suggested that I learn Java and then > move to obj-c, but all the Java books seem to suggest that C is the > fundamental thing to know Java and Obj-C have a lot of similarities, but if you want to use Obj-C you'll need to learn some C at some point. > The goals of the course would be to write some preposterously simple > original programs (can command line apps be written in obj-c?) Yes, but OpenStep is one of those rare environments in which it's probably easier to make your very first app graphical instead of command-line based. An example of a very simple first-time app, try the temperature converter demo (Fahrenheit<->Celsius). > If you were going to design this course, how would you go about it? Read the Apple documentation: http://developer.apple.com/techpubs/macosxserver/macosxserver.html Start with "Object-Oriented Programming and the Objective-C Language". Read "Discovering OPENSTEP". Look at the Foundation and AppKit APIs, but not too deeply at first (especially since your goal is to learn Obj-C, not OpenStep); better to just jump in and implement some of the demo apps found in tutorials. > and a larger "rapidly developed" app that shows off the power of PB/IB > and all those cool frameworks, like maybe a text editor where I could > add some features. Do the famous "5-minute word processor" demo. :) For an example of how easy this sort of thing is, look at Scott Anguish's "Working with NSDocument - A Practical Primer" and his followup tutorials on HTMLedit, in the Stepwise archives at: http://www.stepwise.com/Articles/Technical/HTMLEdit/ > Do I have unrealistic expectations for a single semester? No. You can jump right into the OOP, and learn the C as you go along. All the complexity in Objective-C is in the C part, not the Objective part, and if you stick to very OOP applications you don't need to know a whole lot of the C, mostly just enough for control structures (if/while/for). Programming isn't as hard as you might think. Just ask questions.
From: Richard <chergr@lure.latrobe.edu.au> Newsgroups: comp.sys.next.programmer Subject: miniSQL porting to Rhapsody Date: Mon, 07 Dec 1998 12:27:18 +1000 Organization: Faraday R&D Message-ID: <366B3D05.561B6780@lure.latrobe.edu.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; x-mac-type="54455854"; x-mac-creator="4D4F5353" Content-Transfer-Encoding: 7bit NNTP-Posting-Date: 7 Dec 1998 08:30:13 GMT I am looking for instructions to compile miniSQL for Rhapsody DR2. The present sticking point is the identity of the library that contains msync. (the header declaration for msync has one less that the corrrect number of parameters as well, but I fixed that) Any clues gratefully recieved.
Message-ID: <366B70FE.CAF7DB9@austin.rr.com> Date: Mon, 07 Dec 1998 00:09:02 -0600 From: Futeh Kao <futeh@austin.rr.com> MIME-Version: 1.0 Subject: WebObject 4.0 & JDK1.2 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Newsgroups: comp.sys.next.programmer Does WebObject 4.0 support JDK1.2. If not, when will the support be available? At work, we have WebObject 3.5.1 and it supports JDK1.1.3. Moreover, I've not seen any problems using JavaVM.framework with JDK1.1.7a and swing-1.0.3. However, it complains bitterly when I tried to use JDK1.2. I know Apple only supports the JDK that comes with WebObject. However, it is critical for us ( I can imagine for other people as well) that Apple supports JDK 1.2 in WebObject. Futeh
From: pete@ohm.york.ac.uk (-bat.) Newsgroups: comp.sys.next.programmer Subject: Re: NSDrawBitmap() and 5-bps. Date: 8 Dec 1998 12:09:45 GMT Organization: The University of York, UK Sender: pcf1@york.ac.uk Distribution: world Message-ID: <74j4u9$d43$1@pump1.york.ac.uk> References: <SCOTT.98Dec4170901@slave.doubleu.com> NNTP-Posting-Date: 8 Dec 1998 12:09:45 GMT scott@nospam.doubleu.com (Scott Hess) writes: > In the interests of being fast on different screens, I've been > experimenting with 555/16 mode. Since it's easy for me to modify > whether the code's spitting out 444/16 or 888/32, I figured I'd add > 555/16 _if_ that's what the screen is at. Hmmm, I found that using 444/16 or 222/8 always comes out fast, even on higher depth screens. 222/8 is a superb colour format if you are trying to squeeze down a picture for storage on disc by the way... -bat.
From: eric@skatter.USask.Ca Newsgroups: comp.sys.next.programmer Subject: egcs-1.1.1 Date: 8 Dec 1998 15:34:34 GMT Organization: University of Saskatchewan Message-ID: <74jgua$nls$1@tribune.usask.ca> Has anyone had any luck getting egcs-1.1.1 to build on an OPENSTEP 4.2/Intel machine? My attempts to build the compiler end with a core dump from the gencheck program built by the the stage 1 compiler. I'm using gcc-2.8.1 to build the egcs-1.1.1 stage1 compiler. ============================================================== rm -f cpp ln cccp cpp stage1/xgcc -Bstage1/ -c -DIN_GCC -O2 -g -O2 -DHAVE_CONFIG_H -I. -I../. ./egcs-1.1.1/gcc -I../../egcs-1.1.1/gcc/config ../../egcs-1.1.1/gcc/../../egcs-1 .1.1/gcc/gencheck.c stage1/xgcc -Bstage1/ -DIN_GCC -O2 -g -O2 -DHAVE_CONFIG_H -o gencheck \ gencheck.o ` case "obstack.o" in ?*) echo obstack.o ;; esac ` ` case "" in ?*) echo ;; esac ` ` case "" in ?*) echo ;; esac ` ` case "" in ?*) echo ;; esac ` ` case "" in ?*) echo ;; esac ` ./gencheck > tmp-check.h /bin/sh: 3993 Bus error gmake[2]: *** [s-check] Error 138 gmake[1]: *** [bootstrap] Error 2 gmake: *** [bootstrap] Error 2 ================================================================ I checked with gdb and found that gencheck was dying in the code in /lib/crt0.o. I tried rebuilding gencheck with /lib/crt1.o instead. The program then gets through the crt code but dies later with a segmentation violation. :-( I presume that something has gotten messed up during the link-edit, but I have absolutely no idea where to start. Any suggestions? -- Eric Norum eric@skatter.usask.ca Saskatchewan Accelerator Laboratory Phone: (306) 966-6308 University of Saskatchewan FAX: (306) 966-6058 Saskatoon, Canada.
From: "Stefan Lindgren" <s.lindgren@telia.com> Newsgroups: comp.sys.next.programmer Subject: Pascal compiler? Date: Tue, 8 Dec 1998 18:53:39 +0100 Organization: Telia Internet Services Message-ID: <366d665b.0@d2o49.telia.com> Does anybody know about a Pascal compiler for OPENSTEP? Regards
From: "Rex Dieter" <rdieter@math.unl.edu> Newsgroups: comp.sys.next.programmer Subject: Re: egcs-1.1.1 Date: Tue, 8 Dec 1998 18:23:49 -0600 Organization: University of Nebraska-Lincoln Message-ID: <74kfil$643$1@unlnews.unl.edu> References: <74jgua$nls$1@tribune.usask.ca> eric@skatter.USask.Ca wrote in message <74jgua$nls$1@tribune.usask.ca>... >Has anyone had any luck getting egcs-1.1.1 to build on an OPENSTEP 4.2/Intel >machine? My attempts to build the compiler end with a core dump from the >gencheck program built by the the stage 1 compiler. I'm using gcc-2.8.1 to >build the egcs-1.1.1 stage1 compiler. >I checked with gdb and found that gencheck was dying in the code in >/lib/crt0.o. I tried rebuilding gencheck with /lib/crt1.o instead. The >program then gets through the crt code but dies later with a segmentation >violation. :-( > >I presume that something has gotten messed up during the link-edit, but I >have absolutely no idea where to start I had the same problem, it's because egcs is attempting to link dynamically, and MUST be configured to link statically (in the same/similar manner I hacked gcc-2.8.1). You need to munge the specs file that gets created to do this for you (see the specs file that came with gcc-2.8.1 on where to add the appropriate static flags). You'll also need the file libsys_s.a available for linking against (which is also included with gcc 2.8.1). But, don't bother. )-; I ended up just installing the stage1 compiler, and it produces bad code. Anything other than a trivial "Hello world" program seems to bomb out. I hope that you may have a different result -- Rex Dieter Computer System Manager UNL Math/Stat
From: theisen@akaMail.com (Dirk Theisen) Newsgroups: comp.sys.next.programmer Subject: Re: WebObject 4.0 & JDK1.2 Date: Wed, 9 Dec 1998 12:35:06 +0100 Organization: University of Bonn, Germany Message-ID: <1djr5a3.1wefhwq12gacg0N@ascend-tk-p170.rhrz.uni-bonn.de> References: <366B70FE.CAF7DB9@austin.rr.com> Futeh Kao <futeh@austin.rr.com> wrote: > I know Apple only supports the > JDK that comes with WebObject. However, it is critical for us > ( I can imagine for other people as well) that Apple supports > JDK 1.2 in WebObject. Why? Dirk -- No RISC - No fun http://theisen.home.pages.de/
From: Jo.Hagelberg@t-online.de (Jo Hagelberg) Newsgroups: comp.sys.next.programmer,comp.sys.next.misc,comp.sys.next.software Subject: Help: NS3.3 to OS4.2/NT port? Date: 9 Dec 1998 08:19:11 GMT Organization: Dipl.-Phys. Jo Hagelberg Systemtechnik Message-ID: <74lbpv$pgh$1@news00.btx.dtag.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit hello, I want to port a NextSTep 3.3 application to OpenStep NT 4.2? Do I need Mach 4.2 too? I didn't find any plain answer to this in the newsgroups. I just learned that there are some conversion scripts available, however, not on the OS/NT package. So the OS/Mach package (which I don't have yet) seems to be needed for the conversion. I'm not familiar with neither system, just have some OO experience and a colleague who has done the NS3.3 stuff. I'd appreciate any help you can give to me. best regards Jo Hagelberg
From: DavidShaffer@psu.edu Newsgroups: comp.sys.next.programmer Subject: BPF lks source code? Date: Wed, 09 Dec 1998 14:57:35 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <74m34u$57g$1@nnrp1.dejanews.com> Hello, I seem to recall that a BPF LKS was packaged with PPP but I'm having trouble tracking down source code for it. Does anyone know were I could find such a thing? If not, then, how about a binary version for OPENSTEP 4.2? Thanks! David -----------== Posted via Deja News, The Discussion Network ==---------- http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
Message-ID: <366E8F03.51431732@invensys.com> From: futeh <fkao@invensys.com> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Re: WebObject 4.0 & JDK1.2 References: <366B70FE.CAF7DB9@austin.rr.com> <1djr5a3.1wefhwq12gacg0N@ascend-tk-p170.rhrz.uni-bonn.de> Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Date: Wed, 09 Dec 1998 08:53:55 -0600 NNTP-Posting-Date: Wed, 09 Dec 1998 09:59:49 CDT Because other in-house non OpenStep projects will move toward JDK1.2 and it will be difficult to support and to deploy two JREs. Notice I said it is critical for US, instead of Apple. It's a simple question and I've no intention of starting any debates on Java vs. whatever. Futeh Dirk Theisen wrote: > Futeh Kao <futeh@austin.rr.com> wrote: > > I know Apple only supports the > > JDK that comes with WebObject. However, it is critical for us > > ( I can imagine for other people as well) that Apple supports > > JDK 1.2 in WebObject. > > Why? > > Dirk > > -- > No RISC - No fun > > http://theisen.home.pages.de/
From: no.spam.please@no.spam.period Newsgroups: comp.sys.next.programmer Subject: Re: BPF lks source code? MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit References: <74m34u$57g$1@nnrp1.dejanews.com> Message-ID: <cHBb2.442$tI2.549@news.rdc1.tx.home.com> Date: Wed, 09 Dec 1998 21:10:32 GMT NNTP-Posting-Date: Wed, 09 Dec 1998 13:10:32 PDT Organization: @Home Network The dhcp package on the peak openstep archive (ftp://ftp.next.peak.org/pub/next/new_arrivals_openstep) will install bpf, related /dev files, and modify /etc/rc.local to load bpf and dhclient on boot. You can prevent the dhclient load during boot by commenting out the lines for DHCP client in /etc/rc.local If you don't want to keep the dhcp files on the system, use Installer.app to List the files installed and manually remove them. The package also installs /usr/local/etc/bpf_load and /usr/local/etc/bpf_unload, so you could comment out the bpf load in /etc/rc.local as well, and manually load/unload bpf. These two files are root-executable, so you'd have to make them group executable to allow members of the wheel group to execute them (or su to root in Terminal.app). JP -- Please respond to (slight mods needed to address): jpmeia (@ ) ix.netcom.com NeXTMail/MIME welcome
From: "Yan Laporte" <ylaporte@acm.org> Newsgroups: comp.sys.next.programmer Subject: I need Internship! Date: Wed, 09 Dec 1998 23:25:38 -0500 Organization: Universit=?ISO-8859-1?B?6Q==?= de Moncton Message-ID: <74nevl$i0e@sol.sun.csd.unb.ca> Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit Hi, I am a third year computer science student in University of Moncton New-Brunswick in Canada (I live in Montreal) and I need a nice place for my internship. I really like NeXTSTEP and all it's descendants so I was hoping to find something where I could work with OPENSTEP or WebObject. I have a little experience in Objective-C and with NeXTSTEP 2.2. I have more experience with WindowsNT, Solaris Linux and the MacOS. I also read all the manuals for OPENSTEP 4.2 at least two times. I am willing to travel (in fact I would really like to). Please contact me if you have anything that might interest me. Thank you Yan Laporte ylaporte@acm.org http://eve.info.umoncton.ca:8080/~yanl/
From: tlegvold@c2i.net (Thor Legvold) Newsgroups: comp.sys.next.programmer Subject: "DISK BOOT FAILURE, INSTERT SYSTEM DISK AND PRESS ENTER" Date: Thu, 10 Dec 1998 16:38:28 GMT Organization: Tele2 Norway Public Access Message-ID: <366ee3e9.3136587@news1.c2i.net> NS 3.3 on first IDE disk, no DOS or partitioning info (i.e. a NS-only disk). Worked great for years. FreeBSD/Win98 on first SCSI disk. I switch OS's in the mainboard BIOS, choosing the boot order. The SCSI disk has a bootloader, IDE goes right into NS. After upgrading some hardware NS won't boot. (New mainboard, CPU and RAM). AOpen AX6BC (BX chipset), Celeron 300A. I get the message in the Subject line. I booted the hard disk from my boot floppy, replaced the boot sector by hand several times, and it still doesn't boot. What's wrong? I used: disk -B /usr/standalone/i386/boot1 /dev/rhd0a also tried disk -B0 (the "0" option doesn't show up in the man page), disk -b (for the default bootloader), and tried with both boot0 and boot1 and boot. AFAIK, since I have a NS only disk with no DOS info or partitions, boot should go right to boot1 and then boot2 in order to start the system. What have I forgotten? Please reply to me directly via email. Regards, Thor
From: tlegvold@c2i.net (Thor Legvold) Newsgroups: comp.sys.next.programmer,comp.sys.next.sysadmin Subject: How does NS 3.3 access the floppy? (Intel) Date: Thu, 10 Dec 1998 16:38:28 GMT Organization: Tele2 Norway Public Access Message-ID: <366ee1e5.2620272@news1.c2i.net> I'm asking because I'm trying to debug some problems after upgrading my hardware. I was on an Abit PX5 (TX) mainboard, AMD K6/233 with 64MB SDRAM. I upgraded to a AOpen AX6BC (BX) mainboard, Celeron 300A w/64MB SDRAM, and an ATX tower (I had AT before that). I run NS 3.3, Win98 and FreeBSD 3.0 After upgrading I experienced many strange problems, and often the board wouldn't even boot. I have gotten everything installed and (seemingly) stable now, with one exception. Accessing the floppy for reading or writing lock Windows up very very hard, requiring a hard reset. FreeBSD either panics the kernel upon mounting or shows many many errors on read/write operations. DOS works without any problem, reading writing and formatting. I fired up good old NextStep and tried out a few MSDOS floppies - everything seemed to work perfectly, and without a single error on the console. I could even format. I have been suspecting the mainboard as the cause of the problem (I've tried several floppy units, several cables, different BIOS settings, swapped RAM & CPU, etc), but the service centre say's it's ok (I quote "we managed to install Windows on it, so there's nothing wrong here" - well, slightly exaggerated :-) I'm wondering if maybe the way in which Win/BSD access the floppy (i.e. the driver) can be provoking some bug/problem in the BIOS or mainboard that DOS and NS don't seem to affect, and therefore thought I'd ask if there's anything special about the way NS accesses a floppy. Please email replies, Thor
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: keyview loops don't work + example code: why!? Date: 10 Dec 1998 17:13:03 GMT Organization: Department of Electronics, University of York, UK. Sender: rp9@york.ac.uk Message-ID: <slrn770094.770.rog@talisker.ohm.york.ac.uk> NNTP-Posting-Date: 10 Dec 1998 17:13:03 GMT i am failing dismally to get keyview loops to work properly. i explicitly set the keyview ordering and the keyview loop runs backwards! i'm completely at a loss and would be very glad if anyone could confirm the problem or tell me where i'm going wrong... i've tested this program under openstep 4.2 for intel and webobjects under windows NT. first extract the source below into a file keyviewtst.m to compile under openstep 4.2: cc keyviewtst.m -framework AppKit to compile under webobjects/yellowbox under windows NT (under the Bourne shell): gcc keyviewtst.m -o keyviewtst.exe -win -framework Foundation \ -framework AppKit $Lib\\libNSWinMain.a when you run it, try tabbing between the fields; it *should* go through the fields from top to bottom, but for some strange reason it goes backwards. thanks for any help, rog. // **** snip here, start of keyviewtst.m #import <AppKit/AppKit.h> NSWindow *new_window(NSSize sz); @interface FlippedView: NSView - (BOOL)isFlipped; @end // This is the template for the view hierarchy to be created; // the initial letters correspond to the names of the variables // in the function create_window(). // a NSView {{0, 0}, {173, 61}} // { // b NSTextField {{0, 0}, {151, 21}} 1st in keyview loop // c NSPopUpButton {{0, 21}, {94, 20}} 2nd in keyview loop // d NSButton {{0, 41}, {173, 20}} 3rd in keyview loop // } void create_window(void) { id a, b, c, d; NSWindow *win; /* * create all the views in the view hierarchy */ // a NSView frame [(1, 9), (173, 61)] (flipped) // the content view of the window a = [[[FlippedView alloc] initWithFrame:(NSRect){{1,9}, {173,61}}] autorelease]; // b NSTextField {{0, 0}, {151, 21}} // an editable text field b = [[[NSTextField alloc] initWithFrame:(NSRect){{0,0}, {151, 21}}] autorelease]; [b setEditable:YES]; // c NSPopUpButton {{0, 21}, {94, 20}} // a popup button (with only one item) c = [[[NSPopUpButton alloc] initWithFrame:(NSRect){{0,21}, {94,20}}] autorelease]; [c setTitle:@"Popup list"]; // d NSButton {{0, 41}, {173, 20}} // a tickbox d = [[[NSButton alloc] initWithFrame:(NSRect){{0, 41}, {173, 20}}] autorelease]; [d setButtonType:NSToggleButton]; [d setImage:[NSImage imageNamed:@"NSSwitch"]]; [d setAlternateImage:[NSImage imageNamed:@"NSHighlightedSwitch"]]; [d setTitle:@"Always do something"]; [d setImagePosition:NSImageLeft]; /* * actually place the created views into their correct * place in the hierarchy */ win = new_window((NSSize){173, 61}); [win setTitle:@"Keyview loop testing"]; [win setContentView:a]; [a addSubview:b]; // add text field [a addSubview:c]; // add popup button [a addSubview:d]; // add tickbox /* * attempt to set up the key view loop correctly. * NB THIS DOESN'T WORK! - the keyview loop is set up backwards. */ [b setNextKeyView:c]; // textfield to popup button [c setNextKeyView:d]; // popup button to tickbox [d setNextKeyView:b]; // tickbox back to textfield [win makeKeyAndOrderFront:nil]; } int main(int argc, char **argv) { NSAutoreleasePool *pool; id menu, item; pool = [[NSAutoreleasePool alloc] init]; [NSApplication sharedApplication]; menu = [[NSMenu alloc] initWithTitle:@"Keyview Test"]; item = [menu addItemWithTitle:@"Quit" action:@selector(stop:) keyEquivalent:@"q"]; [item setTarget:NSApp]; [NSApp setMainMenu:menu]; create_window(); [NSApp run]; [pool release]; return 0; } #define STYLE (NSTitledWindowMask \ | NSClosableWindowMask \ | NSMiniaturizableWindowMask \ | NSResizableWindowMask) NSSize screensize(NSWindow *win) { NSScreen *screen = [win screen]; if (screen == nil) screen = [NSScreen mainScreen]; return [screen frame].size; } NSWindow *new_window(NSSize sz) { NSRect cr; NSSize scrsz; NSWindow *win; scrsz = screensize(nil); cr = (NSRect){{(scrsz.width-sz.width)/2, (scrsz.height-sz.height)/2}, sz}; win = [[NSWindow alloc] initWithContentRect:cr styleMask:STYLE backing:NSBackingStoreBuffered defer:NO]; return win; } @implementation FlippedView: NSView - (BOOL)isFlipped { return YES; } @end // *** end of keyviewtst.m
From: "Mark Bessey" <mbessey@apple.com> Newsgroups: comp.sys.next.programmer Subject: Re: keyview loops don't work + example code: why!? Date: Thu, 10 Dec 1998 12:01:25 -0800 Organization: Apple Computer, Inc. Message-ID: <74p9e0$dtk$1@news.apple.com> References: <slrn770094.770.rog@talisker.ohm.york.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit You need to set the window's initialFirstResponder to one of the views. If you don't do that, the AppKit completely ignores the nextKeyView stuff, and selects keyviews in an arbitrary order. Just add a line like the following: [win setInitialFirstResponder: b]; to the code where you're setting the nextKeyViews. Hope this helps, -Mark ---------- In article <slrn770094.770.rog@talisker.ohm.york.ac.uk>, rog@ohm.york.ac.uk (Roger Peppe) wrote: > i am failing dismally to get keyview loops to work properly. > i explicitly set the keyview ordering and the keyview loop runs > backwards! i'm completely at a loss and would be very glad if anyone > could confirm the problem or tell me where i'm going wrong...
From: Bill <Bill@software.com> Newsgroups: comp.sys.next.programmer,comp.sys.psion.programmer,comp.unix.programmer,comp.unix.sco.programmer,comp.windows.ms.programmer,de.comp.os.ms-windows.programmer,de.comp.os.os2.programmer Subject: Internet Explorer 5 news! Date: Thu, 10 Dec 1998 22:02:33 +0100 Organization: Software System Message-ID: <367036E8.D1DC2C51@software.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Internet Explorer 5 news! Important news about Internet Explorer 5: http://www.dds.nl/~brk
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer,comp.sys.psion.programmer,comp.unix.programmer,comp.unix.sco.programmer,comp.windows.ms.programmer,de.comp.os.ms-windows.programmer,de.comp.os.os2.programmer Subject: cmsg cancel <367036E8.D1DC2C51@software.com> Control: cancel <367036E8.D1DC2C51@software.com> Date: 10 Dec 1998 21:09:38 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.367036E8.D1DC2C51@software.com> Sender: Bill <Bill@software.com> Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: ecarter@eric.acs.uci.edu (Eric Carter) Newsgroups: comp.sys.next.programmer Subject: Considering WebObjects Date: 11 Dec 1998 06:54:07 GMT Organization: University of California, Irvine Message-ID: <74qfif$iit@news.service.uci.edu> Hello, I've just returned from a seminar presented by Apple on the new WebObjects 4. We currently serve all our courses via the web using Apache/Perl solutions. We are considering taking the step towards a more rapid development environment since staff resources are low. Can anyone provide some personal experience with WebObjects and suggest whether is would be a good solution for accomplishing our goal? Thanks, Eric Carter
From: lcwpod@aol.com Newsgroups: comp.sys.next.programmer Subject: 100% FREE XXX SITE 9364 Date: 11 Dec 1998 11:43:59 GMT Message-ID: <3671057f.0@194.65.24.2> The sex gallery ! It's all FREE here ! FREE xxx memberships, passwords,pic,thumbnails,and much more ! Now featuring live hott erotic talk ! Cum by today ! WE'RE WAITING ! GO TO http://www.adultisp.com/sexgallery/index.html sbixkvslrlqumxexrwppstgiycpjjirdsusotxz
Newsgroups: comp.sys.net-computer.misc,comp.sys.newton.marketplace,comp.sys.newton.misc,comp.sys.newton.programmer,comp.sys.next.advocacy,comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer Subject: cmsg cancel <yG1c2.1017$Vw4.83@nsw.nnrp.telstra.net> ignore no reply Control: cancel <yG1c2.1017$Vw4.83@nsw.nnrp.telstra.net> Message-ID: <cancel.yG1c2.1017$Vw4.83@nsw.nnrp.telstra.net> Date: Fri, 11 Dec 1998 05:02:36 +0000 Sender: streamgold@yaybtwbj.ca From: andrew@erlenstar.demon.co.uk Organization: Annihilator v0.3 Spam (EMP) cancelled - multiposted binary files BI=11972.326/15 SPAM ID=AeaMKZSLPQNO8hqeLm5PvQ==
From: tgia@saltmine.radix.net (Anthony Giaccone) Newsgroups: comp.sys.next.programmer Subject: WebObjects and EnterpriseObjects Date: 11 Dec 1998 12:07:57 -0500 Organization: RadixNet Message-ID: <74rjhd$k8c$1@saltmine.radix.net> Ok folks, as some of you may realize from the last message. I'm trying to evaluate WebObjects. I've gotten pretty far, but reached an impasse which I'm hoping you all can help me with. I created a data model, using the eomodler tool. Originally I accepted the default class for the entities EOGenericRecord. Now I want to go back and change those to class specific objects. So I opened up the eomodler from the project builder and changed the classes to the new types I wanted to use. Here's an example Entity Table Class User UserTable EOGenericRecord (before) User UserTable EOUser (after my object) Now in my code for a particular page I have the following code. protected EOUser aUser protected EOUser ourUser protected WODisplayGroup usersDisplayGroup; protected NSMutableArray selectedUsers; public void selectUser() { if (selectedUsers.count() > 0) { ourUser = (EOUser) selectedUsers.objectAtIndex(0); // Bomb! ((Session)this.session).setUser(ourUser); } } In the browser the following attributes are set: attribute Binding -------------------------------------- displayString aUser.name item aUser list userDisplayGroup.displayedObjects selections selectedUsers This code bombs on the line noted. And it bombs because the return value from the selectedUsers is of type EOGenericRecord. Now the questions 1) Will the WODisplayGroup ever but any kind of object other then EOGenericRecord in it's selection Array? 2) If it will how do I get it to return my Objects, instead of EOGenericUser 3) If it doesn't, how can I use the EOGenericRecord that is returned to retrieve the EOUser object from the DB? I've read through the manual for the last few days, and either I don't understand the terminology and object relationship in EO (Strike that I clearly don't understand it very well) and I can't recognize the sections I should be reading, or it's doesn't seem to be ther. Can anyone shed some light on this? Thanks Tony Giaccone
From: tw-TakeThisOut-@hrl.com (Williamson, Weldon S) Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.hardware Subject: Serial-Port Problem Under NS 3.3 Date: Fri, 11 Dec 1998 15:05:27 -0700 Organization: HRL Laboratories Message-ID: <tw-TakeThisOut--1112981505270001@matod.hrl.hac.com> I am running NeXTStep 3.3 on an Intel motherboard, using the on-board COM1 and COM2 ports and the standard-release On-Board Serial Port driver, a single driver that responds to IRQs 3 and 4 and handles both ports. I need to operate both ports at 57,600 baud, communicating with a DOS machine a few feet away. This was all working fine on an old computer (200 MHz Pentium-S), but on a new machine (300-MHz Pentium II on Intel PD440FX motherboard), the OS crashes (even alt-numlock gets no response) within a few minutes or seconds of launching the code that accesses the ports (same code and driver as in the old Pentium-S machine). Running at a lower baud rate causes it to take longer to crash. I have tried using the Mux driver (v. 1.9), which I downloaded from Peak. I used a separate instance of the driver for COM1 and COM2, setting the IRQs, port addresses and instance numbers appropriately (instance 0: 0x3F8-0x3Ff: IRQ4; instance 1: 0x2F8-0x2FF, IRQ3). COM1 works fine with this driver, but when the code tries to do an ioctl(TIOCSETD) for COM2, it gets an error from Mach, and the port fails to communicate. If I try to drive _only_ COM2 with a single instance of the Mux driver, I get the same error. I tried using an ISA serial board addressed as COM1 and COM2, after disabling the onboard ports. With either the ISA serial port driver or the Mux driver, the behavior was the same as with the onboard ports. Can anyone help with this vexing problem?! TIA, Tod Williamson tw@hrl.com _____________________________________________________ - Tod Williamson Please take out the anti-spam phrase in the reply-to address before emailing. Thanks.
From: "Jin Kim" <kwazyjin@hotmail.com> Newsgroups: comp.sys.next.programmer Subject: where is libc.a in OpenStep4.2? Message-ID: <92lc2.1118$4w2.4950930@news.itd.umich.edu> Date: Fri, 11 Dec 1998 22:00:10 -0500 NNTP-Posting-Date: Fri, 11 Dec 1998 22:03:33 EDT Hello, I need libc.a to compile some source code that I've downloaded. I can't seem to find it anywhere in the OpenStep4.2 install. Do I need to get it from elsewhere, or is it in the regular install and I'm just not being able to find it? I did however find libcc.a. Wondering if I could just substitute without problem. Thanks! - Jin
From: echo4@NOSPAM.erols.com --delete NOSPAM Newsgroups: comp.sys.next.programmer Subject: some help with posix? Date: 12 Dec 1998 04:31:23 GMT Message-ID: <74srir$2h8$1@winter.news.rcn.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NNTP-Posting-Date: 12 Dec 1998 04:31:23 GMT Hey guys, I have OS4.2 running over here. I wanted to compile some code, but the broken, or the lack of posix in 4.2 is killing me. I managed to hack around alot of it, but I don't know how to handle the signal stuff. If i could get my hands the libposix.a library in 3.3, it may solve some of my problems (sigaction for sure). Any ideas would be apreciated. I am assuming that I could also use 3.3's libposix.a on 4.2? If anybody is willing to donate just libposix.a or the needed objects, I would appreciate it. Here are the undefined symbols. /bin/ld: Undefined symbols: _setpgid _waitpid _sigaction _sigaddset _sigemptyset -- |========================================| J. Lee echo4@NOSPAM.erols.com --delete NOSPAM NeXTMail,SUN,MIME |========================================|
From: luomat@peak.org.obvious.portion (TjL) Newsgroups: comp.sys.next.programmer Subject: Re: where is libc.a in OpenStep4.2? Organization: would be nice MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit References: <92lc2.1118$4w2.4950930@news.itd.umich.edu> Message-ID: <pNmc2.424$vg4.660743@newshog.newsread.com> Date: Sat, 12 Dec 1998 05:02:13 GMT NNTP-Posting-Date: Sat, 12 Dec 1998 00:02:13 EDT In <92lc2.1118$4w2.4950930@news.itd.umich.edu> "Jin Kim" wrote: > I need libc.a to compile some source code that I've downloaded. I can't > seem to find it anywhere in the OpenStep4.2 install. Do I need to get it > from elsewhere, or is it in the regular install and I'm just not being able > to find it? FWI I don't seem to have it either for 3.3 or 4.2 > I did however find libcc.a. Wondering if I could just > substitute without problem. Thanks! No idea TjL -- Spam-altered address in effect, remove obvious portion if replying by email.
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <12129803.2653@none444.yet> Control: cancel <12129803.2653@none444.yet> Date: 12 Dec 1998 09:26:49 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.12129803.2653@none444.yet> Sender: no.email.address.entered@none444.yet Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: luomat@peak.org.obvious.portion (TjL) Newsgroups: comp.sys.next.programmer Subject: Re: some help with posix? Organization: would be nice MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit References: <74srir$2h8$1@winter.news.rcn.net> Message-ID: <Xcwc2.522$vg4.962553@newshog.newsread.com> Date: Sat, 12 Dec 1998 15:45:59 GMT NNTP-Posting-Date: Sat, 12 Dec 1998 10:45:59 EDT In <74srir$2h8$1@winter.news.rcn.net> echo4@NOSPAM.erols.com --delete NOSPAM wrote: > If anybody is willing to donate just libposix.a or the needed objects don't know if this will help but ftp://enterprise.apple.com/pub/Patches/NEXTSTEP3.3Patch1/libposix.a TjL -- Spam-altered address in effect, remove obvious portion if replying by email.
From: billg@microsoft.com (MevDev) Newsgroups: comp.sys.next.programmer Subject: Come to the new NeXT Forever site Date: Sat, 12 Dec 1998 11:00:39 -0600 Organization: [poster's organization not specified] Message-ID: <billg-1212981100400001@dial-151.meltel.com> I set up a NeXT forever site for everyone to benefit from. Please visit it at: http://www.melrose.k12.mn.us/geek/index.html If you have additions, subtractions, articles, or anything of use please email me at : peecee@hempseed.com, or visit my site, there is a link at the bottom of the page. Thanks.
From: asmang@CRAZYSPAMmail.wm.edu (Arun Mangalam) Newsgroups: comp.sys.next.programmer Subject: [Q] WebObjects 4.0?!?! Date: Sat, 12 Dec 1998 13:52:08 -0500 Organization: William & Mary Message-ID: <asmang-1212981352080001@192.168.0.22> NNTP-Posting-Date: 12 Dec 1998 18:46:26 GMT Has anyone here used WebObjects 4.0? If you have or know someone who has, can you please post your and/or their experience with it? I'm very interested. Especially those in the education field..., I would like to know what you think about WebObjects. Thanks for any information... :) -- asmang@CRAZYSPAMmail.wm.edu remove the CRAZYSPAM to reply. These spamming buggers are really annoying.
From: echo4@NOSPAM.erols.com --delete NOSPAM Newsgroups: comp.sys.next.programmer Subject: Is there a Otool.app for Intel? removing objects Date: 12 Dec 1998 23:09:31 GMT Message-ID: <74ut3b$6i6$1@winter.news.rcn.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NNTP-Posting-Date: 12 Dec 1998 23:09:31 GMT How you do you remove an object from a library? For example: Removal of sigaction.o from libposix.a I know there was a m68k tool called Otool.app that allowed one to do this, but is there a similiar intel tool? I am pretty sure, you could use the segedit via otool and strip function, but I can figure out how to use the shell tools. Any additional information would be greatly appreciated. Jay -- |========================================| J. Lee echo4@NOSPAM.erols.com --delete NOSPAM NeXTMail,SUN,MIME |========================================|
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <15188912920426@digifix.com> Date: 13 Dec 1998 04:45:59 GMT Organization: Digital Fix Development Message-ID: <13296913525221@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1994. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: "Stefan Lindgren" <s.lindgren@telia.com> Newsgroups: gnu.misc,comp.lang.pascal.misc,gnu.gcc,comp.sys.next.programmer Subject: Compiling GPC on OPENSTEP 4.X Date: Sun, 13 Dec 1998 12:42:22 +0100 Organization: Telia Internet Services Message-ID: <3673a6cf.0@d2o49.telia.com> Hello GNU'ers. I need to get gpc working on OPENSTEP 4.X(Motorola, Black HW). I'm using the proper version of GCC(2.1.7.2.1) as the "backend" for gpc. No problems building GCC. When I build the GPC this error occures: gpc-cccp.c:1240: illegal function definition, found `__attribute__' *** Exit 1 Thanks in advance Stefan s.lindgren@telia.com
From: echo4@NOSPAM.erols.com --delete NOSPAM Newsgroups: comp.sys.next.programmer Subject: libdl.a? dynamic loading..? Date: 13 Dec 1998 17:50:15 GMT Message-ID: <750uon$ht9$2@winter.news.rcn.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NNTP-Posting-Date: 13 Dec 1998 17:50:15 GMT any hacks around dlclose, dlopen, dlsym, dlerror, on OS4.2 Any information would be appreciated. Jay Lee -- |========================================| J. Lee echo4@NOSPAM.erols.com --delete NOSPAM NeXTMail,SUN,MIME |========================================|
Subject: Finding bezier control points ? Newsgroups: comp.sys.next.programmer From: email@end.of.post (Raymond Lutz) Message-ID: <BkXc2.27$Ib4.221@wagner.videotron.net> Date: Sun, 13 Dec 1998 22:37:21 GMT NNTP-Posting-Date: Sun, 13 Dec 1998 17:37:21 EDT Ok, I know this isn't the most relevant group but I did try in comp.graphic.algorithms without any succes... Being about beziers, maybe someone here could help me (they're also called spline and are the parametric curve drawn by the curveto PostScript operator). curveto -> PostScript -> DisplayPostScript -> OPENSTEP -> csn.programmer 8^) My goal: To join N points (sampling a 2D curve) by N-1 beziers for PDF output. At each point, first and second derivative (y' and y") of the sampled 2D curve are known (it's a differential equation solution) . This is ridiculous, but I spent *days* trying to figure how the second derivative y" at a bezier end sets its handle coordinates. All I found is a quartic polynomial... there must be a closed solution! I looked in Foley, van Dam, et al., "Computer Graphics principles and practice" and didn't find anything on this specific problem. Note that the inverse problem is trivial, ie. finding y' and y" knowing x0,y1 .. x3, y3 . Thanks! And Happy winter solstice! Ray -- Raymond Lutz - 9bit.qc.ca@lutzray - www.9bit.qc.ca/~$myusername - "Les 400 plus fortunes individus de la planete possedent autant que 2.3 MILLIARDS des plus pauvres reunis"
From: Denise Howard <denisehAT@idiomDOT.com> Newsgroups: comp.sys.next.programmer Subject: Re: Finding bezier control points ? Date: 14 Dec 1998 08:27:13 GMT Organization: Remove capital letters to reply Message-ID: <752i51$kss$1@news.idiom.com> References: <BkXc2.27$Ib4.221@wagner.videotron.net> User-Agent: tin/pre-1.4-980117 (UNIX) (FreeBSD/2.2.6-STABLE (i386)) Raymond Lutz <email@end.of.post> wrote: > My goal: > > To join N points (sampling a 2D curve) by N-1 beziers for PDF output. > At each point, first and second derivative (y' and y") of the sampled > 2D curve are known (it's a differential equation solution) . > > I looked in Foley, van Dam, et al., "Computer Graphics principles and > practice" and didn't find anything on this specific problem. That's too general. You need a book on computational geometry or geometric modeling. Try "Geometric Modeling" by Michael E. Mortenson, ISBN 0-471-88279-8. I don't know for sure that your solution is in there, but it discusses these sort of things in some detail. Denise
From: echo4@NOSPAM.erols.com --delete NOSPAM Newsgroups: comp.sys.next.programmer Subject: Posix problem solved...thanks. Date: 13 Dec 1998 02:33:32 GMT Message-ID: <74v91s$et6$1@winter.news.rcn.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NNTP-Posting-Date: 13 Dec 1998 02:33:32 GMT Used a combination of ar, otools, segedit to extract different objects from libposix.a . Otool.app was not needed. Thanks.. -- |========================================| J. Lee echo4@NOSPAM.erols.com --delete NOSPAM NeXTMail,SUN,MIME |========================================|
From: rdieter@math.unl.edu (Rex Dieter) Newsgroups: comp.sys.next.programmer Subject: Re: where is libc.a in OpenStep4.2? Date: 14 Dec 1998 14:52:56 GMT Organization: University of Nebraska-Lincoln Message-ID: <7538o8$5tf$1@unlnews.unl.edu> References: <92lc2.1118$4w2.4950930@news.itd.umich.edu> In article <92lc2.1118$4w2.4950930@news.itd.umich.edu> "Jin Kim" <kwazyjin@hotmail.com> writes: > Hello, > > I need libc.a to compile some source code that I've downloaded. I can't > seem to find it anywhere in the OpenStep4.2 install. Do I need to get it > from elsewhere, or is it in the regular install and I'm just not being able > to find it? I did however find libcc.a. Wondering if I could just > substitute without problem. Thanks! There is no libc.a on NEXTSTEP/OpenStep. The standard C library on openstep it is /lib/libsys_s.dylib aka /NextLibrary/Frameworks/System.framework/System FYI, You need to install OpenStep 4.2 Developer (a separate package from OpenStep 4.2 User) to be able to compile anything -- Rex A. Dieter rdieter@math.unl.edu Computer System Manager http://www.math.unl.edu/~rdieter/ Mathematics and Statistics University of Nebraska-Lincoln
Newsgroups: comp.sys.next.programmer From: mark@ns1.oaai.com (Mark Onyschuk) Subject: Re: Finding bezier control points ? Message-ID: <slrn77ad8s.lge.mark@ns1.oaai.com> Sender: news@T-FCN.Net Organization: M. Onyschuk and Associates Inc. References: <BkXc2.27$Ib4.221@wagner.videotron.net> <752i51$kss$1@news.idiom.com> Date: Mon, 14 Dec 1998 15:54:39 GMT In article <752i51$kss$1@news.idiom.com>, Denise Howard wrote: >Raymond Lutz <email@end.of.post> wrote: >> My goal: >> >> To join N points (sampling a 2D curve) by N-1 beziers for PDF output. >> At each point, first and second derivative (y' and y") of the sampled >> 2D curve are known (it's a differential equation solution) . >> >> I looked in Foley, van Dam, et al., "Computer Graphics principles and >> practice" and didn't find anything on this specific problem. > >That's too general. You need a book on computational geometry or >geometric modeling. Try "Geometric Modeling" by Michael E. Mortenson, >ISBN 0-471-88279-8. I don't know for sure that your solution is in there, >but it discusses these sort of things in some detail. > >Denise > You can find an iterative solution to the problem of interpolating digitized curves using Beziers in Graphics Gems I. In general, you'll find good treatments of whole slews of graphics problems in this series of books. Mark
From: heller@lrz.de (Helmut Heller) Newsgroups: comp.sys.next.programmer Subject: Re: where is libc.a in OpenStep4.2? Date: 14 Dec 1998 18:11:18 GMT Organization: [posted via] Leibniz-Rechenzentrum, Muenchen (Germany) Distribution: world Message-ID: <753kc6$br4$1@sparcserver.lrz-muenchen.de> References: <7538o8$5tf$1@unlnews.unl.edu> In article <7538o8$5tf$1@unlnews.unl.edu> rdieter@math.unl.edu (Rex Dieter) writes: > In article <92lc2.1118$4w2.4950930@news.itd.umich.edu> "Jin Kim" > <kwazyjin@hotmail.com> writes: > > Hello, > > > > I need libc.a to compile some source code that I've downloaded. I can't > > seem to find it anywhere in the OpenStep4.2 install. Do I need to get it > > from elsewhere, or is it in the regular install and I'm just not being > able > > to find it? I did however find libcc.a. Wondering if I could just > > substitute without problem. Thanks! > > There is no libc.a on NEXTSTEP/OpenStep. The standard C library on > openstep it is > /lib/libsys_s.dylib > aka > /NextLibrary/Frameworks/System.framework/System > > FYI, You need to install OpenStep 4.2 Developer (a separate package from > OpenStep 4.2 User) to be able to compile anything > True. But sometimes a libc.a is needed in config scripts and there are a few functions (mmap, setenv, strdup, etc. comes to mind) which are missing in NeXTSTEP. So I made a little dummy libc.a which comes in handy sometimes. You can download it here: http://www.lrz-muenchen.de/~heller/NeXT/libc.tar.gz If you have any functions to contribute, please email them to me and I might include them! -- Servus, Helmut (DH0MAD) ______________NeXT-mail welcome_________________ FAX: +49-89-280-9460 "Knowledge must be gathered and cannot be given" heller@lrz.de ZEN, one of BLAKES7 Phone: +49-89-289-28823 ------------------------------------------------ Dr. Helmut Heller Leibniz-Rechenzentrum (LRZ)
From: bellj@mikado.dialup.ais.net (J. Shan Bell) Newsgroups: comp.sys.next.programmer Subject: Re: Considering WebObjects Date: 15 Dec 1998 00:26:47 GMT Organization: EnterAct L.L.C. Turbo-Elite News Server Distribution: us Message-ID: <754ac7$e8r$1@eve.enteract.com> References: <74qfif$iit@news.service.uci.edu> In article <74qfif$iit@news.service.uci.edu> ecarter@eric.acs.uci.edu (Eric Carter) writes: > Hello, > > I've just returned from a seminar presented by Apple on the > new WebObjects 4. We currently serve all our courses via > the web using Apache/Perl solutions. We are considering > taking the step towards a more rapid development environment > since staff resources are low. Can anyone provide some > personal experience with WebObjects and suggest whether is > would be a good solution for accomplishing our goal? > > Thanks, Eric Carter I doubt you'll get any negative responses from this news group. I developed one application using WO. I'm very pleased with it. It allows you to write code in either Objective C or Java. However, WO is written in Obj C, so you'll find things are easier (maybe even less buggy) than Java. Obj C is pretty easy to learn if you know OO already. The framework of classes that WO is abstracts things like the HTML page, the user session and the request/response loop. It really makes it feel like you're writing an application that would run stand along (instead of being held hostage to the HTTP event loop). Finally, the Enterprise Object Framework that comes with WO gives you the most powerful abstraction of a relationship database I think is available anywhere. I highly recommend WO. Good luck. Shan -- J. Shan Bell NeXTmail/MIME and PGP encrypted mail encouraged Software Engineer finger bellj@cs.indiana.edu for my PGP public key at large http://www.cs.indiana.edu/hyplan/bellj.html
From: far_no@spam.ix.netcom.com(Felipe A. Rodriguez) Newsgroups: comp.sys.next.programmer,gnu.gnustep.discuss,comp.os.linux.x Subject: GNUstep AppKit changes 12/14/98 Date: 15 Dec 1998 07:32:03 GMT Organization: ICGNetcom Message-ID: <75539j$eg5@sjx-ixn8.ix.netcom.com> The latest addition to the GNUstep AppKit's feature set is support for NeXT's field editor mechanism. This is a key part of NeXTStep's fabled ability to pretty much edit, copy and paste from anything to anything. You can try out a preliminary implementation by editing filenames in the Workspace's browser. It doesn't actually change the filesystem though. I just didn't have the nerve to open up Pandora's box and allow experimental code to modify a live filesystem. At least not without doing a careful review of the underlying code in NSFileManager. The code now in CVS also has significant improvements in drawing, performance and more than a few bug fixes (some quite significant). But perhaps the most significant change is simply that more people have been contributing recently. Most notably Richard Frith-Macdonald's services and alert panel code. All in all the AppKit's prospects seem much better than they did even just one month ago :-) That said we could use many more hands so if you're interested info is available at: GNUstep home page http://www.gnustep.org/ Unofficial site http://gnustep.current.nu/ WindowMaker (GNUstep windowmanger) http://www.windowmaker.org/ My own page (just screen shots) http://pweb.netcom.com/~far/far.html Another important change has been the movement of NSDPSContext out of the front end of the AppKit. Basically, NeXT defined the concept of this class as one which represents a Postscript drawing destination. But given the front/back end implementation of GNUstep along with the lack of a viable free DPS it makes more sense to define this class in the XDPS backend. In other words the concept makes very little sense in the case of a drawing engine built on top of Xlib. And by not having to link in the Adobe Client lib it becomes possible to write more of the AppKit abstractly at the root while also sharing more code between the backends. So I extended the concept slightly and at the root of the AppKit you'll now find a GSContext class whose concrete implementation is backend dependent. Conceptually, that means GSContext represents a drawing destination of unknown type. Instances of the XRAW subclass XRContext represent one connection to an X server (multiple connections are possible). So the AppKit inheritance heirarchy ends up as: Abstract XRAW (Xlib) XDPS root or or layer Win32 (GDI) XGPS (client side DPS emulator) NSObject -> XRObject -> XPObject | | | -> XGObject | -> WRObject -> WPObject | -> WGObject I should note that in the above diagram the Win32 backends are only at the conceptual stage. Though given how small and lite XRAW is this should not be very difficult (I hope ;-). The forthcoming client side DPS emulator is being developed by Adam Fedor. Ideally in the future XDPS should be able to inherit much of it's X code from XRAW. And in most cases every attempt should be made to write code abstractly at the root of the AppKit. Given the above I think it becomes apparent that the AppKit can be a useful interface to anything from an embedded application to a full DPS. Felipe A. Rodriguez far@ix.netcom.com Agoura Hills, CA (NeXTmail preferred) (MIMEmail welcome)
#################################################################### From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: Is there a Otool.app for Intel? removing objects Organization: Is a sign of weakness Message-ID: <SCOTT.98Dec14082501@slave.doubleu.com> References: <74ut3b$6i6$1@winter.news.rcn.net> In-reply-to: echo4@NOSPAM.erols.com --delete NOSPAM's message of 12 Dec 1998 23:09:31 GMT Date: 14 Dec 98 08:25:01 NNTP-Posting-Date: Tue, 15 Dec 1998 02:12:35 PDT In article <74ut3b$6i6$1@winter.news.rcn.net>, echo4@NOSPAM.erols.com --delete NOSPAM writes: How you do you remove an object from a library? For example: Removal of sigaction.o from libposix.a I know there was a m68k tool called Otool.app that allowed one to do this, but is there a similiar intel tool? I am pretty sure, you could use the segedit via otool and strip function, but I can figure out how to use the shell tools. Any additional information would be greatly appreciated. Not sure what you really want, but you could use "libtool" or "ar" to work with libraries. To remove an object file from a library, you'll probably have to retrieve all files from the library and build a new one without the file you don't want. I wouldn't expect segedit to be any help when working with a _library_, nor would I expect otool to be any help when trying to modify a library, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: scott@nospam.doubleu.com (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: where is libc.a in OpenStep4.2? Organization: Is a sign of weakness Message-ID: <SCOTT.98Dec14082224@slave.doubleu.com> References: <92lc2.1118$4w2.4950930@news.itd.umich.edu> In-reply-to: "Jin Kim"'s message of Fri, 11 Dec 1998 22:00:10 -0500 Date: 14 Dec 98 08:22:24 NNTP-Posting-Date: Tue, 15 Dec 1998 02:12:35 PDT In article <92lc2.1118$4w2.4950930@news.itd.umich.edu>, "Jin Kim" <kwazyjin@hotmail.com> writes: I need libc.a to compile some source code that I've downloaded. I can't seem to find it anywhere in the OpenStep4.2 install. Do I need to get it from elsewhere, or is it in the regular install and I'm just not being able to find it? I did however find libcc.a. Wondering if I could just substitute without problem. libc.a is just the default C stuff. Most compilers, including NeXT's, include libc.a by default, unless you request otherwise. The actual location of the libc stuff on NeXTSTEP and OpenStep is libsys_s (/lib/libsys_s.a on NeXTSTEP, /lib/libsys_s.dylib on OpenStep). Again, 99% of the time you can just delete the libc.a reference, or -lc if that's how it's referenced, -- scott hess <scott@doubleu.com> (408) 739-8858 http://www.doubleu.com/ <Favorite unused computer book title: The Compleat Demystified Idiots Guide to the Zen of Dummies in a Nutshell in Seven Days, Unleashed>
From: Christian Neuss <neuss.@informatik.th-darmstadt.de.nos-pam> Newsgroups: comp.sys.next.programmer Subject: Re: Considering WebObjects Date: 15 Dec 1998 10:50:48 GMT Organization: Technische Universitaet Darmstadt Message-ID: <755eu8$4ft$1@sun27.hrz.tu-darmstadt.de> References: <74qfif$iit@news.service.uci.edu> ecarter@eric.acs.uci.edu (Eric Carter) wrote: >Hello, > >I've just returned from a seminar presented by Apple on the >new WebObjects 4. We currently serve all our courses via >the web using Apache/Perl solutions. We are considering >taking the step towards a more rapid development environment >since staff resources are low. Can anyone provide some >personal experience with WebObjects and suggest whether is >would be a good solution for accomplishing our goal? Well, it took us a little over a man month to build www.sendamac.de. Highly recommended. Chris -- // Christian Neuss "static typing? how quaint.." // fax: (+49) 6151 16 5472
From: Pegasus <veasara@tin.it> Newsgroups: comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software Subject: Progetto Date: Tue, 15 Dec 1998 10:00:50 +0100 Organization: TIN Message-ID: <36762542.8AAF8383@tin.it> Mime-Version: 1.0 Content-Type: multipart/alternative; boundary="------------E31A5B1CA48BF98A6FD7E82B" --------------E31A5B1CA48BF98A6FD7E82B Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit PROGETTO! LEGGIMI CON ATTENZIONE SENZA SALTARE NEMMENO UNA PAROLA E… E’ sufficiente trovare 10 / 15 persone (ti spiegherò poi come) che abbiano una buona apertura mentale e l’intelligenza di comprendere un metodo matematico infallibile. Di primo acchito si potrebbe avere l’impressione di avventurarsi in un progetto impossibile. Io ti garantisco che lavorando poche ore al computer si raggiungono risultati inimmaginabili, io sono un’agente di commercio da vent’anni, negli ultimi mesi ho dedicato moltissimo tempo a questo progetto - all’inizio ero scettico ma i risultati mi hanno fatto cambiare opinione – oggi, comodamente seduto a casa mia, raggiungo risultati di reddito migliori di quelli che ottenevo lavorando 12-14 ore al giorno sempre in auto e alle prese con clienti che… In Internet ci sono ogni giorno migliaia di nuovi ingressi, migliaia di persone che come noi navigano alla ricerca di informazioni, contatti, amicizie e a volte ci sfiorano opportunità che non cogliamo perché presi dalle mille preoccupazioni e attività quotidiane. Questo progetto funziona benissimo per tutti ma più ancora a chi lavorando in ufficio ha modo di usare il computer e conosce molte persone anche al di fuori dell’azienda per cui lavora e il passaparola è la migliore pubblicità. Io lo faccio da casa. Avventurarti in questo progetto non ti costerà nulla e ti farà guadagnare moltissimo, con i primi risultati ti accorgerai che diventerà per te una “droga” e impegnerai sempre più tempo in modo piacevole e autonomo, non dovrai MAI rendere conto (se non a te stesso) del tuo operato a NESSUNO! Se questo mio messaggio ti ha incuriosito scrivimi e ti darò le necessarie istruzioni per iniziare il progetto. Nella richiesta di informazioni come oggetto metti PROGETTO veasara@tin.it Ciao a presto. PEGASUS -- Rimuovi XXX dall'indirizzo E-Mail Remove XXX from E-Mail address. --------------E31A5B1CA48BF98A6FD7E82B Content-Type: text/html; charset=iso-8859-1 Content-Transfer-Encoding: 8bit <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> <HTML> <B><FONT COLOR="#3333FF">PROGETTO!</FONT></B> <BR><B><FONT COLOR="#3333FF">LEGGIMI CON ATTENZIONE SENZA SALTARE NEMMENO UNA PAROLA E…</FONT></B> <BR>E’ sufficiente trovare 10 / 15 persone (ti spiegherò poi come) che abbiano una buona apertura mentale e l’intelligenza di <BR>comprendere un metodo matematico infallibile. Di primo acchito si potrebbe avere l’impressione di avventurarsi in un progetto <BR>impossibile. Io ti garantisco che lavorando poche ore al computer si raggiungono risultati inimmaginabili, io sono un’agente di <BR>commercio da vent’anni, negli ultimi mesi ho dedicato moltissimo tempo a questo progetto - all’inizio ero scettico ma i risultati <BR>mi hanno fatto cambiare opinione – oggi, comodamente seduto a casa mia, raggiungo risultati di reddito migliori di quelli che <BR>ottenevo lavorando 12-14 ore al giorno sempre in auto e alle prese con clienti che… <BR>In Internet ci sono ogni giorno migliaia di nuovi ingressi, migliaia di persone che come noi navigano alla ricerca di informazioni, <BR>contatti, amicizie e a volte ci sfiorano opportunità che non cogliamo perché presi dalle mille preoccupazioni e attività quotidiane. <BR>Questo progetto funziona benissimo per tutti ma più ancora a chi lavorando in ufficio ha modo di usare il computer e conosce <BR>molte persone anche al di fuori dell’azienda per cui lavora e il passaparola è la migliore pubblicità. Io lo faccio da casa. <BR>Avventurarti in questo progetto non ti costerà nulla e ti farà guadagnare moltissimo, con i primi risultati ti accorgerai che <BR>diventerà per te una “droga” e impegnerai sempre più tempo in modo piacevole e autonomo, non dovrai MAI rendere conto (se <BR>non a te stesso) del tuo operato a NESSUNO! Se questo mio messaggio ti ha incuriosito scrivimi e ti darò le necessarie <BR>istruzioni per iniziare il progetto. <P>Nella richiesta di informazioni come oggetto metti PROGETTO <BR><A HREF="mailto:veasara@tin.it">veasara@tin.it</A> <P><I>Ciao a presto. PEGASUS</I> <P>-- <BR>Rimuovi XXX dall'indirizzo E-Mail <BR>Remove XXX from E-Mail address. <BR> </HTML> --------------E31A5B1CA48BF98A6FD7E82B--
From: spamcancel@wupper.com Newsgroups: comp.sys.next.hardware,comp.sys.next.marketplace,comp.sys.next.misc,comp.sys.next.programmer,comp.sys.next.software Subject: cmsg cancel <36762542.8AAF8383@tin.it> Control: cancel <36762542.8AAF8383@tin.it> Date: 15 Dec 1998 11:17:44 GMT Message-ID: <cancel.36762542.8AAF8383@tin.it> Sender: Pegasus <veasara@tin.it> Excessive Multi-Posted spam article exceeding a BI of 20 cancelled by spamcancel@wupper.com. From was: Pegasus <veasara@tin.it> Subject was: Progetto NNTP-Posting-Host was: a-fi14-15.tin.it
Message-ID: <3676AE99.D2AB9D7F@postoffice.pacbell.net> From: kzworld@postoffice.pacbell.net MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Objective C NeXTStep Programmers Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Tue, 15 Dec 1998 10:46:49 -0800 NNTP-Posting-Date: Tue, 15 Dec 1998 10:45:49 PDT Organization: SBC Internet Services Hello, I am seeking programmers with the skill set which includes Objective C in a NeXT/OpenStep environment. If you have this skill set or are aware of anyone who does please contact me. Kay Lynn Gabaldon Recruiter Los Angeles, California (310) 337-0305 kzworld@pacbell.net
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer Subject: Re: where is libc.a in OpenStep4.2? Date: 15 Dec 1998 18:57:57 GMT Organization: Technische Universitaet Berlin, Deutschland Message-ID: <756bfl$hg0$1@news.cs.tu-berlin.de> References: <92lc2.1118$4w2.4950930@news.itd.umich.edu> <SCOTT.98Dec14082224@slave.doubleu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit NNTP-Posting-Date: 15 Dec 1998 18:57:57 GMT scott@nospam.doubleu.com (Scott Hess) writes: >Again, 99% of the time you can just delete the libc.a reference, or >-lc if that's how it's referenced, Also, if you you have a distribution where it's difficult to remove the reference, you can easily create a dummy libc.a and place it in your /usr/local/lib. (I did this for libm.a a while back, don't know what the package was that was giving me grief). I think this was enough: > echo -n >dummy.o > ar -r libc.a dummy.o ar: creating archive libc.a > ranlib lic.a ranlib: warning for library: libc.a the table of contents is empty (no object file members in the library) The warning can be ignored. You might need to compile a dummy c file instead of just echo-ing, I don't remember exactly. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: Rick_Olsen@mail.ci.baltimore.md.us Newsgroups: comp.sys.next.programmer Subject: Page Not Found Date: Wed, 16 Dec 1998 15:20:24 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <758j3k$i3a$1@nnrp1.dejanews.com> We just took over a system that has WebObjects on a solaris server running Netscape. The application worked and now doesnt. We have been removing the next boxes on the network but that is all I can prove has happened. The obj.conf does not have a transname entry for cgi-bin, WebObjects or the combination. Should it? Any help would be appreciated. rick. -----------== Posted via Deja News, The Discussion Network ==---------- http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
From: Rick_Olsen@mail.ci.baltimore.md.us Newsgroups: comp.sys.next.programmer Subject: HELP --- obj.conf whats wrong with it --- HELP Date: Wed, 16 Dec 1998 16:43:24 GMT Organization: Deja News - The Leader in Internet Discussion Message-ID: <758nvc$mdt$1@nnrp1.dejanews.com> I get the error message "conf_init: 3: bad parameters while processing load file obj.conf". my obj.conf is as follows: <Object name=default> NameTrans fn=pfx2dir from=/mc-icons dir="/usr/local/netscape/ns-home/mc-icons" NameTrans fn=document-root root="/export/htdocs" NameTrans from="/cgi-bin/WebObjects" fn="WONSInterfaceFindWebObjects" name="webobjects" NameTrans from="/cgi-bin" fn="pfx2dir dir="/export/cgi-bin" name="cgi" PathCheck fn=unix-uri-clean PathCheck fn=find-pathinfo PathCheck fn=find-index index-names="index.html,home.html" ObjectType fn=type-by-extension ObjectType fn=force-type type=text/plain Service method=(GET|HEAD) type=magnus-internal/imagemap fn=imagemap Service method=(GET|HEAD) type=magnus-internal/directory fn=index-common Service method=(GET|HEAD) type=*~magnus-internal/* fn=send-file AddLog fn=common-log Init fn=load-modules shlib=/export/opt/WOF/NextLibrary/WOAdaptors/NSAPI/2.0/WebObjects-NSAPI.so funcs="WONetscapeInterface,WONSInterfaceFindWebObjects" </Object> <Object name=cgi> ObjectType fn=force-type type=magnus-internal/cgi Service fn=send-cgi </Object> <Object name="webobjects"> Service fn="WONetscapeInterface" </Object> ------------- magnus.conf: #ServerRoot /usr/local/netscape/ns-home/httpd-80.IPADDR Port 80 Address 169.156.33.12 LoadObjects obj.conf RootObject default ErrorLog /usr/local/netscape/ns-home/httpd-80.169.156.33.12/logs/errors PidLog /usr/local/netscape/ns-home/httpd-80.169.156.33.12/logs/pid User nobody ServerName orion.ci.baltimore.md.us MinProcs 16 MaxProcs 32 DNS on Init fn=load-types mime-types=mime.types Init fn=init-clf global=/usr/local/netscape/ns-home/httpd-80.169.156.33.12/logs/access Init fn=load-modules shlib=/export/opt/WOF/NextLibrary/WOAdaptors/NSAPI/1.1/WebObjects-NSAPI.so funcs="WONetscapeInterface,WONSInter faceFindWebObjects" -----------== Posted via Deja News, The Discussion Network ==---------- http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
From: epeyton@apple.com (Eric Peyton) Newsgroups: comp.sys.next.programmer Subject: Re: WebObjects and EnterpriseObjects Date: 16 Dec 1998 18:49:38 GMT Organization: Apple Computer Message-ID: <758vc2$h9m$1@news.apple.com> References: <74rjhd$k8c$1@saltmine.radix.net> Mime-Version: 1.0 Content-Type: Text/Plain; charset=US-ASCII Did you generate an EOUser class inside EOModeler? I.e. Generate an EOUser.java file and then compile that into the application? (I would highly recommend that you don't name this class EOUser, EO is reserved for the internal Enterprise BOjects classes) Eric Peyton In article <74rjhd$k8c$1@saltmine.radix.net>, tgia@saltmine.radix.net says... > > >Ok folks, as some of you may realize from the last message. >I'm trying to evaluate WebObjects. I've gotten pretty far, but >reached an impasse which I'm hoping you all can help me with. > >I created a data model, using the eomodler tool. Originally >I accepted the default class for the entities EOGenericRecord. > >Now I want to go back and change those to class specific objects. > >So I opened up the eomodler from the project builder and changed the >classes to the new types I wanted to use. Here's an example > >Entity Table Class >User UserTable EOGenericRecord (before) >User UserTable EOUser (after my object) > >Now in my code for a particular page I have the following code. > > > protected EOUser aUser > protected EOUser ourUser > protected WODisplayGroup usersDisplayGroup; > protected NSMutableArray selectedUsers; > > >public void >selectUser() >{ > > if (selectedUsers.count() > 0) > { > ourUser = (EOUser) selectedUsers.objectAtIndex(0); // Bomb! > ((Session)this.session).setUser(ourUser); > } >} > >In the browser the following attributes are set: > >attribute Binding >-------------------------------------- >displayString aUser.name >item aUser >list userDisplayGroup.displayedObjects >selections selectedUsers > > > >This code bombs on the line noted. And it bombs because the return >value from the selectedUsers is of type EOGenericRecord. > >Now the questions > >1) Will the WODisplayGroup ever but any kind of object other then > EOGenericRecord in it's selection Array? > >2) If it will how do I get it to return my Objects, instead of > EOGenericUser > >3) If it doesn't, how can I use the EOGenericRecord that is returned > to retrieve the EOUser object from the DB? > >I've read through the manual for the last few days, and either I don't >understand the terminology and object relationship in EO (Strike that I >clearly don't understand it very well) and I can't recognize the sections >I should be reading, or it's doesn't seem to be ther. > >Can anyone shed some light on this? > >Thanks > > >Tony Giaccone
Message-ID: <3679287F.DC6419BE@pacbell.net> From: Kay Lynn Gabaldon <kaylynn@pacbell.net> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Objective C NeXTStep/OpenStep Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Thu, 17 Dec 1998 07:51:27 -0800 NNTP-Posting-Date: Thu, 17 Dec 1998 07:40:59 PDT Organization: SBC Internet Services Attention Road Warriors: I am seeking programmers/coders with Objective C who have worked in a NeXT/OpenStep environment for a long term assignment in the MidWest region of the United States. If you have this skill set or know anyone who does please contact me. I am looking to create a team to submit to the client/s in January, 1999. Respectfully, Kay Lynn Gabaldon WnC, Los Angeles (310) 337-0305
From: "=?big5?B?r6vz56tMq1E=?=" <hkh07382@ms23.hinet.net> Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin,comp.sys.northstar,comp.sys.nsc.32k,comp.sys.oric,comp.sys.palmtops,comp.sys.palmtops.pilot,comp.sys.pen,comp.sys.powerpc,comp.sys.powerpc.advocacy,comp.sys.powerpc.misc,c Subject: =?big5?B?UmU6IKXOpHC/+qjTwcikar/6qrqtULRJpOiqaw==?= Date: Sun, 20 Dec 1998 01:49:06 +0800 Organization: DCI HiNet Distribution: inet Message-ID: <75gota$9ou@netnews.hinet.net> References: <366A55FF.15D79C16@ms24.hinet.net> <74vhp0$3ea@netnews.hinet.net> Mime-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0131_01BE2BBA.EDD938A0" This is a multi-part message in MIME format. ------=_NextPart_000_0131_01BE2BBA.EDD938A0 Content-Type: text/plain; charset="big5" Content-Transfer-Encoding: quoted-printable =A7=DA=A4]=A6=AC=A8=EC=BF=FA=A4F=A1I =A7A=A9=CE=B3\=B7|=A5H=AC=B0=B3o=A5u=ACO=AD=D3=C4F=A4H=AA=BA=BE=BB=C0Y=A1= A=A5=FD=ABe=A7=DA=AC=DD=A8=EC=C3=FE=A6=FC=AA=BA=ABH=A5=F3=AE=C9=A1A=A4]=B4= =BF=A6=B3=B3o=BA=D8=B7Q=AAk=A1A=A6=FD=A4p=A4p=AA=BA=A7=EB=B8=EA=C3B=A4=DE= =B0_=A7=DA=AA=BA=BF=B3=BD=EC=A1A=B4N=BA=E2=BD=DF=A4F=A4]=A5u=ACO500=A4=B8= =A5=E1=A5X=A5h=B6R=B1=D0=B0V=A1C =A9=EA=B5=DB=A4S=B4=C1=AB=DD=A4S=A9=C8=A8=FC=B6=CB=AE`=AA=BA=A4=DF=B1=A1=A1= A=A7=DA=A6U=B1H100=A4=B8=B5=B9=ABH=A5=F3=A6W=B3=E6=A4=A4=AA=BA=A4=AD=A6W=A6= =AC=A5=F3=AA=CC=A1A=B1=B5=A4U=A8=D3=AA=BA=A8=E2=A4T=A4=D1=A1A=A7=DA=B6}=A9= l=A6b=A6U=AD=D3=B7s=BBD=B0Q=BD=D7=B0=CF=B5o=B3o=AB=CA=ABH=A1A=B2=C4=A4@=AD= =D3=C2=A7=AB=F4=A7=DA=AAG=AFu=A6=AC=A8=EC=BF=FA=A4F=A1A=A7=DA=A6=AC=A8=EC= =A4K=AB=CA=B8=CB=A6=B3100=A4=B8=B2{=AA=F7=AA=BA=ABH=A1A=B5=A5=A9=F3=BB=A1= =A1A=A7=DA=A4w=B8g=C1=C8=A6^=A7=DA=A9=D2=A7=EB=B8=EA=AA=BA=AA=F7=C3B=A1A=C1= =D9=A5~=A5[300=A4=B8=AA=BA=A6=AC=A4J=A1C =B2=C4=A4G=AD=D3=C2=A7=AB=F4=AB=E1=A1A=A7=DA=AA=BA=A6=AC=A4J=A7=F3=C5=E5=A4= H=A1A=B3=BA=B5M=A6=AC=A8=EC=A4G=A4Q=A4=AD=AB=CA=AA=BA=ABH=A1]=A4]=B4N=ACO= 2500=A1^=A1A=A7=DA=A9=F3=ACO=A5=B4=C5K=B6X=BC=F6=A6b=B5o=A5X=B3\=A6h=AA=BA= =ABH=A5=F3=A6b=BA=F4=B8=F4=A4W=A1A=A7A=AC=DB=ABH=B6=DC=A1H =A6b=A4@=AD=D3=A4=EB=AB=E1=A7=DA=A5=CE500=A4=B8=C1=C8=B6i=A4F27500=A4=B8=A1= A=A7=EB=B8=EA=A9=EA=B9S=B2v=A4=A7=B0=AA=A1A=B3s=AA=D1=B2=BC=B3=A3=B1=E6=A4= =A7=B2=F6=A4=CE=A1C =AB=E1=A8=D3=A4S=C2_=C2_=C4=F2=C4=F2=A6=AC=A8=EC=BF=FA=A1A=A5=D8=ABe=A4w=B8= g=B2=D6=BFn=AA=F1=A4Q=A6=EC=BC=C6=A1A=A5=BB=A8=D3=A6]=AC=B0=A4=F7=A5=C0=B8= g=B0=D3=A5=A2=B1=D1=A6=D3=A4T=C0\=B3=A3=A6=A8=B0=DD=C3D=AA=BA=A7=DA=A1A=B2= {=A6b=BE=C7=B6O=A9M=A5=CD=AC=A1=B6O=A5=FE=A6=B3=B5=DB=B8=A8=A4F=A1I =C5=A5=A7=DA=BB=A1=A7=B9=A6=DB=A4v=AA=BA=BF=CB=A8=AD=B8g=BE=FA=A1A=A7A=A4= @=A9w=A4]=C5D=C5D=B1=FD=B8=D5=BD}=A1I =B3o=BA=D8=A7=EB=B8=EA=A4=E8=AAk=AB=DC=C2=B2=B3=E6=A1A=AD=BA=A5=FD=A7A=A5= =B2=B6=B7=AD@=B5=DB=A9=CA=A4l=AC=DD=A7=B9=A7=DA=A7Y=B1N=A7i=B6D=A7A=AA=BA= =A8=E2=A5=F3=A8=C6=A1]=B3=CC=A6n=A7=E2=A5=A6=A6C=A6L=A5X=A8=D3=A1^=A1G =A2=B0.=B3o=AD=D3=ADp=B5e=AA=BA=B9=EA=A6=E6=A4=E8=AAk=A1C =A2=B1.=B3o=AD=D3=ADp=B5e=A5i=A6=E6=AA=BA=AD=EC=A6]=A1C 1.=B3o=AD=D3=ADp=B5e=AA=BA=B9=EA=A6=E6=A4=E8=AAk = =A2=CF.=B7=C7=B3=C6=A7=F7=AE=C6=A1G=A4=AD=AD=D3=BC=D0=B7=C7=ABH=AB=CA=A1B= =A4=AD=B1i=A4=AD=A4=B8=AA=BA=B6l=B2=BC=A1B=A4=AD=B1i=C3C=A6=E2=AB=DC=B2`=AA= =BA=AF=C8 = =A1]=ADn=AF=E0=A7=E2=A2=B0=A2=AF=A2=AF=A4=B8=A5]=A6=ED=A1A=B5L=AAk=AC=DD=A8= =A3=AA=BA=B2`=A6=E2=A1^=A1B=A4=AD=B1i=ABK=B1=F8=AF=C8=A1C = =A2=D0.=A4=E8=AAk=A1G=A6b=A8C=B1i=ABK=B1=F8=AF=C8=A4W=BCg=B5=DB=A1y=BD=D0= =B1N=A7=DA=A9=F1=A4J=A7A=AA=BA=B6l=BB=BC=A6W=B3=E6=A4=A4=A1z=A1F=B1N=A2=B0= =A2=AF=A2=AF=A4=B8=A5=CE=B2`=A6=E2 =AF=C8=A5]=B0_=A8=D3=A1A = =A4=C0=A7O=A9=F1=A4J=A8C=AD=D3=ABH=AB=CA=A4=A4=A1]=A4@=A9w=ADn=C5=FD=A2=B0= =A2=AF=A2=AF=A4=B8=B5L=AAk=B3Q=B3z=B5=F8=A1A=A7_=ABh=AEe=A9=F6=B3Q=A4=A3=A8= v=A4=A7=AE{=AE=B3=A8=AB=A1^=A1C = =A2=D1.=B1N=B3o=A4=AD=AB=CA=ABH=A4=C0=A7O=B1H=B5=B9=A5H=A4U=AA=BA=A4=AD=AD= =D3=A4H=A1]=AA`=B7N=A1G=A4=A3=A5i=A5H=C0H=ABK=B1H=A1^=A1G 1.=B6=C0=A7g=B0=B6 = =A5x=A4=A4=A5=AB=A5_=A4=D9=B0=CF=A6Z=C9=DC=B8=F4128=B8=B97=BC=D3 = =B6l=BB=BC=B0=CF=B8=B9 406 2. =AAL=B4=AD=AA@ = =B0=AA=B6=AF=BF=A4=A9=A3=A4s=C2=ED=A9=A3=A4s=B8=F4287=B8=B9 = =B6l=BB=BC=B0=CF=B8=B9820 3.=A7=F5=A9y=AF=D5 =A5x=A5_=A5=AB=AAQ=A6=BF=B8=F4510=B8=B93=BC=D3 = =B6l=BB=BC=B0=CF=B8=B9104 4.=AAL=A9=FA=BC=DD = =A5x=A4=A4=BF=A4=AFQ=A4=E9=B6m=A5_=A8=BD=A7=F8=B7=CB=ABn=B8=F463-2=B8=B9 = =B6l=BB=BC=B0=CF=B8=B9 414 5.=B6=C0=AB=B6=BD=F7 = =AE=E7=B6=E9=BF=A4=A4K=BCw=A5=AB=A9M=A5=AD=B8=F4621=AB=D149=B8=B9 = =B6l=BB=BC=B0=CF=B8=B9334 (=B0O=B1o=BCg=B6l=BB=BC=B0=CF=B8=B9 = =A7_=ABh=B6l=A7=BD=B1N=A9=B5=AA=F8=A4u=A7@=A4=D1=B3=E1!) =A2=D2.=A7=F3=A7=EF=A6a=A7}=A1G =A1@=A1@=B1H=A7=B9=ABH=AB=E1=A1A=A6A=B3z=B9L=B9q=B8=A3=B1N=A4W=AD=B1=B2=C4= =A4@=AD=D3=A4H=AA=BA=A6W=A6r=BD=F0=B1=BC=A1A=A8=C3=A8=CC=A7=C7=B1N=A5|=AD= =D3=A4H=AA=BA=A6W=A6r=A9=B9=A4W=B2=BE=A4G=B2=BE=A8=EC=A4@=A1F=A4T=B2=BE=A8= =EC=A4G=A1F=A5|=B2=BE=A8=EC=A4T=A1F=A4=AD=B2=BE=A8=EC=A5|=A1^=A1A=A6=D3=B2= =C4=A4=AD=AD=D3=BCg=A4W=A7A=AA=BA=A6W=A6r=A1B=A6a=A7}=A1A=ADn=AA`=B7N=A6a= =A7}=ACO=A5=BF=BDT=AA=BA=A1A=A7_=ABh=A7A=B7|=A6=AC=A4=A3=A8=EC=ABH=B3=E1=A1= I = =A2=D3.=A5=E1=A4J=B7s=BBD=B8s=B2=D5=A1G=A4=E5=B3=B9=A4=A4=B6=C8=BB=DD=AD=D7= =A7=EF=A1u=A2=D2=A1B=A7=F3=A7=EF=A6a=A7}=AA=BA=B3=A1=A5=F7=A1v=A9=D2=B4=A3= =A8=EC=AA=BA=B3=A1=A5=F7=A1A=ABe=AD=B1=A9=D2 =BB=A1=AA=BA=B8g=C5=E7=B3=A1=A5=F7=A1A=A7A=A5i=A5H=AE=DA=BE=DA=A6=DB=A4v=AA= =BA=B9=EA=AAp=B0=B5=C2=B2=A9=F6=A7=F3=A7=EF=A1A=A8=E4=A5L=AA=BA=B3=A1=A5=F7= =B3=CC=A6n=ABO=AB=F9=AD=EC=AA=AC=A1A=B5M=AB=E1=A7A=A5=B2=B6=B7=B1N=ABH=A5= =F3=A6=DC=A4=D6=A4=BD=A7G=A6b250=AD=D3=B7s=BBD=B0Q=BD=D7=B0=CF=A1]=B3=CC=A4= =D6=ADn=A6=B3=A8=E2=A6=CA=AD=D3=B0Q=BD=D7=B0=CF=A1A=A5=FE=A5@=AC=C9=AC=F9= =A6=B325000=AD=D3=B7s=BBD=B0Q=BD=D7=B0=CF=A1C) =A1=AF=AA`=B7N=A1G=A7A=A9=D2=A4=BD=A7G=A5X=A5h=AA=BA=B7s=BBD=B0=CF=B6V=A6= h=A1A=B1z=B1o=A8=EC=AA=BA=A6^=F5X=B6V=A6h=A1C =ADY=A4=A3=AA=BE=A6p=A6=F3=A4=BD=A7G=A8=EC=B7s=BBD=B0=CF=A1A=B1z=A5i=A5H=B7= =D3=A4U=A6C=AA=BA=A4=E8=AAk=A7@=A1G =A1=D7=A1=D7=A1=D7=A1=D7=A8=CF=A5=CE=BA=F4=B4=BA=A2=DC=A2=D3=A2=E2=A2=E1=A2= =D1=A2=CF=A2=DE=A2=D3=A1@=A2=DC=A2=CF=A2=E4=A2=D7=A2=D5=A2=CF=A2=E2=A2=DD= =A2=E0=AA=BA=A8=CF=A5=CE=AA=CC=A1=D7=A1=D7=A1=D7=A1=D7 =A2=B0. =A5=FD=A5=B4=B6}NETSCAPE = NEWS=A1A=A6A=A6b=BA=F4=B4=BA=C2s=C4=FD=BE=B9=A4=A4=B1q=A5\=AF=E0=BF=EF=B3= =E6=A4=A4=A5=FD=BF=EF=A8=FAWindows=A1A=B1=B5=B5=DB=A6A=BF=EF=A8=FANETSCAP= E NEWS=A1A=A7Y=A5i=B1=D2=B0=CANETSCAPE NEWS=A1C =A2=B1. =C2I=A8=FA=B3=CC=A5=AA=C3=E4=AA=BA=AB=F6=B6s=A1yPOST = NEWS=A1z=A6=B9=AE=C9=B7|=A5X=B2{=A4@=AD=D3=BDs=BF=E8=B7s=A4=E5=B3=B9=AA=BA= =B5=F8=B5=A1=A1C =A2=B2. = =AC=B0=A7A=AA=BA=A4=E5=B3=B9=BF=EF=A4@=AD=D3=A7l=A4=DE=A4H=AA=BA=BC=D0=C3= D=A1A=B1N=A7A=A5=D8=ABe=AC=DD=A8=EC=AA=BA=B3o=BDg=A4=E5=B3=B9COPY=B5M=AB=E1= PASTE=A8=EC=B0T=AE=A7=C4=E6=A4=BA=A1A=B7=ED=B5M=A7A=A5i=A5H=BEA = =B7=ED=AA=BA=AD=D7=A7=EF=B3o=BDg=A4=E5=B3=B9=A1]=A7A=A5=B2=B6=B7=B1N=A7A=AA= =BA=A6W=A6r=A6a=A7}=A5[=A6b=A4=E5=B3=B9=AA=BA=B2=C4=A4=AD=A6=EC=A1A=A8=C3= =B1N=B3=D1=A4U=A5|=AD=D3=A4H=AA=BA=A6W=A6r=A6a=A7}=A9=B9=A4W=B5=A5=B2=BE=A4= @=A6=EC=A1^=A1C=B7=ED=A7A=A7=E2=BE=E3=BDg=A4=E5=B3=B9=A5=B4=A6n=A9=CECOPY= =A4=A7=AB=E1,=C2I=BF=EF=A1ySEND = NOW=A1z.=B2{=A6b=B6}=A9l=B3=B0=C4=F2=B6K=A9=B9=A4=A3=A6P=AA=BA=B7s=BBD=B0= Q=BD=D7=B0=CF=A7a=A1I =A1=D7=A1=D7=A1=D7=A1=D7 = =A8=CF=A5=CE=B7L=B3n=AA=BA=B1=B4=C0I=AEa=C2s=C4=FD=BE=B9 Microsoft = Internet Explorer=A1=D7=A1=D7=A1=D7=A1=D7 =A2=B0.=A4=E8=AAk=A4]=ACO=A4@=BC=CB=AA=BA=C2=B2=B3=E6=A1A=AB=F6=A6=ED=B7=C6= =B9=AB=A5=AA=C1=E4=A4=A3=A9=F1=A1A=B1N=BE=E3=AD=D3=A4=E5=B3=B9=A4=CF=A5=D5= =B0_=A8=D3=A1A=B5M=AB=E1=A6P=AE=C9=AB=F6=B5=DB =A1yCTRL=A1z=A4=CE=A1yC=A1z=C1=E4=B1N = =BE=E3=BDg=A4=E5=B3=B9=AB=FE=A8=A9=A8=EC=B0=C5=B6K=C3=AF=A1A=A6=B9=AE=C9=A6= A=B1N=A5=BB=A4=E5=A6C=A6L=A5X=A8=D3=A1A=B3o=BC=CB=A7A=B4N =A6=B3=A4@=A5=F7=A7A=B4=BF=B8g=B1H=A5X=AA=BA=A6W=B3=E6=A4F=A1C=A1]=B0O=B1= o=B1N=A7A=A6=DB=A4v=AA=BA=A6W=A6r=A6a=A7}=A5[=A6b=A6W=B3=E6=AA=BA=B3=CC=AB= =E1=A1I=A1^ =A2=B1. = =B3o=AE=C9=AD=D4=A6A=A6=B8=B1N=A4=E5=B3=B9=A4=CF=A5=D5=B0_=A8=D3=A1A=AD=AB= =B7sCOPY=A4@=A6=B8=A1A=A5H=BDT=BB{=A7A=A4w=B8gCOPY=A4F=AD=D7=A5=BF=B9L=AA= =BA=A4=E5=B3=B9=A4=BA=AEe=A1C=B3o=AE=C9=AD=D4=A7A=A5u=BB=DD=ADn=A6b=A7A=B7= =C7=B3=C6=B6K=A4=E5=B3=B9=AA=BA=B0Q=BD=D7=B0=CF=BF=EF=BE=DC=B5o=AA=ED=B7s= =AA=BA=A4=E5=B3=B9=A1A=B5M=AB=E1=A5HCTRL+V=A1z=B1N=A4=E5=B3=B9=A4@=B0_=B6= K(PASTE)=A5X=A1C =A1@=A1@=B4N=ACO=B3o=BB=F2=C2=B2=B3=E6=A1A=A7A=A5u=BB=DD=ADn=BF=EF=BE=DC=A4= =A3=A6P=AA=BA=B7s=BBD=B0Q=BD=D7=B0=CF=A1A=B5M=AB=E1=BF=EF=BE=DC=A1yPOST=A1= z=B5o=AA=ED=A4=E5=B3=B9=B5=A5=A7A=B2=DF=BAD=BE=E3=AD=D3=B9L=B5{=AB=E1=A1A= =A8C=B5o=A4@=AB=CA=ABH=A4j=AC=F9=A5u=BB=DD=ADn30=AC=ED=AA=BA=AE=C9=B6=A1=A9= O=A1I =A1@=A1@=A5=D1=A9=F3=A5=BB=A4=E5=AC=B0=A4=A4=A4=E5=A1A=ACG=BE=A8=B6q=B5o=A4= =A4=A4=E5=B0=CF=B8s=B2=D5=B8=FB=A6=B3=AE=C4=A1C=A7A=A5i=A5H=C2I=BF=EF=A1e= =B7s=BBD=A1f=A1=D0=A1e=BF=EF=BE=DC=B7s=BBD=B8s=B2=D5=A1f=A1A=A6=B9=AE=C9=A5= X=B2{=B3\=A6h=B0Q=BD=D7=B0=CF=A1A=B1z=A5i=A5H=A6b=A4W=A4=E8=BF=E9=A4J=B7j= =B4M"ROC" =A1A = "TAIWAN"=B5=A5=C3=F6=C1=E4=A6r=A1A=ABh=A9=D2=A8=A3=A6C=A5X=AA=BA=B4N=ACO=A5= x=C6W=AC=DB=C3=F6=AA=BA=B0Q=BD=D7=B0=CF=A1A=B3o=AE=C9=B4N=A8=CC=B1z=AD=D3= =A4H=BB=DD=ADn=ACD=BF=EF=A8=C3=B1i=B6K=A4=E5=B3=B9=A4F=A1C=BD=D0=B0O=A6=ED= =A1A=A7A=B6K=AA=BA=B7s=BBD=B0Q=BD=D7=B0=CF=B6V=A6h=A1A=AC=DB=B9=EF=A6a=A7= A=B4N=B7|=B1o=A8=EC=B6V=A6h=AA=BA=A6^=C0=B3=A1C=A6P=AE=C9=A7A=A4]=B7|=B1o= =A8=EC=B6V=A6h=AA=BA=A6^=C5T=A1I=B4N=ACO=B3o=BB=F2=C2=B2=B3=E6=A1A=A7A=B1= N=A6=AC=A8=EC=A8=D3=A6=DB=A5=FE=BBO=C6W=A6U=A6a=A4D=A6=DC=A9=F3=A5=FE=B2y= =A6U=A6a=AA=BA=A6^=ABH,=A7=C6=B1=E6=AF=E0=A5[=A4J=A7A=AA=BA"=B6l=BB=BC=A6= W=B3=E6"!=A1]=B7=ED=B5M,=BD=D0=B0=C8=A5=B2=BDT=BB{=B1H=B0e=A6a=A7}=AA=BA=A5= =BF=BDT!=A1^=A1@=A1@=A1@=A1@ 2.=B3o=AD=D3=ADp=B5e=A5i=A6=E6=AA=BA=AD=EC=A6] = =A6b=A7A=B5o=A5X250=AB=CA=A4=E5=B3=B9=A4=A4=A1A=B0=B2=B3]=A5u=A6=B3=A4=AD= =AD=D3=A4H=A6^=C0=B3=A7A=AA=BA=A4=E5=B3=B9=A1]=B3o=AD=D3=C1|=A6C=A5=CE=AA= =BA=A4=F1=A8=D2=AC=DB =B7=ED=A7C=A1^=A1A=B7=ED=A7=DA=AA=BA=A6W=A6r=A6b=B2=C4=A4=AD=A6=EC=AE=C9=A7= =DA=C1=C8=B6i=A4F250=A4=B8=A1C=B2{=A6b=A6^=C0=B3=A7=DA=A4=E5=B3=B9=AA=BA=A4= =AD=AD=D3=A4H=A1A=A4]=B4N=ACO=B1H=B5=B9=A7=DA=B7s=BBO=B9=F4100=A4=B8=AA=BA= =B3o=A4=AD=AD=D3=A4H=A4S=A6U=A6=DB=A6=DC=A4=D6=B6K=A5X250=AB=CA=A4F=A4=E5= =B3=B9=A1A=A6=B9=AE=C9=A7=DA=AA=BA=A6W=A6r=B1N=A4W=B2=BE=A6=DC=B2=C4=A5|=A6= =EC=A1F=B0=B2=B3]=A4]=A5u=A6=B3=A4=AD=AD=D3=A4H=A6U=A6=DB=A6^=C0=B3=B5=B9= =A5L=AD=CC=B3o=A4=AD=AD=D3=A4H=A1]=A4]=B4N=ACO=B3=CC=A6=AD=B1H=BF=FA=B5=B9= =A7=DA=AA=BA=A4H=A1^=A1A=A8=BA=A7=DA=A4S=A6=AC=B6i=A4F=B7s=BBO=B9=F42500=A4= =B8=A1A=ADY=B3o=A2=B1=A2=B4=AD=D3=A4H=A4]=A6U=A6=DB=B6K=A5X=A4F250=AB=CA=A4= =E5=B3=B9=A1A=A6=B9=AE=C9=A7=DA=AA=BA=A6W=A6r=A6b=A6W=B3=E6=A4=A4=AA=BA=B2= =C4=A4T=A6=EC=A1F=ADY=A4]=ACO=A5u=A6=B3=A4=AD=AD=D3=A4H=A6U=A6=DB=A6^=C0=B3= =A5L=AD=CC=AA=BA=A4=E5=B3=B9=A1A=A8=BA=A7=DA=B1N=A6A=A6=AC=A8=EC=A4F62500= =A4=B8=A1A=A6=B9=AE=C9=B3o125=A4H=A6b=A6U=A6=DB=B6K=A5X250=AB=CA=A4=E5=B3= =B9=A1A=A6=B9=AE=C9=A7=DA=AA=BA=A6W=A6r=A6b=B2=C4=A4G=A6=EC=A1F=A6P=BC=CB= =A5L=AD=CC=A8C=AD=D3=A4H=A4]=A6U=A6=DB=A6=AC=A8=EC=A2=B4=AD=D3=A6^=C0=B3=A1= A=A8=BA=A7=DA=A4S=B1N=A6=AC=A8=EC12500=A4=B8=A1C=A6P=B2z=A1A2625=A4H=A4]=A6= U=A6=DB=B6K=A5X250=BDg=A4=E5=B3=B9=A1A=A6=B9=AE=C9=A7=DA=AA=BA=A6W=A6r=A6= b=B2=C4=A4@=A6=EC=A1A=A8=BA=A7=DA=B1N=A6A=A6=AC=A4J315200=A4=B8=A1C=A6=D3= =A7=DA=B3=CC=AA=EC=AA=BA=A7=EB=B8=EA=A5u=A6=B3500=A4=B8=A1A=B3o=B9=EA=A6b= =ACO=AC=DB=B7=ED=A5O=A4H=A4=A3=A5i=AB=E4=C4=B3=A1C=A7=F3=A6=F3=AAp=B0=B2=B3= ]=A5u=A6=B3=A4=AD=AD=D3=A4H=A6^=C0=B3=AA=BA=A4=F1=A8=D2=A9=FA=C5=E3=B0=BE= =A7C=A1C=A5=AD=A7=A1=A8=D3=BB=A1=A4j=AC=F9=B7|=A6=B320~30=A4H=A6^=C0=B3=A1= A=B0=B2=B3]=ACO15=A4H=AA=BA=B8=DC=A1A=A7A=B1N=B7|=A6=AC=A8=EC=A6h=A4=D6=A1= H ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^= =A6W=A6r=A6b=B2=C4=A4=AD=A6=EC=AE=C9 2000=A4=B8 =A6W=A6r=A6b=B2=C4=A5|=A6=EC=AE=C9 40000=A4=B8 =A6W=A6r=A6b=B2=C4=A4T=A6=EC=AE=C9 800000=A4=B8 =A6W=A6r=A6b=B2=C4=A4G=A6=EC=AE=C9 1600000=A4=B8 =A6W=A6r=A6b=B2=C4=A4@=A6=EC=AE=C9 = =A6n=A6h=A1K=A1K=A6n=A6h=A1K=A1K=A4=B8 =A2w=A2w=A2w=A2w=A2w=A2w $$ = =B3o=ACO=A4@=AD=D3=A6X=AAk=C2=B2=A9=F6=A4S=A7=D6=B3t=AA=BA=B6=B0=B8=EA=A4= =E8=AAk $$=A1@=A2w=A2w=A2w=A2w=A2w=A2w =A1@=A1@=B7=ED=A7A=AA=BA=A6W=A6r=A4w=B8g=B1=BC=A5X=A6W=B3=E6=A4=A7=A5~=AE= =C9=A1A=A7A=B4N=B1q=B7s=BBD=B0Q=BD=D7=B0=CF=A4=A4=A6A=A7=E4=A4@=BDg=A4=E5= =B3=B9=A6^=C2=D0=A1A=A6A=A6=B8=B1N=A7A=AA=BA=A6W=A6r=A9=F1=A6b=A6W=B3=E6=A4= =A4=AA=BA=B2=C4=A4=AD=A6=EC=A1A=AD=AB=B7s=A6A=A8=D3=A4@=A6=B8=A1C=B7Q=B7Q= =AC=DD=A8C=A4=D1=B3=A3=A6=B3=A6=A8=A4d=A4W=B8U=AA=BA=A4H=A7=EB=A4J=BA=F4=BB= =DA=BA=F4=B8=F4=A4=A7=A4=A4=A1A=A6P=AE=C9=A8C=A4=D1=A4]=A6=B3=B3\=B3\=A6h= =A6h=AA=BA=A4H=A6b=BE\=C5=AA=B7s=BBD=B0Q=BD=D7=B0=CF=AA=BA=A4=E5=B3=B9=A1= A=A7=DA=B7Q=A7A=C0=B3=B8=D3=ACO=ADt=BE=E1=AA=BA=B0_=B3o=A2=B4=A2=AF=A2=AF= =A4=B8=AA=BA=A4p=A4p=A7=EB=B8=EA=A1A=A6p=AAG=A7A=C1=D9=A9=EA=AB=F9=B5=DB=C3= h=BA=C3=AA=BA=BAA=AB=D7=A1A=AC=B0=A4=B0=BB=F2=A4=A3=B8=D5=B8=D5=AC=DD=A9O= =A1H=A7A=AB=DC=A7=D6=B4N=B7|=AA=BE=B9D=B5=B2=AAG=A4F=A1C=B3o=ACO=A6=E6=AA= =BA=B3q=AA=BA=A1I=B3=CC=AB=E1=BD=D0=B1z=B0O=A6=ED=A1y=B8=DB=B9=EA=AC=B0=A4= W=A1z=AA=BA=AD=EC=ABh=A1A=BD=D0=B0=C8=A5=B2=B1N=BF=FA=B1H=B5=B9=A6W=B3=E6= =A4W=AA=BA=A4H=A1A=A7_=ABh=A7A=B1N=A4=A3=B7|=B1o=A8=EC=A4=D3=A6h=AA=BA=A6= ^=C0=B3=AA=BA=A1I=A6]=AC=B0=A6p=AAG=B1z=A4=A3=A7=E2=BF=FA=B1H=B5=B9=A6W=B3= =E6=A4W=AA=BA=A4H=A1A=ABo=A6b=A6W=B3=E6=A4=A4=A5[=A4J=B1z=AA=BA=A9m=A6W=A6= =D3=B1i=B6K=A6b=B7s=BBD=B8s=B2=D5=AA=BA=B8=DC=A1A=C5=FD=A8=BA=A8=C7=AD=EC= =A5=BB=B1z=C0=B3=B8=D3=A6^=F5X=B5=B9=A5L=AA=BA=A4H=A4W=BA=F4=B5o=B2{=A1A=AB= h=A9=D2=A6=B3=A4H=B1N=B7|=A6b=BA=F4=B8=F4=A4W=A4j=A4O=B5=FB=C0=BB=B1z=A1A= =A4=A3=A6=FD=C5=FD=B1z=B5L=AAk=A6=AC=A8=EC=BF=FA=A1A=A5t=A5~=A7=F3=A6W=C5= A=B1=BD=A6a=A1A=A9=D2=A5H=BD=D0=B1z=B8=DB=B9=EA=AC=B0=A4W=A1C=A8=C3=A5B=B5= =B9=B1z=A4@=AD=D3=B0J=A4=DF=AA=BA=AB=D8=C4=B3--=BD=D0=A4=A3=ADn=B5S=BF=DD= =A1I=B0=A8=A4W=A6=E6=B0=CA=A1I=ADY=A6]=B1z=BAC=A4@=AC=ED=C4=C1=C5=FD=A6=B3= =A4=DF=B0=D1=BBP=AA=CC=A5=FD=AC=DD=A8=EC=A5L=A4H=A6=AD=A4@=A8B=B1i=B6K=AA= =BA=A4=E5=B3=B9=A1A=A6=A8=AC=B0=A5L=A4H=B6l=BB=BC=A6W=B3=E6=A4=A4=AA=BA=A4= @=AD=FB=A1A=C1=F6=B5M=B5L=B7l=A6=DB=A8=AD=AA=BA=A7Q=AFq=A1A=A6=FD=B7Q=A8=D3= =C1`=A6=B3=A8=C7=BE=D2=B4o=A7a=A1I=A4=A3=ADn=BF=E9=A6b=B0_=B6]=C2I=A4W=A1= A=A7=D6=A8B=B8=F2=B6i=A7a=A1I=A6A=A6=B8=A1A=C1=C2=C1=C2=A7A=AD@=A4=DF=AC=DD= =A7=B9=A5=BB=A4=E5=A1A=AF=AC=A7A=A6n=B9B=A1I=A8=C3=B5=BD=A5=CE=B3o=A8=C7=A4= =C0=A8=C9=A9=D2=B1o=AA=BA=B8=EA=B7=BD=A1A=A6@=A6P=AC=B0=BBO=C6W=A6A=B3y=A5= t=A4@=AD=D3=C4=DD=A9=F3=A7=DA=AD=CC=A6=DB=A4v=BE=C4=B0=AB=AA=BA=A9_=C2=DD= ! ------=_NextPart_000_0131_01BE2BBA.EDD938A0 Content-Type: text/html; charset="big5" Content-Transfer-Encoding: quoted-printable <!DOCTYPE HTML PUBLIC "-//W3C//DTD W3 HTML//EN"> <HTML> <HEAD> <META content=3Dtext/html;charset=3Dbig5 http-equiv=3DContent-Type> <META content=3D'"MSHTML 4.72.3110.7"' name=3DGENERATOR> </HEAD> <BODY bgColor=3D#ffffff> <DIV> <DIV> <DIV> </DIV> <DIV> <DIV>=A7=DA=A4]=A6=AC=A8=EC=BF=FA=A4F=A1I<BR>=A7A=A9=CE=B3\=B7|=A5H=AC=B0= =B3o=A5u=ACO=AD=D3=C4F=A4H=AA=BA=BE=BB=C0Y=A1A=A5=FD=ABe=A7=DA=AC=DD=A8=EC= =C3=FE=A6=FC=AA=BA=ABH=A5=F3=AE=C9=A1A=A4]=B4=BF=A6=B3=B3o=BA=D8=B7Q=AAk=A1= A=A6=FD=A4p=A4p=AA=BA=A7=EB=B8=EA=C3B=A4=DE=B0_=A7=DA=AA=BA=BF=B3=BD=EC=A1= A=B4N=BA=E2=BD=DF=A4F=A4]=A5u=ACO500=A4=B8=A5=E1=A5X=A5h=B6R=B1=D0=B0V=A1= C<BR>=A9=EA=B5=DB=A4S=B4=C1=AB=DD=A4S=A9=C8=A8=FC=B6=CB=AE`=AA=BA=A4=DF=B1= =A1=A1A=A7=DA=A6U=B1H100=A4=B8=B5=B9=ABH=A5=F3=A6W=B3=E6=A4=A4=AA=BA=A4=AD= =A6W=A6=AC=A5=F3=AA=CC=A1A=B1=B5=A4U=A8=D3=AA=BA=A8=E2=A4T=A4=D1=A1A=A7=DA= =B6}=A9l=A6b=A6U=AD=D3=B7s=BBD=B0Q=BD=D7=B0=CF=B5o=B3o=AB=CA=ABH=A1A=B2=C4= =A4@=AD=D3=C2=A7=AB=F4=A7=DA=AAG=AFu=A6=AC=A8=EC=BF=FA=A4F=A1A=A7=DA=A6=AC= =A8=EC=A4K=AB=CA=B8=CB=A6=B3100=A4=B8=B2{=AA=F7=AA=BA=ABH=A1A=B5=A5=A9=F3= =BB=A1=A1A=A7=DA=A4w=B8g=C1=C8=A6^=A7=DA=A9=D2=A7=EB=B8=EA=AA=BA=AA=F7=C3= B=A1A=C1=D9=A5~=A5[300=A4=B8=AA=BA=A6=AC=A4J=A1C<BR>=B2=C4=A4G=AD=D3=C2=A7= =AB=F4=AB=E1=A1A=A7=DA=AA=BA=A6=AC=A4J=A7=F3=C5=E5=A4H=A1A=B3=BA=B5M=A6=AC= =A8=EC=A4G=A4Q=A4=AD=AB=CA=AA=BA=ABH=A1]=A4]=B4N=ACO2500=A1^=A1A=A7=DA=A9= =F3=ACO=A5=B4=C5K=B6X=BC=F6=A6b=B5o=A5X=B3\=A6h=AA=BA=ABH=A5=F3=A6b=BA=F4= =B8=F4=A4W=A1A=A7A=AC=DB=ABH=B6=DC=A1H<BR>=A6b=A4@=AD=D3=A4=EB=AB=E1=A7=DA= =A5=CE500=A4=B8=C1=C8=B6i=A4F27500=A4=B8=A1A=A7=EB=B8=EA=A9=EA=B9S=B2v=A4= =A7=B0=AA=A1A=B3s=AA=D1=B2=BC=B3=A3=B1=E6=A4=A7=B2=F6=A4=CE=A1C<BR>=AB=E1= =A8=D3=A4S=C2_=C2_=C4=F2=C4=F2=A6=AC=A8=EC=BF=FA=A1A=A5=D8=ABe=A4w=B8g=B2= =D6=BFn=AA=F1=A4Q=A6=EC=BC=C6=A1A=A5=BB=A8=D3=A6]=AC=B0=A4=F7=A5=C0=B8g=B0= =D3=A5=A2=B1=D1=A6=D3=A4T=C0\=B3=A3=A6=A8=B0=DD=C3D=AA=BA=A7=DA=A1A=B2{=A6= b=BE=C7=B6O=A9M=A5=CD=AC=A1=B6O=A5=FE=A6=B3=B5=DB=B8=A8=A4F=A1I<BR>=C5=A5= =A7=DA=BB=A1=A7=B9=A6=DB=A4v=AA=BA=BF=CB=A8=AD=B8g=BE=FA=A1A=A7A=A4@=A9w=A4= ]=C5D=C5D=B1=FD=B8=D5=BD}=A1I<BR>=B3o=BA=D8=A7=EB=B8=EA=A4=E8=AAk=AB=DC=C2= =B2=B3=E6=A1A=AD=BA=A5=FD=A7A=A5=B2=B6=B7=AD@=B5=DB=A9=CA=A4l=AC=DD=A7=B9= =A7=DA=A7Y=B1N=A7i=B6D=A7A=AA=BA=A8=E2=A5=F3=A8=C6=A1]=B3=CC=A6n=A7=E2=A5= =A6=A6C=A6L=A5X=A8=D3=A1^=A1G<BR>=A2=B0.=B3o=AD=D3=ADp=B5e=AA=BA=B9=EA=A6= =E6=A4=E8=AAk=A1C<BR>=A2=B1.=B3o=AD=D3=ADp=B5e=A5i=A6=E6=AA=BA=AD=EC=A6]=A1= C<BR>1.=B3o=AD=D3=ADp=B5e=AA=BA=B9=EA=A6=E6=A4=E8=AAk<BR> =20 =A2=CF.=B7=C7=B3=C6=A7=F7=AE=C6=A1G=A4=AD=AD=D3=BC=D0=B7=C7=ABH=AB=CA=A1B= =A4=AD=B1i=A4=AD=A4=B8=AA=BA=B6l=B2=BC=A1B=A4=AD=B1i=C3C=A6=E2=AB=DC=B2`=AA= =BA=AF=C8<BR> =20 =A1]=ADn=AF=E0=A7=E2=A2=B0=A2=AF=A2=AF=A4=B8=A5]=A6=ED=A1A=B5L=AAk=AC=DD=A8= =A3=AA=BA=B2`=A6=E2=A1^=A1B=A4=AD=B1i=ABK=B1=F8=AF=C8=A1C<BR> =20 =A2=D0.=A4=E8=AAk=A1G=A6b=A8C=B1i=ABK=B1=F8=AF=C8=A4W=BCg=B5=DB=A1y=BD=D0= =B1N=A7=DA=A9=F1=A4J=A7A=AA=BA=B6l=BB=BC=A6W=B3=E6=A4=A4=A1z=A1F=B1N=A2=B0= =A2=AF=A2=AF=A4=B8=A5=CE=B2`=A6=E2<BR>=AF=C8=A5]=B0_=A8=D3=A1A  = ; =20 =A4=C0=A7O=A9=F1=A4J=A8C=AD=D3=ABH=AB=CA=A4=A4=A1]=A4@=A9w=ADn=C5=FD=A2=B0= =A2=AF=A2=AF=A4=B8=B5L=AAk=B3Q=B3z=B5=F8=A1A=A7_=ABh=AEe=A9=F6=B3Q=A4=A3=A8= v=A4=A7=AE{=AE=B3=A8=AB=A1^=A1C<BR> =20 =A2=D1.=B1N=B3o=A4=AD=AB=CA=ABH=A4=C0=A7O=B1H=B5=B9=A5H=A4U=AA=BA=A4=AD=AD= =D3=A4H=A1]=AA`=B7N=A1G=A4=A3=A5i=A5H=C0H=ABK=B1H=A1^=A1G<BR> = 1.=B6=C0=A7g=B0=B6 =20 =A5x=A4=A4=A5=AB=A5_=A4=D9=B0=CF=A6Z=C9=DC=B8=F4128=B8=B97=BC=D3 &nb= sp; =20 =B6l=BB=BC=B0=CF=B8=B9 406<BR> 2. = =AAL=B4=AD=AA@ =20 =B0=AA=B6=AF=BF=A4=A9=A3=A4s=C2=ED=A9=A3=A4s=B8=F4287=B8=B9 &n= bsp; &nb= sp; =20 =B6l=BB=BC=B0=CF=B8=B9820<BR> = 3.=A7=F5=A9y=AF=D5 =20 =A5x=A5_=A5=AB=AAQ=A6=BF=B8=F4510=B8=B93=BC=D3 &nb= sp; &nbs= p; =20 =B6l=BB=BC=B0=CF=B8=B9104<BR> = 4.=AAL=A9=FA=BC=DD=20 =A5x=A4=A4=BF=A4=AFQ=A4=E9=B6m=A5_=A8=BD=A7=F8=B7=CB=ABn=B8=F463-2=B8=B9&= nbsp; =B6l=BB=BC=B0=CF=B8=B9 414</DIV> <DIV> <FONT color=3D#000000 size=3D2> = <FONT=20 size=3D3>5.=B6=C0=AB=B6=BD=F7 =20 =AE=E7=B6=E9=BF=A4=A4K=BCw=A5=AB=A9M=A5=AD=B8=F4621=AB=D149=B8=B9 &n= bsp; =20 =B6l=BB=BC=B0=CF=B8=B9334</FONT></FONT></DIV> <DIV> = (=B0O=B1o=BCg=B6l=BB=BC=B0=CF=B8=B9 = =A7_=ABh=B6l=A7=BD=B1N=A9=B5=AA=F8=A4u=A7@=A4=D1=B3=E1!)<BR> =20 =A2=D2.=A7=F3=A7=EF=A6a=A7}=A1G<BR>=A1@=A1@=B1H=A7=B9=ABH=AB=E1=A1A=A6A=B3= z=B9L=B9q=B8=A3=B1N=A4W=AD=B1=B2=C4=A4@=AD=D3=A4H=AA=BA=A6W=A6r=BD=F0=B1=BC= =A1A=A8=C3=A8=CC=A7=C7=B1N=A5|=AD=D3=A4H=AA=BA=A6W=A6r=A9=B9=A4W=B2=BE=A4= G=B2=BE=A8=EC=A4@=A1F=A4T=B2=BE=A8=EC=A4G=A1F=A5|=B2=BE=A8=EC=A4T=A1F=A4=AD= =B2=BE=A8=EC=A5|=A1^=A1A=A6=D3=B2=C4=A4=AD=AD=D3=BCg=A4W=A7A=AA=BA=A6W=A6= r=A1B=A6a=A7}=A1A=ADn=AA`=B7N=A6a=A7}=ACO=A5=BF=BDT=AA=BA=A1A=A7_=ABh=A7A= =B7|=A6=AC=A4=A3=A8=EC=ABH=B3=E1=A1I<BR> =20 =A2=D3.=A5=E1=A4J=B7s=BBD=B8s=B2=D5=A1G=A4=E5=B3=B9=A4=A4=B6=C8=BB=DD=AD=D7= =A7=EF=A1u=A2=D2=A1B=A7=F3=A7=EF=A6a=A7}=AA=BA=B3=A1=A5=F7=A1v=A9=D2=B4=A3= =A8=EC=AA=BA=B3=A1=A5=F7=A1A=ABe=AD=B1=A9=D2<BR>=BB=A1=AA=BA=B8g=C5=E7=B3= =A1=A5=F7=A1A=A7A=A5i=A5H=AE=DA=BE=DA=A6=DB=A4v=AA=BA=B9=EA=AAp=B0=B5=C2=B2= =A9=F6=A7=F3=A7=EF=A1A=A8=E4=A5L=AA=BA=B3=A1=A5=F7=B3=CC=A6n=ABO=AB=F9=AD= =EC=AA=AC=A1A=B5M=AB=E1=A7A=A5=B2=B6=B7=B1N=ABH=A5=F3=A6=DC=A4=D6=A4=BD=A7= G=A6b250=AD=D3=B7s=BBD=B0Q=BD=D7=B0=CF=A1]=B3=CC=A4=D6=ADn=A6=B3=A8=E2=A6= =CA=AD=D3=B0Q=BD=D7=B0=CF=A1A=A5=FE=A5@=AC=C9=AC=F9=A6=B325000=AD=D3=B7s=BB= D=B0Q=BD=D7=B0=CF=A1C)<BR>=A1=AF=AA`=B7N=A1G=A7A=A9=D2=A4=BD=A7G=A5X=A5h=AA= =BA=B7s=BBD=B0=CF=B6V=A6h=A1A=B1z=B1o=A8=EC=AA=BA=A6^=F5X=B6V=A6h=A1C<BR>= =ADY=A4=A3=AA=BE=A6p=A6=F3=A4=BD=A7G=A8=EC=B7s=BBD=B0=CF=A1A=B1z=A5i=A5H=B7= =D3=A4U=A6C=AA=BA=A4=E8=AAk=A7@=A1G<BR>=A1=D7=A1=D7=A1=D7=A1=D7=A8=CF=A5=CE= =BA=F4=B4=BA=A2=DC=A2=D3=A2=E2=A2=E1=A2=D1=A2=CF=A2=DE=A2=D3=A1@=A2=DC=A2= =CF=A2=E4=A2=D7=A2=D5=A2=CF=A2=E2=A2=DD=A2=E0=AA=BA=A8=CF=A5=CE=AA=CC=A1=D7= =A1=D7=A1=D7=A1=D7<BR>=A2=B0.=20 =A5=FD=A5=B4=B6}NETSCAPE = NEWS=A1A=A6A=A6b=BA=F4=B4=BA=C2s=C4=FD=BE=B9=A4=A4=B1q=A5\=AF=E0=BF=EF=B3= =E6=A4=A4=A5=FD=BF=EF=A8=FAWindows=A1A=B1=B5=B5=DB=A6A=BF=EF=A8=FANETSCAP= E NEWS=A1A=A7Y=A5i=B1=D2=B0=CANETSCAPE=20 NEWS=A1C<BR>=A2=B1. =C2I=A8=FA=B3=CC=A5=AA=C3=E4=AA=BA=AB=F6=B6s=A1yPOST = NEWS=A1z=A6=B9=AE=C9=B7|=A5X=B2{=A4@=AD=D3=BDs=BF=E8=B7s=A4=E5=B3=B9=AA=BA= =B5=F8=B5=A1=A1C<BR>=A2=B2.=20 =AC=B0=A7A=AA=BA=A4=E5=B3=B9=BF=EF=A4@=AD=D3=A7l=A4=DE=A4H=AA=BA=BC=D0=C3= D=A1A=B1N=A7A=A5=D8=ABe=AC=DD=A8=EC=AA=BA=B3o=BDg=A4=E5=B3=B9COPY=B5M=AB=E1= PASTE=A8=EC=B0T=AE=A7=C4=E6=A4=BA=A1A=B7=ED=B5M=A7A=A5i=A5H=BEA=20 =B7=ED=AA=BA=AD=D7=A7=EF=B3o=BDg=A4=E5=B3=B9=A1]=A7A=A5=B2=B6=B7=B1N=A7A=AA= =BA=A6W=A6r=A6a=A7}=A5[=A6b=A4=E5=B3=B9=AA=BA=B2=C4=A4=AD=A6=EC=A1A=A8=C3= =B1N=B3=D1=A4U=A5|=AD=D3=A4H=AA=BA=A6W=A6r=A6a=A7}=A9=B9=A4W=B5=A5=B2=BE=A4= @=A6=EC=A1^=A1C=B7=ED=A7A=A7=E2=BE=E3=BDg=A4=E5=B3=B9=A5=B4=A6n=A9=CECOPY= =A4=A7=AB=E1,=C2I=BF=EF=A1ySEND=20 NOW=A1z.=B2{=A6b=B6}=A9l=B3=B0=C4=F2=B6K=A9=B9=A4=A3=A6P=AA=BA=B7s=BBD=B0= Q=BD=D7=B0=CF=A7a=A1I<BR>=A1=D7=A1=D7=A1=D7=A1=D7 = =A8=CF=A5=CE=B7L=B3n=AA=BA=B1=B4=C0I=AEa=C2s=C4=FD=BE=B9 Microsoft = Internet=20 Explorer=A1=D7=A1=D7=A1=D7=A1=D7<BR>=A2=B0.=A4=E8=AAk=A4]=ACO=A4@=BC=CB=AA= =BA=C2=B2=B3=E6=A1A=AB=F6=A6=ED=B7=C6=B9=AB=A5=AA=C1=E4=A4=A3=A9=F1=A1A=B1= N=BE=E3=AD=D3=A4=E5=B3=B9=A4=CF=A5=D5=B0_=A8=D3=A1A=B5M=AB=E1=A6P=AE=C9=AB= =F6=B5=DB<BR>=A1yCTRL=A1z=A4=CE=A1yC=A1z=C1=E4=B1N =20 =BE=E3=BDg=A4=E5=B3=B9=AB=FE=A8=A9=A8=EC=B0=C5=B6K=C3=AF=A1A=A6=B9=AE=C9=A6= A=B1N=A5=BB=A4=E5=A6C=A6L=A5X=A8=D3=A1A=B3o=BC=CB=A7A=B4N<BR>=A6=B3=A4@=A5= =F7=A7A=B4=BF=B8g=B1H=A5X=AA=BA=A6W=B3=E6=A4F=A1C=A1]=B0O=B1o=B1N=A7A=A6=DB= =A4v=AA=BA=A6W=A6r=A6a=A7}=A5[=A6b=A6W=B3=E6=AA=BA=B3=CC=AB=E1=A1I=A1^<BR= >=A2=B1.=20 =B3o=AE=C9=AD=D4=A6A=A6=B8=B1N=A4=E5=B3=B9=A4=CF=A5=D5=B0_=A8=D3=A1A=AD=AB= =B7sCOPY=A4@=A6=B8=A1A=A5H=BDT=BB{=A7A=A4w=B8gCOPY=A4F=AD=D7=A5=BF=B9L=AA= =BA=A4=E5=B3=B9=A4=BA=AEe=A1C=B3o=AE=C9=AD=D4=A7A=A5u=BB=DD=ADn=A6b=A7A=B7= =C7=B3=C6=B6K=A4=E5=B3=B9=AA=BA=B0Q=BD=D7=B0=CF=BF=EF=BE=DC=B5o=AA=ED=B7s= =AA=BA=A4=E5=B3=B9=A1A=B5M=AB=E1=A5HCTRL+V=A1z=B1N=A4=E5=B3=B9=A4@=B0_=B6= K(PASTE)=A5X=A1C<BR>=A1@=A1@=B4N=ACO=B3o=BB=F2=C2=B2=B3=E6=A1A=A7A=A5u=BB= =DD=ADn=BF=EF=BE=DC=A4=A3=A6P=AA=BA=B7s=BBD=B0Q=BD=D7=B0=CF=A1A=B5M=AB=E1= =BF=EF=BE=DC=A1yPOST=A1z=B5o=AA=ED=A4=E5=B3=B9=B5=A5=A7A=B2=DF=BAD=BE=E3=AD= =D3=B9L=B5{=AB=E1=A1A=A8C=B5o=A4@=AB=CA=ABH=A4j=AC=F9=A5u=BB=DD=ADn30=AC=ED= =AA=BA=AE=C9=B6=A1=A9O=A1I<BR>=A1@=A1@=A5=D1=A9=F3=A5=BB=A4=E5=AC=B0=A4=A4= =A4=E5=A1A=ACG=BE=A8=B6q=B5o=A4=A4=A4=E5=B0=CF=B8s=B2=D5=B8=FB=A6=B3=AE=C4= =A1C=A7A=A5i=A5H=C2I=BF=EF=A1e=B7s=BBD=A1f=A1=D0=A1e=BF=EF=BE=DC=B7s=BBD=B8= s=B2=D5=A1f=A1A=A6=B9=AE=C9=A5X=B2{=B3\=A6h=B0Q=BD=D7=B0=CF=A1A=B1z=A5i=A5= H=A6b=A4W=A4=E8=BF=E9=A4J=B7j=B4M"ROC"=20 =A1A=20 "TAIWAN"=B5=A5=C3=F6=C1=E4=A6r=A1A=ABh=A9=D2=A8=A3=A6C=A5X=AA=BA= =B4N=ACO=A5x=C6W=AC=DB=C3=F6=AA=BA=B0Q=BD=D7=B0=CF=A1A=B3o=AE=C9=B4N=A8=CC= =B1z=AD=D3=A4H=BB=DD=ADn=ACD=BF=EF=A8=C3=B1i=B6K=A4=E5=B3=B9=A4F=A1C=BD=D0= =B0O=A6=ED=A1A=A7A=B6K=AA=BA=B7s=BBD=B0Q=BD=D7=B0=CF=B6V=A6h=A1A=AC=DB=B9= =EF=A6a=A7A=B4N=B7|=B1o=A8=EC=B6V=A6h=AA=BA=A6^=C0=B3=A1C=A6P=AE=C9=A7A=A4= ]=B7|=B1o=A8=EC=B6V=A6h=AA=BA=A6^=C5T=A1I=B4N=ACO=B3o=BB=F2=C2=B2=B3=E6=A1= A=A7A=B1N=A6=AC=A8=EC=A8=D3=A6=DB=A5=FE=BBO=C6W=A6U=A6a=A4D=A6=DC=A9=F3=A5= =FE=B2y=A6U=A6a=AA=BA=A6^=ABH,=A7=C6=B1=E6=AF=E0=A5[=A4J=A7A=AA=BA"=B6= l=BB=BC=A6W=B3=E6"!=A1]=B7=ED=B5M,=BD=D0=B0=C8=A5=B2=BDT=BB{=B1H=B0e= =A6a=A7}=AA=BA=A5=BF=BDT!=A1^=A1@=A1@=A1@=A1@<BR>2.=B3o=AD=D3=ADp=B5e=A5i= =A6=E6=AA=BA=AD=EC=A6]<BR> =20 =A6b=A7A=B5o=A5X250=AB=CA=A4=E5=B3=B9=A4=A4=A1A=B0=B2=B3]=A5u=A6=B3=A4=AD= =AD=D3=A4H=A6^=C0=B3=A7A=AA=BA=A4=E5=B3=B9=A1]=B3o=AD=D3=C1|=A6C=A5=CE=AA= =BA=A4=F1=A8=D2=AC=DB<BR>=B7=ED=A7C=A1^=A1A=B7=ED=A7=DA=AA=BA=A6W=A6r=A6b= =B2=C4=A4=AD=A6=EC=AE=C9=A7=DA=C1=C8=B6i=A4F250=A4=B8=A1C=B2{=A6b=A6^=C0=B3= =A7=DA=A4=E5=B3=B9=AA=BA=A4=AD=AD=D3=A4H=A1A=A4]=B4N=ACO=B1H=B5=B9=A7=DA=B7= s=BBO=B9=F4100=A4=B8=AA=BA=B3o=A4=AD=AD=D3=A4H=A4S=A6U=A6=DB=A6=DC=A4=D6=B6= K=A5X250=AB=CA=A4F=A4=E5=B3=B9=A1A=A6=B9=AE=C9=A7=DA=AA=BA=A6W=A6r=B1N=A4= W=B2=BE=A6=DC=B2=C4=A5|=A6=EC=A1F=B0=B2=B3]=A4]=A5u=A6=B3=A4=AD=AD=D3=A4H= =A6U=A6=DB=A6^=C0=B3=B5=B9=A5L=AD=CC=B3o=A4=AD=AD=D3=A4H=A1]=A4]=B4N=ACO=B3= =CC=A6=AD=B1H=BF=FA=B5=B9=A7=DA=AA=BA=A4H=A1^=A1A=A8=BA=A7=DA=A4S=A6=AC=B6= i=A4F=B7s=BBO=B9=F42500=A4=B8=A1A=ADY=B3o=A2=B1=A2=B4=AD=D3=A4H=A4]=A6U=A6= =DB=B6K=A5X=A4F250=AB=CA=A4=E5=B3=B9=A1A=A6=B9=AE=C9=A7=DA=AA=BA=A6W=A6r=A6= b=A6W=B3=E6=A4=A4=AA=BA=B2=C4=A4T=A6=EC=A1F=ADY=A4]=ACO=A5u=A6=B3=A4=AD=AD= =D3=A4H=A6U=A6=DB=A6^=C0=B3=A5L=AD=CC=AA=BA=A4=E5=B3=B9=A1A=A8=BA=A7=DA=B1= N=A6A=A6=AC=A8=EC=A4F62500=A4=B8=A1A=A6=B9=AE=C9=B3o125=A4H=A6b=A6U=A6=DB= =B6K=A5X250=AB=CA=A4=E5=B3=B9=A1A=A6=B9=AE=C9=A7=DA=AA=BA=A6W=A6r=A6b=B2=C4= =A4G=A6=EC=A1F=A6P=BC=CB=A5L=AD=CC=A8C=AD=D3=A4H=A4]=A6U=A6=DB=A6=AC=A8=EC= =A2=B4=AD=D3=A6^=C0=B3=A1A=A8=BA=A7=DA=A4S=B1N=A6=AC=A8=EC12500=A4=B8=A1C= =A6P=B2z=A1A2625=A4H=A4]=A6U=A6=DB=B6K=A5X250=BDg=A4=E5=B3=B9=A1A=A6=B9=AE= =C9=A7=DA=AA=BA=A6W=A6r=A6b=B2=C4=A4@=A6=EC=A1A=A8=BA=A7=DA=B1N=A6A=A6=AC= =A4J315200=A4=B8=A1C=A6=D3=A7=DA=B3=CC=AA=EC=AA=BA=A7=EB=B8=EA=A5u=A6=B35= 00=A4=B8=A1A=B3o=B9=EA=A6b=ACO=AC=DB=B7=ED=A5O=A4H=A4=A3=A5i=AB=E4=C4=B3=A1= C=A7=F3=A6=F3=AAp=B0=B2=B3]=A5u=A6=B3=A4=AD=AD=D3=A4H=A6^=C0=B3=AA=BA=A4=F1= =A8=D2=A9=FA=C5=E3=B0=BE=A7C=A1C=A5=AD=A7=A1=A8=D3=BB=A1=A4j=AC=F9=B7|=A6= =B320~30=A4H=A6^=C0=B3=A1A=B0=B2=B3]=ACO15=A4H=AA=BA=B8=DC=A1A=A7A=B1N=B7= |=A6=AC=A8=EC=A6h=A4=D6=A1H<BR>^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^= ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^<BR>=A6W=A6r=A6b=B2=C4=A4=AD=A6=EC=AE=C9=20 2000=A4=B8<BR>=A6W=A6r=A6b=B2=C4=A5|=A6=EC=AE=C9 = 40000=A4=B8<BR>=A6W=A6r=A6b=B2=C4=A4T=A6=EC=AE=C9 = 800000=A4=B8<BR>=A6W=A6r=A6b=B2=C4=A4G=A6=EC=AE=C9 = 1600000=A4=B8<BR>=A6W=A6r=A6b=B2=C4=A4@=A6=EC=AE=C9=20 =A6n=A6h……=A6n=A6h……=A4=B8<BR>=A2w=A2w=A2w=A2= w=A2w=A2w $$ = =B3o=ACO=A4@=AD=D3=A6X=AAk=C2=B2=A9=F6=A4S=A7=D6=B3t=AA=BA=B6=B0=B8=EA=A4= =E8=AAk=20 $$=A1@=A2w=A2w=A2w=A2w=A2w=A2w<BR>=A1@=A1@=B7=ED=A7A=AA=BA=A6W=A6r=A4w=B8= g=B1=BC=A5X=A6W=B3=E6=A4=A7=A5~=AE=C9=A1A=A7A=B4N=B1q=B7s=BBD=B0Q=BD=D7=B0= =CF=A4=A4=A6A=A7=E4=A4@=BDg=A4=E5=B3=B9=A6^=C2=D0=A1A=A6A=A6=B8=B1N=A7A=AA= =BA=A6W=A6r=A9=F1=A6b=A6W=B3=E6=A4=A4=AA=BA=B2=C4=A4=AD=A6=EC=A1A=AD=AB=B7= s=A6A=A8=D3=A4@=A6=B8=A1C=B7Q=B7Q=AC=DD=A8C=A4=D1=B3=A3=A6=B3=A6=A8=A4d=A4= W=B8U=AA=BA=A4H=A7=EB=A4J=BA=F4=BB=DA=BA=F4=B8=F4=A4=A7=A4=A4=A1A=A6P=AE=C9= =A8C=A4=D1=A4]=A6=B3=B3\=B3\=A6h=A6h=AA=BA=A4H=A6b=BE\=C5=AA=B7s=BBD=B0Q=BD= =D7=B0=CF=AA=BA=A4=E5=B3=B9=A1A=A7=DA=B7Q=A7A=C0=B3=B8=D3=ACO=ADt=BE=E1=AA= =BA=B0_=B3o=A2=B4=A2=AF=A2=AF=A4=B8=AA=BA=A4p=A4p=A7=EB=B8=EA=A1A=A6p=AAG= =A7A=C1=D9=A9=EA=AB=F9=B5=DB=C3h=BA=C3=AA=BA=BAA=AB=D7=A1A=AC=B0=A4=B0=BB= =F2=A4=A3=B8=D5=B8=D5=AC=DD=A9O=A1H=A7A=AB=DC=A7=D6=B4N=B7|=AA=BE=B9D=B5=B2= =AAG=A4F=A1C=B3o=ACO=A6=E6=AA=BA=B3q=AA=BA=A1I=B3=CC=AB=E1=BD=D0=B1z=B0O=A6= =ED=A1y=B8=DB=B9=EA=AC=B0=A4W=A1z=AA=BA=AD=EC=ABh=A1A=BD=D0=B0=C8=A5=B2=B1= N=BF=FA=B1H=B5=B9=A6W=B3=E6=A4W=AA=BA=A4H=A1A=A7_=ABh=A7A=B1N=A4=A3=B7|=B1= o=A8=EC=A4=D3=A6h=AA=BA=A6^=C0=B3=AA=BA=A1I=A6]=AC=B0=A6p=AAG=B1z=A4=A3=A7= =E2=BF=FA=B1H=B5=B9=A6W=B3=E6=A4W=AA=BA=A4H=A1A=ABo=A6b=A6W=B3=E6=A4=A4=A5= [=A4J=B1z=AA=BA=A9m=A6W=A6=D3=B1i=B6K=A6b=B7s=BBD=B8s=B2=D5=AA=BA=B8=DC=A1= A=C5=FD=A8=BA=A8=C7=AD=EC=A5=BB=B1z=C0=B3=B8=D3=A6^=F5X=B5=B9=A5L=AA=BA=A4= H=A4W=BA=F4=B5o=B2{=A1A=ABh=A9=D2=A6=B3=A4H=B1N=B7|=A6b=BA=F4=B8=F4=A4W=A4= j=A4O=B5=FB=C0=BB=B1z=A1A=A4=A3=A6=FD=C5=FD=B1z=B5L=AAk=A6=AC=A8=EC=BF=FA= =A1A=A5t=A5~=A7=F3=A6W=C5A=B1=BD=A6a=A1A=A9=D2=A5H=BD=D0=B1z=B8=DB=B9=EA=AC= =B0=A4W=A1C=A8=C3=A5B=B5=B9=B1z=A4@=AD=D3=B0J=A4=DF=AA=BA=AB=D8=C4=B3--=BD= =D0=A4=A3=ADn=B5S=BF=DD=A1I=B0=A8=A4W=A6=E6=B0=CA=A1I=ADY=A6]=B1z=BAC=A4@= =AC=ED=C4=C1=C5=FD=A6=B3=A4=DF=B0=D1=BBP=AA=CC=A5=FD=AC=DD=A8=EC=A5L=A4H=A6= =AD=A4@=A8B=B1i=B6K=AA=BA=A4=E5=B3=B9=A1A=A6=A8=AC=B0=A5L=A4H=B6l=BB=BC=A6= W=B3=E6=A4=A4=AA=BA=A4@=AD=FB=A1A=C1=F6=B5M=B5L=B7l=A6=DB=A8=AD=AA=BA=A7Q= =AFq=A1A=A6=FD=B7Q=A8=D3=C1`=A6=B3=A8=C7=BE=D2=B4o=A7a=A1I=A4=A3=ADn=BF=E9= =A6b=B0_=B6]=C2I=A4W=A1A=A7=D6=A8B=B8=F2=B6i=A7a=A1I=A6A=A6=B8=A1A=C1=C2=C1= =C2=A7A=AD@=A4=DF=AC=DD=A7=B9=A5=BB=A4=E5=A1A=AF=AC=A7A=A6n=B9B=A1I=A8=C3= =B5=BD=A5=CE=B3o=A8=C7=A4=C0=A8=C9=A9=D2=B1o=AA=BA=B8=EA=B7=BD=A1A=A6@=A6= P=AC=B0=BBO=C6W=A6A=B3y=A5t=A4@=AD=D3=C4=DD=A9=F3=A7=DA=AD=CC=A6=DB=A4v=BE= =C4=B0=AB=AA=BA=A9_=C2=DD!<BR><BR><BR></DIV></DIV></DIV></DIV></BODY></HT= ML> ------=_NextPart_000_0131_01BE2BBA.EDD938A0--
From: "Wiro van Schaik" <ourxs@xs4all.nl> Newsgroups: comp.sys.next.programmer Subject: Problem: Spontenous death notifications on just one side of NXConnection Date: Sat, 19 Dec 1998 23:28:05 +0100 Organization: XS4ALL, networking for the masses Sender: ourxs@asp54-59.amsterdam.nl.net Message-ID: <75h97p$kq8$1@news2.xs4all.nl> Cross-posted to: mailinglist next-prog@omnigroup.com L.S. I have a problem with a NeXTSTEP 3.3 system which uses Distributed Objects. The system consist of a server part, who receives message through a socket interface and distributes these to a number of clients through a Distributed Objects connection. The server also receives messages from the clients to send through a socket connection to another process as well as distribute these messages again to the other clients. The server actually acts as a notification center to keep the information displayed by the clients in sync with each other as well as with another (Non-NeXTSTEP) process running on a HP-UX machine (hence the socket connection). The problems is that I sometimes get port death notifications on just one side of a NXConnection. For instance, the server would receive a death notification for a connection to a client and consequently clear all references to the client. The client however doesn't get a death notification and waits forever for messages from the server. The same problem but then the other way around also occurred. Some additional clues which might ring a bell with someone: Most clients have a 2Mbit leased line connection with the server. The problems indeed occurs much more frequently with these clients then with clients connected through a regular 10Mb Ethernet. The clients also have to communicate with a database server across this same connection using EOF. The loads over this connection however don't seem to be all that heavy (database queries have good response times and aren't generally that heavy in the amount of data to transfer) The problem occurs most often about 1 minute after the client has started and registered at the server. First I thought it might have something to do with the number of NXConnections as they don't seem to be freed after a client correctly deregisters. At a certain moment in time the problem seem to occur when more then 32 NXConnections were used during the lifetime of the server. But recently the problem occurred after more then 60 NXConnections had been used. Both server and client are single threaded. They have been built using NeXTSTEP 3.3, EOF 1.1, with the latest patches for these version applied. We use both HP 715 workstations as well as HP Pentium PC's. All messages (except the registration message) are asynchronous. They range from 4 to about 60 bytes per message. There are no problems when a client or a server dies itself. The death notification is correctly received and handled on the other end (apart from the fact that the NXConnection doesn't seem to be freed). I've tried enabling the DO debugging but that gave me no clue. I don't have clue were to look for next. If anybody out there could give me any hints on issues I should check, I would but ultimately grateful. If anybody would need more information before they would be able to give some answer, just let me know and I will do my best to provide it. Regretfully, I probably can't publish a lot of source code as it is not our code but from a customer we are trying to help out. Thanks beforehand, Wiro van Schaik, Consulting Software Engineer /* * Shared Objectives bv - Bringing engineering to software * * Transpolis Schiphol Airport, Polaris Avenue 53 * 2132 JH Hoofddorp, The Netherlands * Tel.: +31 (0)23 5685788, Fax: +31 (0)23 5685701 * WWW: <http://www.shared-objectives.com> */
From: ecarter@eric.acs.uci.edu (Eric Carter) Newsgroups: comp.sys.next.programmer Subject: [Q] WebObjects vs ColdFusion Date: 19 Dec 1998 22:34:00 GMT Organization: University of California, Irvine Message-ID: <75h9ko$sj2@news.service.uci.edu> Hi There... awhile back I posted a question about experiences with webobjects. A couple of people let me know they would highly recommend it. However, where I'm at some have suggested ColdFusion over WebObjects. Can anyone relate some differences between these two products? Would there be a significant advantage to use one over another? Any responses HIGHLY appreciated. ecarter@uci.edu
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <13296913525221@digifix.com> Date: 20 Dec 1998 04:45:18 GMT Organization: Digital Fix Development Message-ID: <6039914130021@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1994. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
Subject: Test first, develop later: Sen:te has released OCUnit Newsgroups: comp.sys.next.programmer From: marco@sente.ch.mil (Marco Scheurer) Message-ID: <367d3b1d.0@epflnews.epfl.ch> Date: 20 Dec 1998 18:59:57 +0100 Organization: EPFL Test first, develop later. Sen:te releases OCUnit. OCUnit is a testing framework for Objective C in the MacOSX/Yellow Box environment. It is based on SUnit, Kent Beck's Smalltalk testing framework, also available for Java under the name JUnit. With OCUnit, testing becomes integrated with development. You can test frameworks, bundles, or applications. You can now tell "when something starts working or when something stops working [...] you can cheaply and incrementally build a test suite that will help you measure your progress, spot unintended side effects, and focus your development efforts." (Kent Beck and Erich Gamma, Test Infected: Programmers Love Writing Tests). As a service to the MacOSX/WebObjects developer community, OCUnit is distributed as open source. OCUnit includes: - SenTestingKit, a framework to help you write test cases; - Shikenjo.app, an interactive application to run tests, review results; - otest, a testing tool; - a collection of Makefiles and utilities to seamlessly integrate testing with ProjectBuilder. Some benefits of OCUnit: - Object-oriented tests. - Low overhead. - Tests are in the same language as the implementation. - Test case classes cohabit with classes they test. - Promotes aggressive refactoring. - Helps capture requirements, expose dependencies. - Total integration in the development process. - Tests frameworks, bundles, or applications. - No license fee. OCUnit and more information are available at: http://www.sente.ch/software/ocunit More information about Sen:te is available at: http://www.sente.ch/ -- Marco Scheurer (remove "dot mil" from my address) Sen:te OPENSTEP/MacOSX and WebObjects development, consulting and mentoring.
From: holger@_REMOVE_THIS_.wizards.de (Holger Hoffstaette) Newsgroups: comp.sys.next.programmer Subject: Re: Test first, develop later: Sen:te has released OCUnit Date: 20 Dec 1998 21:55:50 GMT Organization: The secret circle of the NSRC Message-ID: <75jrp6$i8v@ragnarok.en.uunet.de> References: <367d3b1d.0@epflnews.epfl.ch> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Marco Scheurer wrote: > Test first, develop later. > > Sen:te releases OCUnit. > > OCUnit is a testing framework for Objective C in the MacOSX/Yellow Box > environment. It is based on SUnit, Kent Beck's Smalltalk testing > framework, also available for Java under the name JUnit. > (..) What can I say..'excellent' !! :-) I had recently thought about porting JUnit/SUnit myself, but didn't think too many ObjC programmers (except maybe the really crazy ones :) would actually care. I'd be curious to hear if/how you have integrated this into your general dev process (manual (one-shot) testing, automated regression testing, etc.), what kind of problems you have encountered, and what kind of overall, *measurable* benefits you have experienced. Thanks Holger
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer Subject: Re: Test first, develop later: Sen:te has released OCUnit Date: 21 Dec 1998 07:20:37 GMT Organization: Technische Universitaet Berlin, Deutschland Message-ID: <75kss5$q8e$1@news.cs.tu-berlin.de> References: <367d3b1d.0@epflnews.epfl.ch> <75jrp6$i8v@ragnarok.en.uunet.de> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit NNTP-Posting-Date: 21 Dec 1998 07:20:37 GMT holger@_REMOVE_THIS_.wizards.de (Holger Hoffstaette) writes: [Sente releases testing framework] >What can I say..'excellent' !! :-) >I had recently thought about porting JUnit/SUnit myself, but didn't think >too many ObjC programmers (except maybe the really crazy ones :) would >actually care. Well, I was sufficiently interested that I implemented a testing framework a couple of months ago. It differs from the SUnit/JUnit work in that (a) the manual testing overhead is reduced for common cases and (b) the code-base does not become dependent on the testing framework. >I'd be curious to hear if/how you have integrated this into your general >dev process (manual (one-shot) testing, automated regression testing, >etc.), what kind of problems you have encountered, and what kind of >overall, *measurable* benefits you have experienced. Nothing objective from me, but I can tell you that it made me go faster. :-) Very little additional automation so far because saying 'testlogger <framework1 framework2 ... framework1> on the command-line is really not that much trouble. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: theisen@akaMail.com (Dirk Theisen) Newsgroups: comp.sys.next.programmer Subject: Re: Test first, develop later: Sen:te has released OCUnit Date: Mon, 21 Dec 1998 15:00:31 +0100 Organization: University of Bonn, Germany Message-ID: <1dkdpm1.1in9kga60o8gcN@ascend-tk-p83.rhrz.uni-bonn.de> References: <367d3b1d.0@epflnews.epfl.ch> <75jrp6$i8v@ragnarok.en.uunet.de> <75kss5$q8e$1@news.cs.tu-berlin.de> User-Agent: MacSOUP/D-2.4 Hello, Marcel! Marcel Weiher <marcel@cs.tu-berlin.de> wrote: > (b) the code-base does not become dependent on the testing framework. Could you please explain that one a bit further? How is the integration in PB done? Regards, Dirk -- No RISC - No fun http://theisen.home.pages.de/
From: "Wiro van Schaik" <ourxs@xs4all.nl> Newsgroups: comp.sys.next.programmer Subject: Repost: Problem: Spontenous death notifications on just one side of NXConnection Date: Mon, 21 Dec 1998 20:15:51 +0100 Organization: XS4ALL, networking for the masses Sender: ourxs@asp54-4.amsterdam.nl.net Message-ID: <75m6lc$b0t$1@news2.xs4all.nl> Cross-posted to: mailinglist next-prog@omnigroup.com My previous post was probably unreadable because of all the html-formatting so this is my second try in plain ASCII. I apologize for any inconveniences I caused. L.S. I have a problem with a NeXTSTEP 3.3 system which uses Distributed Objects. The system consist of a server part, who receives message through a socket interface and distributes these to a number of clients through a Distributed Objects connection. The server also receives messages from the clients to send through a socket connection to another process as well as distribute these messages again to the other clients. The server actually acts as a notification center to keep the information displayed by the clients in sync with each other as well as with another (Non-NeXTSTEP) process running on a HP-UX machine (hence the socket connection). The problems is that I sometimes get port death notifications on just one side of a NXConnection. For instance, the server would receive a death notification for a connection to a client and consequently clear all references to the client. The client however doesn't get a death notification and waits forever for messages from the server. The same problem but then the other way around also occurred. Some additional clues which might ring a bell with someone: Most clients have a 2Mbit leased line connection with the server. The problems indeed occurs much more frequently with these clients then with clients connected through a regular 10Mb Ethernet. The clients also have to communicate with a database server across this same connection using EOF. The loads over this connection however don't seem to be all that heavy (database queries have good response times and aren't generally that heavy in the amount of data to transfer) The problem occurs most often about 1 minute after the client has started and registered at the server. First I thought it might have something to do with the number of NXConnections as they don't seem to be freed after a client correctly deregisters. At a certain moment in time the problem seem to occur when more then 32 NXConnections were used during the lifetime of the server. But recently the problem occurred after more then 60 NXConnections had been used. Both server and client are single threaded. They have been built using NeXTSTEP 3.3, EOF 1.1, with the latest patches for these version applied. We use both HP 715 workstations as well as HP Pentium PC's. All messages (except the registration message) are asynchronous. They range from 4 to about 60 bytes per message. There are no problems when a client or a server dies itself. The death notification is correctly received and handled on the other end (apart from the fact that the NXConnection doesn't seem to be freed). I've tried enabling the DO debugging but that gave me no clue. I don't have clue were to look for next. If anybody out there could give me any hints on issues I should check, I would but ultimately grateful. If anybody would need more information before they would be able to give some answer, just let me know and I will do my best to provide it. Regretfully, I probably can't publish a lot of source code as it is not our code but from a customer we are trying to help out. Thanks beforehand, Wiro van Schaik, Consulting Software Engineer /* * Shared Objectives bv - Bringing engineering to software * * Transpolis Schiphol Airport, Polaris Avenue 53 * 2132 JH Hoofddorp, The Netherlands * Tel.: +31 (0)23 5685788, Fax: +31 (0)23 5685701 * WWW: <http://www.shared-objectives.com> */
From: John Hornkvist <sorry@no.more.spams> Newsgroups: comp.sys.next.programmer Subject: Re: Repost: Problem: Spontenous death notifications on just one side of NXConnection Date: Mon, 21 Dec 1998 21:24:32 GMT Organization: Chalmers Tekniska Högskola Sender: john@haddock.cd.chalmers.se (John Hprnkvist) Message-ID: <F4C3Gw.J7@haddock.cd.chalmers.se> References: <75m6lc$b0t$1@news2.xs4all.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: ourxs@xs4all.nl In <75m6lc$b0t$1@news2.xs4all.nl> "Wiro van Schaik" wrote: > Cross-posted to: mailinglist next-prog@omnigroup.com > > My previous post was probably unreadable because of all the html-formatting > so this is my second try in plain ASCII. I apologize for any inconveniences > I caused. > > L.S. It made it fine to where I am at. I even replied to it; I guess your email address is a spam bouncer. >The problems is that I sometimes get port death notifications on just one >side of a NXConnection. >For instance, the server would receive a death notification for a connection >to a client and consequently clear all references to the client. The client >however doesn't get a death notification and waits forever for messages from >the server. The same problem but then the other way around also occurred. At least on OPENSTEP you shouldn't expect to get a nofication on both sides; in my experience -- and I believe it is in the docs somewhere -- only the sender will notice when a connection goes down. As a remedy, run a timer that fires at a suitable interval and tries sends a message over the connection. If the connection is down, you catch the exception, or notification, and bring it up again. If you need to avoid that clients get out of sync, you can log messages in the notification center, and pass them once the client connects. This requires that the client keeps track of when it got the last notification -- in the server's time frame; a time stamp in the userInfo part of a notification is useful for this -- and the server needs to log all messages with the same time stamp. Regards, John Hornkvist Name:nhoj Address:cd.chalmers.se
From: marcel@cs.tu-berlin.de (Marcel Weiher) Newsgroups: comp.sys.next.programmer Subject: Re: Test first, develop later: Sen:te has released OCUnit Date: 21 Dec 1998 19:32:46 GMT Organization: Technische Universitaet Berlin, Deutschland Message-ID: <75m7ou$4ul$1@news.cs.tu-berlin.de> References: <367d3b1d.0@epflnews.epfl.ch> <75jrp6$i8v@ragnarok.en.uunet.de> <75kss5$q8e$1@news.cs.tu-berlin.de> <1dkdpm1.1in9kga60o8gcN@ascend-tk-p83.rhrz.uni-bonn.de> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit NNTP-Posting-Date: 21 Dec 1998 19:32:46 GMT theisen@akaMail.com (Dirk Theisen) writes: >Marcel Weiher <marcel@cs.tu-berlin.de> wrote: >> (b) the code-base does not become dependent on the testing framework. >Could you please explain that one a bit further? After first looking at XP and the Java testing framework, I was excited by the approach but taken aback somewhat by the additional dependency (which didn't seem necessary to me) and by the overhead needed for even simple testing. It may be that there is a rationale that I haven't discovered yet, but I tried to implement an approach that had lower overhead and would therefore not discourage me from testing. My current approach is to put all functionality in frameworks, with executable tools/applications only providing the thinnest of wrappers. The testing framework loads one or more frameworks, asks all classes in that framework for their test-selectors and a fixture object, creates tests grouped by framework/class and then runs the tests. For many classes, the class object itself is sufficient as a fixture, for those where it isn't a simple NSObject subclass should usually be fine. >How is the integration in PB done? Currently not at all, but that is an orthogonal issues as far as I can tell. I simply haven't found the need yet. Marcel -- Java and C++ make you think that the new ideas are like the old ones. Java is the most distressing thing to hit computing since MS-DOS. - Alan Kay -
From: "Wiro van Schaik" <ourxs@xs4all.nl> Newsgroups: comp.sys.next.programmer Subject: Re: Repost: Problem: Spontenous death notifications on just one side of NXConnection Date: Tue, 22 Dec 1998 22:40:59 +0100 Organization: XS4ALL, networking for the masses Sender: ourxs@hgl99-4.hengelo.nl.net Message-ID: <75p3hf$gkh$1@news2.xs4all.nl> References: <75m6lc$b0t$1@news2.xs4all.nl> <F4C3Gw.J7@haddock.cd.chalmers.se> John Hornkvist wrote in message ... >In <75m6lc$b0t$1@news2.xs4all.nl> "Wiro van Schaik" wrote: > >>The problems is that I sometimes get port death notifications on just one >>side of a NXConnection. >>For instance, the server would receive a death notification for a connection >>to a client and consequently clear all references to the client. The client >>however doesn't get a death notification and waits forever for messages from >>the server. The same problem but then the other way around also occurred. > >At least on OPENSTEP you shouldn't expect to get a notification on both sides; >in my experience -- and I believe it is in the docs somewhere -- only the >sender will notice when a connection goes down. > In generally that sounds a bit unlikely to my. How would the receiver ever know that its sender has died? >As a remedy, run a timer that fires at a suitable interval and tries sends a >message over the connection. If the connection is down, you catch the >exception, or notification, and bring it up again. > How would I bring the connection back up if the initiative for the connection in the first place came from the receiver side. My simplified situation is as follows: I have several clients that register at a central server to receive notifications of events happening in the system. The sender would be the server, the receivers would be the clients. How could the server restart a connection with a client for which it received a death notification. How would the client know that the server got a death notification. Note that the problem I have is getting death notifications for clients which didn't actually die. So the client is still alive and waiting for messages. >If you need to avoid that clients get out of sync, you can log messages in >the notification center, and pass them once the client connects. This >requires that the client keeps track of when it got the last notification -- >in the server's time frame; a time stamp in the userInfo part of a >notification is useful for this -- and the server needs to log all messages >with the same time stamp. > I'm already programming towards such a solution but it still feels very unreliable especially when compared with the socket-based communication we use for communicating with a non-NeXSTEP based system. It also forces the server to remember which clients registered and have them register with some sort of unique id. This poses the problem of choosing unique id's and how long to keep this information around. All in all not the transparency of communication that DO promises. Wiro
From: "Wiro van Schaik" <ourxs@xs4all.nl> Newsgroups: comp.sys.next.programmer Subject: Problem: Spontenous death notifications on just one side of NXConnection Control: cancel <75h97p$kq8$1@news2.xs4all.nl> Date: 22 Dec 1998 21:37:30 GMT Organization: XS4ALL, networking for the masses Sender: ourxs@hgl99-4.hengelo.nl.net Message-ID: <75p3eq$gh0$1@news2.xs4all.nl> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit
From: Cosmo Roadkill <cosmo.roadkill%bofh.int@rauug.mil.wi.us> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <75t0iu$ds5454@vaasa.vaasa.fi> Control: cancel <75t0iu$ds5454@vaasa.vaasa.fi> Date: 24 Dec 1998 09:24:42 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.75t0iu$ds5454@vaasa.vaasa.fi> Sender: billyo@jumbat.com Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Macghod@concentric.net (Steve) Newsgroups: comp.sys.next.programmer,comp.sys.mac.programmer.codewarrior Subject: ADC student program (does it suck????) Date: Thu, 24 Dec 1998 00:21:40 -0800 Organization: Concentric Message-ID: <Macghod-2412980021410001@sdn-ar-002casbarp328.dialsprint.net> I am thinking about joing this program, but frankly, as far as I can tell it really really blows. I could really give a flying duck about the demo programs, the seeds would be nice, but I could live without them, but the tool cd's would be very usefull. From reading the web page tho, it seems as if Apple is only including a tiny bit of the tools and samples, kind of a enticement to get people to pay the $500 for the select program. Read the web page, and I think it is very obvious they are using this to a) get some money b) give only a small sampling of what you would get if you join the more expensive programs, hopefully enticing people to join. I was seriously considering joining the developer program before they doubled the price and took away the hardware discounts, and frankly this program just seems lame as hell. This isnt meant as a inflamatory post, but is more of a opportunity for people to convince me (mainly by showing cool things you get that I did not know you get) that this program is worth it.
From: gericom@usa.net Newsgroups: comp.sys.next.programmer Subject: The Interactive Dance Compilation you can play and Mix on your PC!! 7648 Date: 24 Dec 1998 12:19:53 GMT Organization: Centro Servizi Interbusiness Message-ID: <75tbh9$gjq$2185@fe2.cs.interbusiness.it> The Interactive Dance Compilation you can play and Mix on your PC!! The best dance hits of '90 and 98 together in Discoparade the first double compilation that you can play and Mix ! Discoparade is the one and only interactive compilation ! http://www.discoparade.com/ http://www.discoparade.com/ bwpfpsncfljgirxtcobigz
From: holger@_REMOVE_THIS_.wizards.de (Holger Hoffstaette) Newsgroups: comp.sys.next.programmer Subject: Re: Test first, develop later: Sen:te has released OCUnit Date: 24 Dec 1998 04:17:53 GMT Organization: The secret circle of the NSRC Message-ID: <75sf9h$et@ragnarok.en.uunet.de> References: <367d3b1d.0@epflnews.epfl.ch> <75jrp6$i8v@ragnarok.en.uunet.de> <75kss5$q8e$1@news.cs.tu-berlin.de> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Marcel Weiher wrote: > holger@_REMOVE_THIS_.wizards.de (Holger Hoffstaette) writes: > > [Sente releases testing framework] > > >What can I say..'excellent' !! :-) > >I had recently thought about porting JUnit/SUnit myself, but didn't think > >too many ObjC programmers (except maybe the really crazy ones :) would > >actually care. > > Well, I was sufficiently interested that I implemented a testing framework > a couple of months ago. It differs from the SUnit/JUnit work in that > (a) the manual testing overhead is reduced for common cases and (b) the > code-base does not become dependent on the testing framework. Sounds good. Up to now my test cases have been stand-alone and not automated (i.e. a simple aggregate project of tools). That's still better than nothing, I guess. ;) Working on the framework is done by copying the individual test subproject out of the aggregate, developing the code in the mini-project until it works 100%, and putting the new, extended test case back into the aggregate and the code into the framework. A clean rebuild of both the famework and the test suite must not yield any warnings or errors. In 99% of all cases this 'Just Works'. The remaining 1% is mostly typo-related. > Nothing objective from me, but I can tell you that it made me go > faster. :-) Very little additional automation so far because > saying 'testlogger <framework1 framework2 ... framework1> on > the command-line is really not that much trouble. I found my procedure to be ideal if several well-defined code changes had to be done; however, constantly changing ideas are better left alone in the test case until the idea and hence the code has stabilized in my feeble brain. Also works well together with CVS. I'm currently also thinking of putting the test code into the framework as well, so that automation would be possible. We'll see.. Holger
From: sanguish@digifix.com Newsgroups: comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: NEXTSTEP/OpenStep Resources on the Net Supersedes: <6039914130021@digifix.com> Date: 27 Dec 1998 04:45:38 GMT Organization: Digital Fix Development Message-ID: <14228914734832@digifix.com> Topics include: Major OpenStep/NEXTSTEP World Wide Web Sites OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups Major OpenStep/NEXTSTEP FTP sites NeXTanswers Major OpenStep/NEXTSTEP World Wide Web Sites ============================================ The following sites are a sample of the OpenStep related WWW sites available. A comprehensive list is available on Stepwise. Stepwise OpenStep/Rhapsody Information Server http://www.stepwise.com Stepwise has been serving the OpenStep/NEXTSTEP community since March 1994. Some of the many resources on the site include: original YellowBox and Rhapsody articles, mailing list information, extensive listing of FTP and WWW sites related to Rhapsody, OpenStep and NEXTSTEP, OpenStep related Frequently Asked Questions. NeXT Software Archives @ Peak.org http://www.peak.org/next http://www.peak.org/openstep http://www.peak.org/rhapsody PEAK is the premier NeXTStep/OpenStep FTP site in North America. Apple Enterprise Software Group (formerly NeXT Computer, Inc.) http://enterprise.apple.com http://enterprise.apple.com/NeXTanswers Here is where you'll find the NeXTanswers archive, with information on OpenStep installation, drivers and software patches. Apple Computer's Rhapsody Developer Release Site http://gemma.apple.com/rhapsody/rhapdev/rhapsody.html These pages are focused on tools and resources you need to develop great Rhapsody software products.. Rhapsody Developer Documentation http://gemma.apple.com/techinfo/techdocs/rhapsody/rhapsody.html WebObjects Documentation http://gemma.apple.com/techinfo/techdocs/enterprise/enterprise.html OpenStep/NEXTSTEP/Rhapsody Related Usenet Newsgroups ==================================================== COMP.SYS.NEXT.ADVOCACY This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. COMP.SYS.NEXT.ANNOUNCE Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Messages posted to announce should NOT be posted or crossposted to any other comp.sys.next groups. COMP.SYS.NEXT.BUGS A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT-specific groups as well. COMP.SYS.NEXT.HARDWARE Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. COMP.SYS.NEXT.MARKETPLACE NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. COMP.SYS.NEXT.MISC For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! COMP.SYS.NEXT.PROGRAMMER Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. COMP.SYS.NEXT.SOFTWARE This is a place to talk about [third party] software products that run on NEXTSTEP systems. COMP.SYS.NEXT.SYSADMIN Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. ** RELATED NEWSGROUPS ** COMP.SOFT-SYS.NEXTSTEP Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. COMP.LANG.OBJECTIVE-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. COMP.OBJECT Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original comp.sys.next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News =========================================== Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep next-advocacy next-announce next-bugs next-hardware next-marketplace next-misc next-programmer next-software next-sysadmin object lang-objective-c (For a full description, send mail to listserv@antigone.com). The subscription syntax is essentially the same as Majordomo's. To subscribe, send a message to *-request@lists.best.com saying: subscribe where * is the name of the list e.g. next-programmer-request@lists.best.com Major OpenStep/NEXTSTEP FTP sites ================================= ftp://ftp.next.peak.org The main site for North American submissions formerly ftp.cs.orst.edu ftp://ftp.peanuts.org: (Peanuts) Located in Germany. Comprehensive archive site. Very well maintained. ftp://ftp.nluug.nl/pub/comp/next NeGeN/NiNe (NEXTSTEP Gebruikers Nederland/NeXTSTEP in the Netherlands) ftp://cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp://ftp.nmr.embl-heidelberg.de/pub/next eduStep ftp://ftp.next.com: See below ftp.next.com and NextAnswers@next.com ===================================== [from the document ftp://ftp.next.com/pub/NeXTanswers/1000_Help] Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, read them on the world-wide web, transfer them by anonymous ftp, or download them from the BBS. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to nextanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification REPLY-TO address sets the e-mail address NeXTanswers uses These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO If you have any problem with the system or suggestions for improvement, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS VIA THE WORLD-WIDE WEB To use NeXTanswers via the Internet World-Wide Web connect to NeXT's web server at URL http://www.next.com. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to nextanswers-request@next.com. USING NEXTANSWERS BY MODEM To use NeXTanswers via modem call the NeXTanswers BBS at (415) 780-2965. Log in as the user "guest", and enter the Files section. From there you can download NeXTanswers documents. FOR MORE HELP... If you need technical support for NEXTSTEP beyond the information available from NeXTanswers, call the Support Hotline at 1-800-955-NeXT (outside the U.S. call +1-415-424-8500) to speak to a NEXTSTEP Technical Support Technician. If your site has a NeXT support contract, your site's support contact must make this call to the hotline. Otherwise, hotline support is on a pay-per-call basis. Thanks for using NeXTanswers! _________________________________________________________________ Written by: Eric P. Scott ( eps@toaster.SFSU.EDU ) and Scott Anguish ( sanguish@digifix.com ) Additions from: Greg Anderson ( Greg_Anderson@afs.com ) Michael Pizolato ( alf@epix.net ) Dan Grillo ( dan_grillo@next.com )
From: nvjghn@ggggg.ggg Newsgroups: comp.sys.next.programmer Subject: FREE 4 you! 3614 Date: 27 Dec 1998 23:46:16 GMT Organization: SIAST Message-ID: <766gs8$rk8680@www1.siast.sk.ca> Choose from 100's of FREE catalogs: Outdoor Collectibiles Automotive Business Children Entertainment Toys Games Pets Fashion Gifts Much Much More!!!!!! Come to http://www.freeyellow.com/members3/entertainmentcity/page10.html bbqdyyfqfhlujcnkgttuqpbtrkncsnpxigzgxemgubikqvipxwzkujfxgiolilntsjsnqhguxohzgyket
From: Cosmo Roadkill <cosmo.roadkill%bofh.int%rauug.mil.wi.us@bofh.int> Newsgroups: news.software.nntp,alt.gothic,soc.culture.jewish,uk.test,soc.culture.israel,comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: cmsg cancel <IE01dgNyXq9YtOt.0iyxNIurE3UHFu.Ha3uZSkU@digifix.com> Control: cancel <IE01dgNyXq9YtOt.0iyxNIurE3UHFu.Ha3uZSkU@digifix.com> Date: 28 Dec 1998 11:51:38 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.IE01dgNyXq9YtOt.0iyxNIurE3UHFu.Ha3uZSkU@digifix.com> Sender: mrsam@concentric.net My From: line has been fudged because many test newsgroup autoresponders respond to control messages. My apologies! Please see the X-Cancelled-By: line for my proper address. Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
From: Cosmo Roadkill <cosmo.roadkill%bofh.int%rauug.mil.wi.us@bofh.int> Newsgroups: news.software.nntp,alt.gothic,soc.culture.jewish,uk.test,soc.culture.israel,comp.sys.next.advocacy,comp.sys.next.announce,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.programmer Subject: cmsg cancel <Kojgs1vg06v8N.QC2dj6tNc.SRZZWG8T6dp3@digifix.com> Control: cancel <Kojgs1vg06v8N.QC2dj6tNc.SRZZWG8T6dp3@digifix.com> Date: 28 Dec 1998 12:01:23 GMT Organization: BOFH Space Command, Usenet Division Message-ID: <cancel.Kojgs1vg06v8N.QC2dj6tNc.SRZZWG8T6dp3@digifix.com> Sender: mrsam@concentric.net My From: line has been fudged because many test newsgroup autoresponders respond to control messages. My apologies! Please see the X-Cancelled-By: line for my proper address. Article cancelled as EMP/ECP, exceeding a BI of 20. The "Current Usenet spam thresholds and guidelines" FAQ is available at http://www.uiuc.edu/ph/www/tskirvin/faqs/spam.html Please include the X-CosmoTraq header of this message in any correspondence specific to this spam. Sick-O-Spam, Spam-B-Gon!
These are the contents of the former NiCE NeXT User Group NeXTSTEP/OpenStep software archive, currently hosted by Marcel Waldvogel and Netfuture.ch.