This is Prog-10.gz in view mode; [Up]
From: anderson@sapir.ling.yale.edu (Stephen Anderson) Newsgroups: comp.sys.next.programmer Subject: 4.0 despair Date: 1 Oct 1996 00:10:16 GMT Organization: Yale University Message-ID: <52pnh8$o8o@news.ycc.yale.edu> Please, can someone help me figure out how to recompile code that worked under 3.3, now that I've made the horrible mistake of upgrading to 4.0? In particular, emacs used to be build with little or no problem - so little that a non-programmer like me could usually get it running (except for the actual NS code, which others dealt with). But in 4.0, NeXT provides only a terribly out-of-date emacs (based on FSF emacs 18.59), which means the elisp packages I depend on won't work. The executables built under 3.3 for later versions of emacs all crash under at least some circumstances that turn out to matter to me. And when I try to rebuild emacs, either the NextStep-ified version or the basic FSF emacs 19.34, with no window code at all, I keep running into strange gotchas connected with linking/loading. I've spent hours trying to puzzle out from NeXT's GNU sources just how they got 18.59 to build, so as to apply the same lesson to the 19.34 sources, but to no avail. Perhaps someone at NeXT, or someone elsewhere with some understanding of what stuff changed from 3.3 to 4.0 concerning the standard C and math libraries, would be so kind as to spend a couple of hours starting from the FSF emacs-19.34 sources and building a simple --with-x=no version of emacs (even better a version with X, based on the coXist libraries, but a simple terminal based version would do), and then let me know what the magic is. There's no ObjC involved, no class structure that might have changed, etc., as far as I know. It's all plain C, and I'm sure there's somebody out there who knows what to do with it. I'd sure be grateful...... --Steve Anderson
From: "Hiroki Yamaberi" <yamaberi@gokhu.com> Newsgroups: comp.sys.next.programmer,comp.sys.next.software Subject: CMU Common Lisp for NS FIP 3.3 Date: 1 Oct 1996 00:30:47 GMT Organization: GOKHU Message-ID: <01bbaf2f$beb6df00$020000c0@mini> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-2022-JP Content-Transfer-Encoding: 7bit Does anyone ported CMU Common Lisp for x86 to NEXTSTEP 3.3 for Intel Processor? Thanks -- /// Hiroki Yamaberi /// POWERED BY NEXTSTEP AND NAVITIMER
From: aisbell@ix.netcom.com (Art Isbell) Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: Re: Printf %g format still broken in OPENSTEP 4.0 Date: 1 Oct 1996 02:33:01 GMT Organization: Netcom Distribution: world Message-ID: <52pvst$7jo@dfw-ixnews12.ix.netcom.com> References: <52p5vp$1sq@tribune.usask.ca> eric@skatter.USask.Ca wrote: > I can hardly believe this. It is now *SEVEN YEARS* since I first reported this bug (in > NextStep 1.0). I have reported this error in every subsequent release up to and > including NEXTSTEP 3.3. In May, 1994 I received mail from NeXT saying that this would be > fixed in 4.0. Looks like they missed it again. May, 1994!! Wasn't that back when 4.0 was supposed to have an object-oriented file systems and all sorts of other wiz-bang features? Is the person who wrote that still at NeXT ?-) But 4.0 isn't really "4.0". I think the real 4.0 will be released shortly as 4.1. So keep your test program warm... -- Art Isbell NeXT/MIME Mail: aisbell@ix.netcom.com Trego Systems Voice/Fax: +1 408 335 2515 CaseServ: OPENSTEP Voice Mail: +1 408 335 1154 managed care solutions US Mail: Felton, CA 95018-9442
From: shill@iphysiol.unil.ch (Sean Hill) Newsgroups: comp.sys.next.programmer Subject: Re: Palettes with OpenStep Date: 1 Oct 1996 08:32:32 GMT Organization: University of Lausanne CH (Switzerland) Message-ID: <52qkv0$b72@cisun2000.unil.ch> References: <52o7jh$1l3@cisun2000.unil.ch> There were a few typos in that last post (typing from memory, while asleep). However... Just so you know... NSObject doesn't conform to the NSCoding protocol and therefore you don't send a [super encodeWithCoder:coder] message for subclasses of NSObject... There seems to be some problem with the archiving that's happening (maybe one of my objects stored in the dictionary...) Thanks for the response- Sean The real code (from subclass of NSView) - initWithCoder:(NSCoder *)coder { [super initWithCoder:coder]; plotData = [[coder decodeObject] retain]; [self _setup]; [self _scaleToView]; return self; } - (void)encodeWithCoder:(NSCoder *)coder { [super encodeWithCoder:coder]; [coder encodeObject:plotData]; } ---- The real code (subclass of NSObject) -(void)encodeWithCoder:(NSCoder *)aCoder { // NSObject does not conform to NSCoding protocol //[super encodeWithCoder:aCoder]; [aCoder encodeObject:attributeDictionary]; } - initWithCoder:(NSCoder *)aDecoder { // NSObject does not conform to NSCoding protocol //self = [super initWithCoder:aDecoder]; attributeDictionary = [[aDecoder decodeObject] retain]; [self updateAttributeDictionary]; return self; }
From: ians@cam-ani.co.uk (Ian Stephenson) Newsgroups: comp.sys.next.programmer Subject: Re: Printf %g format still broken in OPENSTEP 4.0 Date: Tue, 1 Oct 1996 09:54:32 GMT Organization: Cambridge Animation Systems Ltd Sender: news@cam-ani.co.uk Message-ID: <DyLCux.HKz@cam-ani.co.uk> References: <52pvst$7jo@dfw-ixnews12.ix.netcom.com> In article <52pvst$7jo@dfw-ixnews12.ix.netcom.com> aisbell@ix.netcom.com (Art Isbell) writes: > May, 1994!! Wasn't that back when 4.0 was supposed to have an > object-oriented file system It does! (or at least it kinda does). The OO-Filestore is EOF, or rather EOF is an OO-filestore. EOF is actually really cool hacker technology of the kind NeXT used to deliver. It is the OO-FS (or at least the interface to a DB which together makes the FS) that was rumoured years ago. If they marketed it honestly it would be awesome - for transfering data (offline) between apps it is fantasic. Unfortunatly NeXT marketing have made sure that no one finds this out (it's for serious DB apps only), and NeXT Sales have unbundled it - pricing it completly out of reach. I only discovered it when I needed to do some DB work, so it was the obvious choice. I now know that it's stength is for non-db apps - allowing REAL apps to organise persistant data in powerfull ways. NeXT have made sure that no hacker will go near it, when in fact it's a hackers dream. > But 4.0 isn't really "4.0". I think the real 4.0 will be released > shortly as 4.1. I've also come to this conclusion (hope!) $an
From: ehutch@hypnos.norden1.com (E. Hutchinson) Newsgroups: comp.sys.next.programmer,comp.sys.next.misc,misc.jobs.contract Subject: NEXTSTEP/Contract/Va Date: 1 Oct 1996 12:46:00 GMT Organization: Norden 1 Communications Message-ID: <52r3q8$fbm@tofu.alt.net> Programmer/analyst/developer NEXTSTEP Objective C Commercial experience EOF---A Plus Contract---Long term Area--Va Start Date---october 1996 Positions---2 To Be Considered---Fax resume or mail a hard copy. -- ehutch@norden1.com (419) 893-6367 [fax] Omni Search (419) 893-6334 [voice] 1310 Craig Maumee, Ohio 43537
From: dekorte@suite.com (Steve Dekorte) Newsgroups: comp.sys.next.programmer Subject: system call in OS4 Date: 1 Oct 1996 19:33:52 GMT Organization: Suite Software Message-ID: <52rrn0$fa2@news.onramp.net> Anyone tried using the system(const char * arg) call from stdlib in OS4? I'm having some odd problems with it. -- Steve Dekorte - OpenStep Developer - Anaheim, CA "Fundamentalism isn't about religion. It's about power." - S. Rushdie
From: jabi@acsu.buffalo.edu Newsgroups: comp.sys.next.programmer Subject: FORTRAN Compiler -> Intel binary Date: 1 Oct 1996 20:53:14 GMT Organization: University at Buffalo Message-ID: <52s0bq$pdc@prometheus.acsu.buffalo.edu> NNTP-Posting-User: jabi URL: http://www-personal.umich.edu/~wjabi/ Hello: Can someone tell me if Absoft's f77 compiler can compile to motorola and intel architectures? If so, what is the option to use? Please e-mail me. Thanks. -- Wassim M. Jabi, Assistant Professor Department of Architecture - 312B Hayes Hall State University of New York at Buffalo Buffalo, NY 14214-3087 Tel: +1 716.829.3485 Ext. 323 Fax: +1 716.829.3256 wjabi@arch.buffalo.edu (Text/MIME/NeXTMail)
From: bbum@friday.com (Bill Bumgarner) Newsgroups: comp.sys.next.programmer Subject: Ghostscript 4.01 or 4.03 anyone? Date: 1 Oct 1996 19:55:29 GMT Organization: Friday Software & Consulting Message-ID: <52rsvh$ftd@chinx10.thoughtport.net> Does anyone have Ghostscript 4.03 or 4.01 available for NEXTSTEP? Without X11 dependencies? I have found that Ghostscript under Linux makes an excellent driver for the HP deskjet series of printers-- I'm wondering if the same holds true for Ghostscript under Linux... I'm not so much interested in hooking up the printer to the parallel port (I have a 486 DX/2 running Linux that makes an excellent lpd server ;-), but am very much interested in using Ghostscript as the rasterization engine... thanks, b.bum
From: Alex Blakemore <alex@genoa.com> Newsgroups: comp.lang.objective-c,comp.sys.next.programmer Subject: Re: Can someone help with an inheritance puzzler? Date: 1 Oct 1996 04:19:26 GMT Organization: Genoa Software Systems Message-ID: <52q64e$att@saturn.genoa.com> References: <slrn53086d.hf7.davet@Fast.thomas.home> <50ps4l$kqe@news.onramp.net> <51k670$nbg@news.onramp.net> <slrn53vapu.8bb.ifeulner@xenon.cube.de> <51rlo8$91k@saturn.genoa.com> <526uri$dh7@news.onramp.net> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: dekorte@suite.com In <526uri$dh7@news.onramp.net> Steve Dekorte wrote: > Actually it's the ugly design and not the performance that bugs me. > But performance rhetoric is usually more persuasive, since > design remains something on which only intuition can currently decide. It was refreshing to read such an honest self-critique of an argument, since so many USENET discussions degrade into emotive attacks (in other groups, of course ;-) Whether a design is elegant (or ugly) depends in large part upon the goals its intended to satisfy. A design can be ideal for one purpose (say extensibility) and poor for another (say performance). I think class clusters are an interesting idea. They simplify some tasks and complicate some others. Rather than just flaming the concept, I'd rather discuss in what circumstances they are useful and not. On the plus side, I see + simple uniform interface to similar concrete classes + efficient use of memory and time (in most circumstances) + hide complexity from user + separate mutable from immutable instances + ability to revise concrete classes without impacting callers On the minus side, I see - tricky, (but possible), to provide new subclasses - difficult to provide both mutable and immutable subclasses without replication - a little more complex to debug, archive custom subclasses Given the subclassing complexity, I'd say class clusters have merit as a design approach when: the class hierarchy is fairly flat (not much subclassing) the best storage/implementation requirements vary dramatically depending upon concrete details users don't need to care about the concrete details, only the behavior defined in the abstract interface Conversely, class clusters are probably a bad idea when: the inheritance hierarchy is deep users need to know representation details about individual subclasses. The Abstract Factory in the Design Patterns book seems to be the same thing as a class cluster. -- Alex Blakemore alex@genoa.com NeXT, MIME and ASCII mail accepted
From: Constantin Szallies <szallies@energotec.de> Newsgroups: comp.sys.next.programmer Subject: Re: Palettes with OpenStep Date: 2 Oct 1996 07:58:50 GMT Organization: Tech Net GmbH Message-ID: <52t7bq$hl1@ddfservb.technet.net> References: <52o7jh$1l3@cisun2000.unil.ch> <52qkv0$b72@cisun2000.unil.ch> shill@iphysiol.unil.ch (Sean Hill) wrote: [snip] >The real code (subclass of NSObject) [snip] >- initWithCoder:(NSCoder *)aDecoder >{ >// NSObject does not conform to NSCoding protocol >//self = [super initWithCoder:aDecoder]; > attributeDictionary = [[aDecoder decodeObject] retain]; > [self updateAttributeDictionary]; > return self; >} Shouldn't you call [super init] (or [self init]) inside the -initWithCoder: message?? -- Constantin Szallies, Energotec GmbH szallies@energotec.de 49211-9144018
From: Erik Doernenburg <erik@object-factory.com> Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: detachNewThreadSelector... retains its target (?) Date: 2 Oct 1996 09:27:39 GMT Organization: Object Factory GmbH (Germany) Message-ID: <52tcib$22l@isabella.object-factory.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit I'm not 100% sure about this, but it seems that the object passed as target to +detachNewThreadSelector:toTarget:withObject in NSThread receives a retain message when the thread is spawned but does not receive a release message when the thread exits. Clearly, after the method is finished the object has an invalid, ie. too high retain count and will not be deallocated properly. Does someone have similar experiences? thanks in advance erik ---------------------------------------------------------------------------- Erik Dörnenburg OBJECT FACTORY Gesellschaft für Informatik und Datenverarbeitung mbH Lohbachstraße 12, 58239 Schwerte, Germany erik@object-factory.com or erik@lagado.free.de (priv)
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.programmer Subject: Re: control a PopUpList Date: 2 Oct 1996 10:30:34 GMT Organization: Department of Electronics, University of York, UK. Message-ID: <52tg8a$i8h@netty.york.ac.uk> References: <52ecdd$24t@rks1.urz.tu-dresden.de> On 26 Sep 1996 16:53:01 GMT, Andreas Dietzsch <dietzsch@rmhs2.urz.tu-dresden.de> wrote: > How can i setup the selected Cell of a PopUpList while Runtime. I try the > "selectCellWithTag:x"-Message but it do not work. > > Thanks for every help > > Andreas void select_tag(id popbutton, int tag) /* * popbutton points to the button that raises the popup * list (could be an outlet set within interface builder */ { id popmatrix = [[popbutton target] itemList]; [popmatrix selectCellWithTag:tag]; [popbutton setTitle:[[popmatrix selectedCell] title]]; } note that any matrix method that sets a selected cell will do instead of selectCellWithTag (e.g. selectCellAt::) cheers, rog.
From: wxb1@wvp_dev_lt_08.wda.disney.com (wxb1) Newsgroups: comp.sys.next.programmer Subject: D'OLE Evaluation Date: 2 Oct 1996 13:53:36 GMT Organization: The Walt Disney Company Distribution: world Message-ID: <52ts50$hjj@louie.disney.com> Keywords: DOLE Hi, I was wondering if anyone has had any experience working with D'OLE. I am about to embark on adventure of having to move an all NEXSTEP company to Windows-NT and am thinking about using D'OLE to ease the transition. The questions I am looking for are as follows: How robust is D'OLE as a product ? How good or bad is the performance ? How well does it work if you have Windows-NT Client (i.e running Word) accessing a NeXSTEP/Mach server (i.e the server runs database reports and returns text or EPS Images) ? Does the existing NS/Code base have to be made OpenStep compliant ? Any answers would be very much appreciated! - willi P.S. I will post a summary of all the responses I get. wxb1@ajax.wda.disney.com
From: royalta@aol.com (RoyalTA) Newsgroups: comp.sys.next.programmer Subject: Re: FORTRAN Compiler -> Intel binary Date: 2 Oct 1996 12:07:10 -0400 Organization: America Online, Inc. (1-800-827-6364) Sender: root@newsbf02.news.aol.com Message-ID: <52u3ve$ac4@newsbf02.news.aol.com> References: <52s0bq$pdc@prometheus.acsu.buffalo.edu> Each platform has its own compiler which must be used on that platform to produce a binary for that platform. Eg. To compile for Intel you must have the Intel compiler and compile using an Intel machine. Tony Royal Jenike & Johanson, Inc.
From: eric@skatter.USask.Ca Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: Re: detachNewThreadSelector... retains its target (?) Date: 2 Oct 1996 16:32:44 GMT Organization: University of Saskatchewan Message-ID: <52u5fc$h9b@tribune.usask.ca> References: <52tcib$22l@isabella.object-factory.com> Erik Doernenburg <erik@object-factory.com> wrote: > >I'm not 100% sure about this, but it seems that the object passed as target >to +detachNewThreadSelector:toTarget:withObject in NSThread receives a retain >message when the thread is spawned but does not receive a release message >when the thread exits. Clearly, after the method is finished the object has >an invalid, ie. too high retain count and will not be deallocated properly. >Does someone have similar experiences? > >thanks in advance >erik > Yes, that's what I found, too (at least in NS 3.3 -- haven't tried it in NS 4.0 yet). Here's what I did in my xmodem file transfer object: - (void)sendCompletionMessage:(NSString *)message { [completionStatus release]; completionStatus = message; [completionStatus retain]; xmodemThread = [NSThread currentThread]; [self autorelease]; /* This matches the retain done by dettachThread */ [xmodemThread exit]; } -- Eric Norum eric@skatter.usask.ca Saskatchewan Accelerator Laboratory Phone: (306) 966-6308 University of Saskatchewan FAX: (306) 966-6058 Saskatoon, Canada. NeXTMail accepted.
From: seanl@carmi.cs.umd.edu (Sean Luke) Newsgroups: comp.sys.next.programmer Subject: Re: A rubix Cube in Pascal...HELP!!!! Date: 2 Oct 1996 16:43:20 GMT Organization: U Maryland at College Park Message-ID: <52u638$rli@mimsy.cs.umd.edu> References: <Pine.NXT.3.95.960914005601.6904F-100000@charisma> <MmCho3u00WBO8141Av@andrew.cmu.edu> <51pn2q$4l7@dfw-ixnews10.ix.netcom.com> <51v9fv$nd3@news2.snfc21.pacbell.net> Mike Paquette (mpaque@pacbell.net) wrote: >johnyreb@ix.netcom.com(Richard Elbert Davis) wrote: >>hello, >> In pascal class right now i have to make a Rubix cube with >>multi-dimensional arrays. Anyone have a sample of a prog I could use. >>Its got me stumped. Thanks alot >Well, a Rubic's Cube is a cube with a 3x3 array of colors for each >face, so the obvious internal representation would be to use a 8x3x3 >array. The cube is logically divided into three planes perpendicular >to each axis, with each plane capable of being rotated in 90 degree >intervals. >The decomposition is intuitively obvious even to the most casual >observer. Once you have followed Mike's suggestions for setting up the cube, need a solution strategy to solve it. This is trivial to even the stupidest of students. One solution is to create an A* search over a set of planning rules, applying either Hierarchical Task Network planning or Partial Order planning to the problem. You should be able to arrange off-the-shelf NeXTSTEP solutions for this through Kutluhan Erol at Intelligent Automation. Another route would be to embed the heuristics found in most cube-solution texts directly into a ruleset and execute the ruleset. An easy way to do this is to code the ruleset in Prolog; for you to embed this properly into Pascal, you'll need to write a Prolog->Pascal compiler. That shouldn't be too tough, even for CS101. :-) Hey, Mike, this is fun! _____________________________________________________________________________ Sean Luke "I've discovered that P==NP, but the proof is too U Maryland at College Park large to fit in the margins of this signature." seanl@cs.umd.edu URL: http://www.cs.umd.edu/~seanl/
From: jalegre@andante-systems.com Newsgroups: comp.sys.next.programmer Subject: Music Kit Question Date: 2 Oct 1996 17:30:38 GMT Organization: SkyPoint Communications, Inc. Message-ID: <52u8ru$n1l@stratus.skypoint.net> What is the current version of the MusicKit? The last version I have is 4.0. How does one get the new version? All attempts to anon ftp to ccrma.stanford.edu have faild. -- John N. Alegre Andante Systems ############################################################### # NeXTMail preferred. | # jalegre@andante-systems.com | If you plant ice, # alegrej@andante.mn.org | you're gonna harvest wind! # jalegre@lenti.med.umn.edu | Hunter/Garcia ############################################################### # URL http://www.andante-systems.com ###############################################################
From: jessica@whitewater.chem.wisc.edu (Jessica Severin) Newsgroups: comp.sys.next.programmer Subject: Openstep 4.1? (was: Which GCC version under 4.0) Date: 2 Oct 1996 18:06:43 GMT Organization: University of Wisconsin, Madison Message-ID: <52uavj$fvi@news.doit.wisc.edu> References: <52bq99$ir9@Holly.aa.net> <1996Sep26.101005.85623@cc.usu.edu> <52ff6o$srt@sjx-ixn2.ix.netcom.com> <01bbacc0$eefa9b00$3e031281@bananajr> "Mark Bessey" <MaRK_BeSSeY@NeXT.CoM> wrote: >Well, the GCC from 4.1 reports: Is Openstep 4.1 already in Beta?!?! Is there any projected release date for it? 6months? 1year? I imagine 4.1 would mostly be bug fixes. I've also heard about a kernel update. Is this in the 4.1 beta? thanks --- Jessica Severin \ Programmer / Analyst jessica@whitewater.chem.wisc.edu \ UW Madison Dept. of Chemistry
From: "Ali Ozer" <Ali_Ozer@next.com> Newsgroups: comp.sys.next.programmer Subject: Re: Rulebooks in Openstep 4.0 Date: 2 Oct 1996 18:04:45 GMT Organization: NeXT Software, Inc Message-ID: <01bbb04a$38087f10$32211281@ant> References: <528op0$mln@noc1.biddeford.com> jeff@tiag.com wrote in article <528op0$mln@noc1.biddeford.com>... > I've read as much as I can find about how Openstep 4.0 for Mach > handles glyphs and characters with the NSLayoutManager etc... > But I don't think the Rulebooks/ (found in NextLibrary/) are currently > documented anywhere. You're right, they are not documented anywhere. In addition, some of the tools for generating those rulebooks are not on the release. > Am I correct or incorrect in thinking one should be able to > create custom rulebooks for mapping glyphs from third party 8-bit fonts into > their proper place in the Unicode 16-bit world so that if we are working with > a file in, for example, TextEdit, and open it as "Unicode" encoding, we would > see our glyphs from our own fonts mapped to Unicode space and displayed in > NSTextView properly? Your assumptions are correct. At some point we plan on making the tools available to allow any font creator to also create a custom rulebook. Note that even without this ability, most third-party fonts will just work, as they have a standard encoding. Currently fonts without a standard-encoding will also work like they might have in the old world, but from an Unicode perspective this isn't ideal. Ali Ali_Ozer@NeXT.com
From: mmalcolm crawford <m.crawford@shef.ac.uk> Newsgroups: comp.sys.next.programmer Subject: Re: Music Kit Question Date: 2 Oct 1996 19:11:02 GMT Organization: University of Sheffield, UK Message-ID: <52ueo6$gjc@bignews.shef.ac.uk> References: <52u8ru$n1l@stratus.skypoint.net> In-Reply-To: <52u8ru$n1l@stratus.skypoint.net> On 10/02/96, jalegre@andante-systems.com wrote: > What is the current version of the MusicKit? The last version I have > is 4.0. How does one get the new version? All attempts to anon ftp > to ccrma.stanford.edu have faild. > Umm ccrma-ftp.stanford.edu ^ It's not at 4.1.1: ftp://ccrma-ftp.stanford.edu/pub/NeXT/MusicKit/MusicKit_4.1.1.pkg.tar Best wishes, mmalc. --
From: Bin <binclare@iconn.net> Newsgroups: comp.sys.next.programmer Subject: error message when booting,need help Date: Wed, 02 Oct 1996 21:25:38 -0400 Organization: i-conn Message-ID: <32531612.5EDA@iconn.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit can anyone help me with an error message i get when i try to start my NeXT Cube? it does not boot into the os, it gets stuck on the "loading from disk" screen and searches for the "boot_rc" did i delete something? what is wrong. i am a new user of the next system can anyone help? please, please try to help, email me -thanx
From: atang@io.org (Andrew Tang) Newsgroups: comp.sys.next.software,comp.sys.next.programmer,comp.sys.next.misc Subject: Does Enterprise Objects Framework 2.0 support Java? Date: Thu, 03 Oct 1996 02:42:32 GMT Organization: Internex Online (io.org), Toronto, Ontario, Canada Message-ID: <32532460.1699515@news.io.org> In WebObjects Enterprise 3.0/Enterprise Objects Framework 2.0, can I write ALL (100%) of my business objects in Java? Basically, we'd like to do all of our implementation in Java; we do not want to write even one line of Objective-C code. Is this possible with WebObjects Enterprise 3.0? I have been trying to get a definitive answer from NeXT for the past week. I left several phone messages and sent e-mail messages to the appropriate people, and what I got is simply two words: NO REPLY. Obviously those guys are quite busy minding their own businesses while I am getting frustrated trying to get a straight answer from them. So I am trying the USENET news community as a last resort. Any input will be greatly appreciated. Thanks! Andrew Tang (atang@io.org)
From: sanjeev@ee.umr.edu (Sanjeev Agarwal) Newsgroups: comp.sys.next.programmer Subject: Help .... Date: 3 Oct 1996 03:11:02 GMT Organization: UMR Missouri's Technological University Message-ID: <52vas6$f56@hptemp1.cc.umr.edu> Hi, I have a software problem. I am working with NextStep 3.3 and I have some features (Image features) on an Image which I want to move around using mouse. At present I am using mouse-down event (double click) to select a feature and then drag it. On dragging I use the location of the event to find the image coordinates. However this gives me Image coordinates in Integers (i.e. non fractions). In order to get exact match I would like to be able to move the features by a fraction of the pixel also. How can I do this. Any help please .... Thank you very much .... Sanjeev Agarwal P.S. I hope I am able to convey what I am trying to do, if not please send me mail at sanjeev@isc.umr.edu
From: wellin@telospub.com (Paul R. Wellin) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.programmer Subject: Compile problems under 3.3 Date: 3 Oct 1996 04:20:09 GMT Organization: Pagesat, Inc. Message-ID: <52vetp$g4n@jolt.pagesat.net> I am having some troubles compiling certain apps and the problems seem to point to the same source. On my black turbo with 3.3, during a compile (of both Gatekeeper and samba) I get the following error: django:6# make Using CFLAGS = -O -DSMBLOGFILE="/usr/local/samba/var/log.smb" ... stuff... Compiling server.c /NextDeveloper/Headers/netdb.h:66: header file 'sys/bitypes.h' not found includes.h:591: warning: could not use precompiled header '/NextDeveloper/Headers/bsd/libc.p', because: includes.h:591: warning: header 'arpa/inet.h' overridden, includes.h:591: warning: /NextDeveloper/Headers/arpa/inet.h vs. /NextDeveloper/Headers/bsd/arpa/inet.h (within the precomp) /NextDeveloper/Headers/arpa/inet.h:68: header file 'sys/bitypes.h' not found *** Exit 1 Stop. I can find the file bitypes.h nowhere on my system and during each of my two recent compiles, this is where the Exit occurs. Am I doing something rather unintelligent here? Has anyone seen this before or have a clue what I can do to prevent this? Thanks, Paul Wellin
Newsgroups: comp.sys.next.programmer Organization: Antigone Press gateway, San Francisco Return-Path: <zleo@dns.istsan.interbusiness.it> Message-ID: <1.5.4.32.19961003054427.006772b0@istsan.interbusiness.it> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Thu, 03 Oct 1996 06:44:27 +0100 From: Zanitti Leo <zleo@dns.istsan.interbusiness.it> Subject: daemon Hello I am developing a program for monitoring my UPS This application consists of two parts: 1) Power.daemon: a daemon responsibility is to check the status of UPS, provide syslog messages, shutdown when the battery runs low 2) PowerMonitor.app : a App only for monitoring the UPS (only report data to the video) The daemon and the App comunicate through Message (implemented with Listener and Speaker) I would try my Power.damon with Terminal App (for debug) is it possible that PowerMonitor.app find the Power.daemon running in Terminal Window? my Power.daemon comunicates to console too, but on my account when I select Console from Workspace nothing happens, why? Thanks ************************************************************************ * Zanitti Leo * * Viale Regina Elena, 299 * * I-00161 ROME * * * * E-Mail: zleo@istsan.interbusiness.it * * * * TEL +39 6 82.09.70.77 * * TEL +39 6 49.90.24.10 * ************************************************************************
From: david@pfi.ibk.baum.ethz.ch (David C. EKCHIAN) Newsgroups: comp.sys.next.programmer,comp.lang.objective-c Subject: multiple declarations for method `addObject:' Date: 3 Oct 1996 14:58:15 GMT Organization: Swiss Federal Institute of Technology (ETHZ) Message-ID: <530ka7$7tc@elna.ethz.ch> Hi all, I get this message at compilation: gi_Graphics.m:219: warning: multiple declarations for method `addObject:' /NextLibrary/Frameworks/System.framework/Headers/objc/List.h: In function `': /NextLibrary/Frameworks/System.framework/Headers/objc/List.h:49: warning: using `-addObject:anObject' /NextLibrary/Frameworks/Foundation.framework/Headers/NSSet.h: In function `': /NextLibrary/Frameworks/Foundation.framework/Headers/NSSet.h:100: warning: also found `-(void)addObject:(id)object' /NextLibrary/Frameworks/Foundation.framework/Headers/NSSet.h: In function `': /NextLibrary/Frameworks/Foundation.framework/Headers/NSSet.h:58: warning: also found `-(void)addObject:(id)object' /NextLibrary/Frameworks/Foundation.framework/Headers/NSArray.h: In function `': /NextLibrary/Frameworks/Foundation.framework/Headers/NSArray.h:73: warning: also found `-(void)addObject:(id)anObject' /NextLibrary/Frameworks/Foundation.framework/Headers/NSAutoreleasePool.h: In function `': /NextLibrary/Frameworks/Foundation.framework/Headers/NSAutoreleasePool.h:18: warning: also found `-(void)addObject:(id)anObject' any idea how to get rid of this? They are declared in the frameworks, I'm certainly not the first one to get these warnings. Thanks, David. -- o _ /-;c ___ David C. EKCHIAN ______________________________________(@)#\(@)___
From: edx@cc.usu.edu Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: cancel <1996Oct3.081316.86050@cc.usu.edu> Message-ID: <1996Oct3.081558.86051@cc.usu.edu> Date: 3 Oct 96 08:15:58 MDT Control: cancel <1996Oct3.081316.86050@cc.usu.edu> cancel <1996Oct3.081316.86050@cc.usu.edu>
Newsgroups: comp.sys.next.programmer From: stes@cwi.nl (David Stes) Subject: Re: A rubix Cube in Pascal...HELP!!!! Message-ID: <DypJ5u.2Gx@cwi.nl> Sender: news@cwi.nl (The Daily Dross) Organization: CWI, Amsterdam References: <51pn2q$4l7@dfw-ixnews10.ix.netcom.com> <51v9fv$nd3@news2.snfc21.pacbell.net> <52u638$rli@mimsy.cs.umd.edu> Date: Thu, 3 Oct 1996 16:01:05 GMT In article <52u638$rli@mimsy.cs.umd.edu> seanl@carmi.cs.umd.edu (Sean Luke) writes: >Mike Paquette (mpaque@pacbell.net) wrote: >>johnyreb@ix.netcom.com(Richard Elbert Davis) wrote: > >>>hello, >>> In pascal class right now i have to make a Rubix cube with >>>multi-dimensional arrays. Anyone have a sample of a prog I could use. >>>Its got me stumped. Thanks alot Also check out GAP (Groups, Algorithms, Programming) by Neubuser and Schoenert. The readme of their computer algebra system for group theory (with a Pascal like programming language and a kernel optimized for permutation and integer arithmetic) is exaclty analyzing in all detail Rubik's Cube... It used to compile on NeXT 3.0. And it's a very portable system. So might still help you out if you're on another NextStep system... ftp is (or was in 1992) samson.math.rwth-aachen.de
From: "Sam R. Thangiah" <thangiah@samuel.cpsc.sru.edu> Newsgroups: comp.sys.next.programmer Subject: C++ rot 4.0 OpenSTEP Date: Thu, 03 Oct 1996 13:50:13 -0700 Organization: Slippery Rock University Message-ID: <32542705.7BCF@samuel.cpsc.sru.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I am trying to get the C++ compiler running on the Openstep (4.0) but it does not seem to like either the c++ command or the file with .c extension. Could somebody give me some pointers on how to get C++ running for the 4.0 Thanks, Sam sam.thangiah@sru.edu or thangiah@samuel.cpsc.sru.edu
From: ckane@next.com (Christopher Kane) Newsgroups: comp.sys.next.programmer Subject: Re: Converting DPSAddFD() to OpenStep 4.0? Date: 3 Oct 1996 18:44:47 GMT Organization: NeXT Software, Inc. Distribution: world Message-ID: <5311iv$oau@news.next.com> References: <Dy6vB8.1CB.0.astra@now.ch> In article <Dy6vB8.1CB.0.astra@now.ch> headi@now.ch (Daniel Scheidegger) writes: > does anybody have an idea how to convert calls of DPSAddFD() to OS4.0? > The ConversionGuide mentions NSRunLoop's addPosixFileDescriptor, > which does only exist in the Conversion Frameworks. > I needed the DPSAddFD to have asynchronous input from TCP/IP sockets... Use NSFileHandle, and -readInBackgroundAndNotify. ------------------------ - (void)readInBackgroundAndNotify Performs an asynchronous availableData operation on a file or communications channel and posts an NSFileHandleReadCompletionNotification to the client process' run loop. The length of the data is limited to the buffer size of the underlying operating system. The notification includes a userInfo dictionary which contains the data read; access this object using the NSFileHandleNotificationDataItem key. Any object interested in receiving this data asynchronously must add itself as an observer of NSFileHandleReadCompletionNotification. In communication via stream-type sockets, the receiver is often the object returned in the userInfo dictionary of NSFileHandleConnectionAcceptedNotification. See also: - acceptConnectionInBackgroundAndNotify, - enqueueNotification:postingStyle:coalesceMask:forModes: (NSNotificationQueue), ------------------------ p NSFileHandleReadCompletionNotification Notification Object The posting NSFileHandle userInfo Key Value NSFileHandleNotificationDataItem An NSData containing the available data read from a socket connection. NSFileHandleNotificationMonitorModes An NSArray containing the run-loop modes in which the notification can be posted. To cause the posting of this notification, you must send either readInBackgroundAndNotify or readInBackgroundAndNotifyForModes: to an appropriate NSFileHandle. This notification is posted when the background thread reads the data currently available in a file or at a communications channel. It makes the data available to observers by putting it in the userInfo dictionary. ------------------------ Note that this is a one-shot operation (ie, the registration is for one read+notification). To "continuously" read in background, the notification handler should decide if it wants to continue background reading on the file handle, and "re-register" the file handle (which is the "object" of the notification object parameter to the handler routine) for another background read by sending it -readInBackgroundAndNotify. > Daniel Scheidegger > Software Engineer, System Administrator > NOW GmbH, Scheideggstr. 73, CH-8038 Zuerich > ++41-1-2898025 / dscheide@now.ch Christopher Kane NeXT Software, Inc.
From: lacourse@midusa.net (Dan LaCourse) Newsgroups: comp.sys.next.programmer Subject: Re: A rubix Cube in Pascal...HELP!!!! Date: 3 Oct 1996 19:04:05 GMT Organization: Netspace Internet Services Message-ID: <5312n5$4fu1@news.midusa.net> References: <Pine.NXT.3.95.960914005601.6904F-100000@charisma> <MmCho3u00WBO8141Av@andrew.cmu.edu> <51pn2q$4l7@dfw-ixnews10.ix.netcom.com> <51v9fv$nd3@news2.snfc21.pacbell.net> <52u638$rli@mimsy.cs.umd.edu> Mike Paquette (mpaque@pacbell.net) wrote: >Well, a Rubic's Cube is a cube with a 3x3 array of colors for each face, so the obvious internal representation would be to use a 8x3x3 array. ^ A cube has six faces! -- Best wishes, Dan LaCourse
From: eric@skatter.USask.Ca Newsgroups: comp.sys.next.programmer Subject: copy vs. retain when unarchiving (OpenStep 4.0) Date: 3 Oct 1996 20:12:37 GMT Organization: University of Saskatchewan Message-ID: <5316nl$t47@tribune.usask.ca> The OPENSTEP 4.0 document ``ImplementingASubclassofNSObject'' gives the following example for unarchiving a subclass of NSObject: - (id)initWithCoder:(NSCoder *)coder { name = [[coder decodeObject] copy]; transportation = [[coder decodeObject] copy]; hotels = [[coder decodeObject] copy]; languages = [[coder decodeObject] copy]; [coder decodeValueOfObjCType:"s" at:&englishSpoken]; currencyName = [[coder decodeObject] copy]; [coder decodeValueOfObjCType:"f" at:¤cyRate]; return self; } I've got a couple of questions about this: 1) Why use `copy' instead of `retain' after decoding all the objects? 2) Shouldn't this method have a`[super init];' at the beginning? -- Eric Norum eric@skatter.usask.ca Saskatchewan Accelerator Laboratory Phone: (306) 966-6308 University of Saskatchewan FAX: (306) 966-6058 Saskatoon, Canada. NeXTMail accepted.
From: dekorte@suite.com (Steve Dekorte) Newsgroups: comp.lang.objective-c,comp.sys.next.programmer Subject: Class Clusters Date: 3 Oct 1996 20:32:01 GMT Organization: Suite Software Message-ID: <5317s1$pmt@news.onramp.net> [Class Clusters] > ... > On the plus side, I see > + simple uniform interface to similar concrete classes > + efficient use of memory and time (in most circumstances) How is either of these not true of subclasses? > + hide complexity from user I disagree. Clusters are complex. Here you say it hides complexity, but below you say it's tricky? Not subclassing at all and using delegation for "inherited" behaivor can be simpler in some ways, but clearly having multiple inheritance mechanisms creates more complexity than it reduces. > + separate mutable from immutable instances > + ability to revise concrete classes without impacting callers Ok. This situation isn't much different from regular subclasses though. > On the minus side, I see > - tricky, (but possible), to provide new subclasses > - difficult to provide both mutable and immutable subclasses > without replication > - a little more complex to debug, archive custom subclasses I agree with these. > Given the subclassing complexity, I'd say class clusters have > merit as a design approach when: > the class hierarchy is fairly flat (not much subclassing) > the best storage/implementation requirements vary dramatically > depending upon concrete details > users don't need to care about the concrete details, only > the behavior defined in the abstract interface Well, the clusters don't matter for deep hierachies. It's only when you subclass straight off the cluster that you have to deal with them, so "flat" hierarchies are actually the most difficult. Now if your just saying that clusters don't cause problems when you don't use them, well, that's kind of obvious. > Conversely, class clusters are probably a bad idea when: > the inheritance hierarchy is deep > users need to know representation details about individual subclasses. Steve
From: clientserver@msn.com (Richard Goode) Subject: L@@K__> OB-C/ WebObjects > N. VA - WASH DC Date: 3 Oct 96 20:30:32 -0700 Message-ID: <00001c42+00000d65@msn.com> Newsgroups: comp.sys.next.programmer Organization: The Microsoft Network (msn.com) Client/Server Resources has cutting edge opportunities in the Washington DC Metro Area for: NeXTStep Developers Responsibilities include design and development of the common object model. Work with other project teams to solidify the design of the common object model through the following development cycles: Requirement Analysis Functional Design Technical Design Construction Application Testing Qualifications: Application of OO design techniques and methodologies 3+ yrs C++ and/or Objective C programming experience 1+ yr UNIX Operating System experience NeXTStep, OpenStep, and Windows NT Operating System experience a plus Knowledge of major RDBMS Experience and/or strong interest in WebObjects e-mail your resume TODAY!!!: clientserver@msn.com Fax: (301) 983-4728 Snail mail to: Client/Server Resources P.O. Box 61351 Potomac, Maryland 20859-1351 Tel: (301) 983-6942 Fax: (301) 983-4728 e-mail: clientserver@msn.com
From: jeremy@inetnebr.com (Jeremy Bettis) Newsgroups: comp.sys.next.programmer Subject: NSTextView under windows NT Date: 4 Oct 1996 01:17:04 GMT Organization: Internet Nebraska Message-ID: <531oig$sgu@hawk.inetnebr.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii NNTP-Posting-User: jeremy Does anyone else have problems with the NSTextView not displaying all the text under WinNT? I often do, I will put a bunch of text in the text view, and the scrollers will not grow. So the only way to get to the bottom of the text is to click in the text and arrow down, or click and drag downward in the text itself. I am using the appendAttributedString and setString methods of NSTextStorage to add the text. -- Jeremy Bettis -*- PGP public key available,send mail with subj "Send pgp key" Home: jeremy@tddi.inetnebr.com Work: jeremy@hksys.com NeXTMAIL/MIME/PGP accepted
From: Mark Anenberg <marka@Eng.Sun.COM> Newsgroups: comp.sys.next.programmer Subject: Re: Converting DPSAddFD() to OpenStep 4.0? Date: Thu, 03 Oct 1996 18:47:13 -0700 Organization: Sun Microsystems Inc. Message-ID: <32546CA1.C9F@Eng.Sun.COM> References: <Dy6vB8.1CB.0.astra@now.ch> <5311iv$oau@news.next.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Christopher Kane wrote: > > In article <Dy6vB8.1CB.0.astra@now.ch> headi@now.ch > (Daniel Scheidegger) writes: > > does anybody have an idea how to convert calls of DPSAddFD() to OS4.0? > > The ConversionGuide mentions NSRunLoop's addPosixFileDescriptor, > > which does only exist in the Conversion Frameworks. > > I needed the DPSAddFD to have asynchronous input from TCP/IP sockets... > > Use NSFileHandle, and -readInBackgroundAndNotify. > > ------------------------ > - (void)readInBackgroundAndNotify > > Performs an asynchronous availableData operation on a file or > communications channel and posts an NSFileHandleReadCompletionNotification > to the client process' run loop. The length of the data is limited to the > buffer size of the underlying operating system. The notification includes > a userInfo dictionary which contains the data read; access this object > using the NSFileHandleNotificationDataItem key. > > Any object interested in receiving this data asynchronously must add > itself as an observer of NSFileHandleReadCompletionNotification. In > communication via stream-type sockets, the receiver is often the object > returned in the userInfo dictionary of > NSFileHandleConnectionAcceptedNotification. > > See also: - acceptConnectionInBackgroundAndNotify, > - enqueueNotification:postingStyle:coalesceMask:forModes: > (NSNotificationQueue), > ------------------------ > p NSFileHandleReadCompletionNotification > > Notification Object The posting NSFileHandle > > userInfo > > Key Value > > NSFileHandleNotificationDataItem An NSData containing the available > data read from a socket connection. > > NSFileHandleNotificationMonitorModes > An NSArray containing the run-loop modes in which the notification can be > posted. > > To cause the posting of this notification, you must send either > readInBackgroundAndNotify or readInBackgroundAndNotifyForModes: to an > appropriate NSFileHandle. This notification is posted when the background > thread reads the data currently available in a file or at a communications > channel. It makes the data available to observers by putting it in the > userInfo dictionary. > ------------------------ > > Note that this is a one-shot operation (ie, the registration is for one > read+notification). To "continuously" read in background, the > notification handler should decide if it wants to continue background > reading on the file handle, and "re-register" the file handle (which is > the "object" of the notification object parameter to the handler routine) > for another background read by sending it -readInBackgroundAndNotify. > > > Daniel Scheidegger > > Software Engineer, System Administrator > > NOW GmbH, Scheideggstr. 73, CH-8038 Zuerich > > ++41-1-2898025 / dscheide@now.ch > > Christopher Kane > NeXT Software, Inc. One word of caution; Whilest this is the way to do in NeXT's OpenStep 4.0, Solaris OpenStep retains the NSPosixFileDescriptor model and we do not currently support NSFileHandle. -Mark Anenberg
From: Mark Anenberg <marka@Eng.Sun.COM> Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: Re: Printf %g format still broken in OPENSTEP 4.0 Date: Thu, 03 Oct 1996 18:56:00 -0700 Organization: Sun Microsystems Inc. Message-ID: <32546EB0.3CD6@Eng.Sun.COM> References: <52p5vp$1sq@tribune.usask.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit eric@skatter.USask.Ca wrote: > > I just upgraded my machine (486DX2/66) to OPENSTEP 4.0. > One of the first things I tried was my printf `%g' format test program: > > #include <stdio.h> > > int main (int argc, char **argv) > { > printf ("Expect 0.00123: %.3g\n", 0.001234567); > printf ("Expect 123: %.3g\n", 123.4567); > printf ("Expect 123.5: %.4g\n", 123.4567); > printf ("Expect 1e+03: %.3g\n", 999.6); > return 0; > } > > Here's the output after compiling the program on NEXTSTEP 4.0: > Expect 0.00123: 0.001 > Expect 123: 123.457 > Expect 123.5: 123.4567 > Expect 1e+03: 999.6 > > I can hardly believe this. It is now *SEVEN YEARS* since I first reported this bug (in > NextStep 1.0). I have reported this error in every subsequent release up to and > including NEXTSTEP 3.3. In May, 1994 I received mail from NeXT saying that this would be > fixed in 4.0. Looks like they missed it again. > > Every other time I have posted an article about this bug I have received mail saying that > the behaviour of the NeXT printf is correct. To try and forestall these replies I > include the following: > > The ANSI standard (X3.159-1989) says (Section 4.9.6.1, page 134, line 33): >.... > > -- > Eric Norum eric@skatter.usask.ca > Saskatchewan Accelerator Laboratory Phone: (306) 966-6308 > University of Saskatchewan FAX: (306) 966-6058 > Saskatoon, Canada. NeXTMail accepted. You could always buy a real Operating system like Solaris and use Solaris Openstep.... (sorry I couldn't resist the jab). Here is the output of your program on Solaris 2.5. /tmp:503> cc x.c /tmp:504> a.out Expect 0.00123: 0.00123 Expect 123: 123 Expect 123.5: 123.5 Expect 1e+03: 1e+03 Mark Anenberg , OpenStep Development Team, Sun Microsystems, Inc. [E,NeXT,Mime]Mail: mark@velodrome.com, http://www.velodrome.com/ Disclaimer: The opinions expressed above are my own and in no way represent those of Sun Microsystems, Inc.
From: Constantin Szallies <szallies@energotec.de> Newsgroups: comp.sys.next.programmer Subject: Re: copy vs. retain when unarchiving (OpenStep 4.0) Date: 4 Oct 1996 08:30:38 GMT Organization: Tech Net GmbH Message-ID: <532hve$nbs@ddfservb.technet.net> References: <5316nl$t47@tribune.usask.ca> <531s97$81e@dfw-ixnews4.ix.netcom.com> aisbell@ix.netcom.com (Art Isbell) wrote: >eric@skatter.USask.Ca wrote: >> The OPENSTEP 4.0 document ``ImplementingASubclassofNSObject'' gives the >> following example for unarchiving a subclass of NSObject: >> >> - (id)initWithCoder:(NSCoder *)coder >> { >> name = [[coder decodeObject] copy]; >> transportation = [[coder decodeObject] copy]; >> hotels = [[coder decodeObject] copy]; >> languages = [[coder decodeObject] copy]; >> [coder decodeValueOfObjCType:"s" at:&englishSpoken]; >> currencyName = [[coder decodeObject] copy]; >> [coder decodeValueOfObjCType:"f" at:¤cyRate]; >> return self; >> } >> >> 1) Why use `copy' instead of `retain' after decoding all the objects? > > The only reason I could figure when I read this was consistency. NeXT >has suggested that, in general, value instance variables should be >independent copies, not shared (i.e., retained) with other objects. So >presumably, if setName: were implemented in the above class: > >- (void)setName:(NSString *)aName >{ > [name autorelease]; > name = [aName copy]; >} > > So to be consistent, if aName is copied in setName:, then the decoded >object should be copied in initWithCoder:. Note that copying an immutable >object can be implemented by merely incrementing its retain count - i.e., >just as if it were retained. If you check some class generated by EOModeller, the setXyz: always looks like this: - (void)setName:(NSString *)aName { [name autorelease]; name = [aName retain]; } Sounds like even some people at NeXT are not sure when to use "copy" and when "retain"! My recommandation: Use "copy" in a setIVar method if the new parameter is no object container. If it's a container, use "retain". Use "retain" in the "initWithCoder:" methods. This avoids unnessary copying of objects and should by safe(?). -- Constantin Szallies, Energotec GmbH szallies@energotec.de 49211-9144018
From: "James P. Klett" <klett002@maroon.tc.umn.edu> Newsgroups: comp.sys.next.programmer Subject: Project Builder Crashes System Over and Over Date: Fri, 04 Oct 1996 02:33:18 -0500 Organization: Emprise Consulting Services, Inc. Message-ID: <3254BDBE.61FD@maroon.tc.umn.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I am in the Conversion process of an app to 4.0 and Project Builder is causing my system to lock to which I must re-boot, fsck, and re-boot again. It does this over and over only at different places, sometimes compiling, sometimes indexing, sometimes just selecting a file in the browser, anyway it has done it about 8 times in 24hours, and it take my system about 10-15 minutes to restart up so I'm losing valuable time. Is there anybody else with this problem? Can it be fixed? Thanks for any help, JIM -- ___ ___....-----'---'-----....___ ========================================= ___'---..._______...---'___ (___) _|_|_|_ (___) \\____.-'_.---._'-.____// klett002@maroon.tc.umn.edu
From: Constantin Szallies <szallies@energotec.de> Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: Re: detachNewThreadSelector... retains its target (?) Date: 4 Oct 1996 08:57:38 GMT Organization: Tech Net GmbH Message-ID: <532ji2$nbs@ddfservb.technet.net> References: <52tcib$22l@isabella.object-factory.com> <52u5fc$h9b@tribune.usask.ca> eric@skatter.USask.Ca wrote: > Here's what I did in my xmodem file transfer object: > >- (void)sendCompletionMessage:(NSString *)message >{ > [completionStatus release]; > completionStatus = message; > [completionStatus retain]; > xmodemThread = [NSThread currentThread]; > [self autorelease]; /* This matches the retain done by dettachThread */ > [xmodemThread exit]; >} Good coding practice would be: Define this somewhere: // There's a bug in NSThread... #define NSTHREAD_EXTRA_RETAIN_BUG - (void)sendCompletionMessage:(NSString *)message { [completionStatus release]; completionStatus = message; [completionStatus retain]; xmodemThread = [NSThread currentThread]; #ifdef NSTHREAD_EXTRA_RETAIN_BUG [self autorelease]; /* This matches the retain done by dettachThread */ #endif [xmodemThread exit]; } This makes maintainance somewhat easier if the bug is fixed by NeXT. You may want to add something like this to the company standard coding rules. -- Constantin Szallies, Energotec GmbH szallies@energotec.de 49211-9144018
From: Constantin Szallies <szallies@energotec.de> Newsgroups: comp.sys.next.programmer Subject: Storing an "NSObject" as "Object": A good idea to do it this way? Date: 4 Oct 1996 13:36:04 GMT Organization: Tech Net GmbH Message-ID: <5333s4$qki@ddfservb.technet.net> I have to store an Enterprice Object (a "pure" NSObject object net) in an old database system which uses "Object"'s "read" and "write" to store "Object"s as BLOBs. As the "HybridWorld" documents recommends, I don't want to store the "NSObject"s by just implementing the old "read" and "write" methods. I came up with this idea: I write an NSObjectWrapper class inheriting from "Object" with a single ivar called "wrappedNSObject" , a pointer to some retained object inherting from NSObject. If NSObjectWrapper implements "write" this way: - write:(NXTypedStream *)stream; { [super write:stream]; NXWriteNSObject(stream, wrappedObject); return self; } and read this way: - read:(NXTypedStream *)stream; { [super read:stream]; wrappedObject = [(NSObject *)NXReadNSObject(stream) retain]; return self; } I can save NSObjects in my "Object" database! Can I? Or did I miss something? -- Constantin Szallies, Energotec GmbH szallies@energotec.de 49211-9144018
From: Joakim Johansson <jocke@rat.se> Newsgroups: comp.sys.next.programmer Subject: Re: DO problem under NEXTSTEP 3.3 "[NXConnection run] - tossing received reply msg" Date: 4 Oct 1996 10:57:58 GMT Organization: Research & Trade AB Distribution: world Message-ID: <532qjm$ctv@baldwin.rat.se> References: <52ot2m$gb1@ddfservb.technet.net> Constantin Szallies <szallies@energotec.de> writes > Looks like the PDO system doesn't add a flag to each request if the message > is oneway or not. So the server thinks it's a syncrounious message and sends > back a reply. Sorry for the confusion, but it is a asynch message from the PDO server to a NS client (which sends back the reply). The PDO system doesn't know whether the client explicitly conforms to the protocol, and shouldn't care. (it's the NS client which erroneously returns a reply message to the asynch request) > What happens if you send a message and the sender wants a proxy > and the server returns a copy (or the other way around)? Could you please rephrase the question? > > I guess one should view > >(that is: "@interface X:Object // <protocolXDO> oops!") > > as a bug in the application program. I dont think so, I haven't seen any references in the docs (those that exist... ;-) that the receiving proxy actually has to explicitly conform to the protocol. Anyway, it's not a serious problem - it was just a suggestion on a reason why <dennis.glatting@plaintalk.bellevue.wa.us> might get that warning message. (which can be quite annoying until you find out what causes it...) Cheers, Joakim -- Joakim Johansson Software Developer, Research & Trade jocke@rat.se <NeXTmail, MIME> http://www.rat.se/
From: gen kiyooka <gen@digigami.com> Newsgroups: comp.sys.next.programmer Subject: OmniWeb plugins? Date: Fri, 04 Oct 1996 06:44:35 -0700 Organization: Digigami, Inc. Message-ID: <325514C3.40A4@digigami.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit all, I'm a PC programmer who just bought black hardware. I had the idea of tinkering with writing a plugin for OmniWeb (I thought I read somewhere that OmniWeb had a plugin architecture of sorts). Can anyone point me toward any info on this? I'm also interested in writing a simple NeXTtime app. Any pointers to FTP source code for, like, a hello world type NeXTtime app? tia, gen
From: gen kiyooka <gen@digigami.com> Newsgroups: comp.sys.next.programmer Subject: Re: NeXTStep Libraries on top of X? Date: Fri, 04 Oct 1996 06:36:48 -0700 Organization: Digigami, Inc. Message-ID: <325512F0.7933@digigami.com> References: <DyJ48L.9xM@devetir.qld.gov.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Stephen Hocking wrote: > > I'm wondering if anyone has implementations (source, of course!) of NeXTStep > lookalike libraries layered on X. I'm looking to make a quick & dirty port of > id's Quake Editor - *blush*Just a thought from a new owner of black hardware, but isn't GNU Step supposed to be an OpenStep implementation that runs on top of X? I had a friend forward me an email that GNUStep were shipping a round of devtools on CDROM. gen
From: Samuel_Figueroa@NeXT.com (Samuel Figueroa) Newsgroups: comp.sys.next.programmer Subject: Re: Which GCC version under 4.0 Date: 4 Oct 1996 16:05:03 GMT Organization: NeXT Software, Inc. Message-ID: <533cjf$1fa@news.next.com> References: <52h84k$72o@news2.snfc21.pacbell.net> Mike Paquette writes > edx@cc.usu.edu wrote: > >The Release notes for 4.0 say that the compiler is based > >on gcc version 2.5.8 > >I guess NeXT isn't in the compiler business any more either. > Just as a point of information, if one were going to redo the code > generator of a compiler, to , oh say, generate highly optimized > position independent code for several architectures, and one knew that > several million lines of code would have to be compiled correctly by > that compiler, then one might want to stop adding random changes from > third parties some time before the compiler was needed for production > work, and concentrate on the tasks at hand. > Another point: Just because the gcc version says 2.5.8 doesn't mean > that later gcc bug fixes wouldn't be incorporated. Many fixes are > incorporated where they make sense. I'm not an official spokesperson for NeXT, but I agree with Mike. NeXT has shipped compilers based on gcc 2.7.2 for some of our platforms, but these are buggier than we would like (though still perfectly usable). On Mach, we can't afford this level of instability, so we will not ship a gcc 2.7.2-based compiler for Mach until it is ready. - Sam Figueroa (figueroa@next.com)
From: headi@now.ch (Daniel Scheidegger) Newsgroups: comp.sys.next.programmer Subject: Re: Converting DPSAddFD() to OpenStep 4.0? Date: Fri, 4 Oct 1996 09:17:24 GMT Organization: NOW GmbH, Baar, Switzerland Message-ID: <Dyqv50.IBs.0.astra@now.ch> References: <Dy6vB8.1CB.0.astra@now.ch> <5311iv$oau@news.next.com> In article <5311iv$oau@news.next.com>, Christopher Kane <ckane@next.com> wrote: >Use NSFileHandle, and -readInBackgroundAndNotify. [description deleted] Thanks for the info. However, I figured that out myself during the last week. Especially the one-shot thing took quite some time to find out. Maybe I should send NeXT a bill :-)))) Is there any way to get the complete documentation of NSFileHandle? Why isn't it documented at all in OS4.0? From the responses I got, I was not alone with the problem... PS: thanks to Mark for the update concerning Sun's implementation. If we really had really portable software, some of us had nothing more to do and loose the job, right ? Kind regards, Daniel -- Daniel Scheidegger Software Engineer, System Administrator NOW GmbH, Scheideggstr. 73, CH-8038 Zuerich ++41-1-2898025 / dscheide@now.ch
From: headi@now.ch (Daniel Scheidegger) Newsgroups: comp.sys.next.programmer Subject: IconKit for OpenStep, anyone? Date: Fri, 4 Oct 1996 09:58:31 GMT Organization: NOW GmbH, Baar, Switzerland Message-ID: <Dyqx1K.JI1.0.astra@now.ch> Hi all, I just wanted to know whether someone is porting or has ported the IconKit to OpenStep. If not, I would try to do it, because I need such a beast... Kind regards, Daniel -- Daniel Scheidegger Software Engineer, System Administrator NOW GmbH, Scheideggstr. 73, CH-8038 Zuerich ++41-1-2898025 / dscheide@now.ch
From: "James P. Klett" <klett002@maroon.tc.umn.edu> Newsgroups: comp.sys.next.programmer Subject: How to run project server on another host? Date: Fri, 04 Oct 1996 12:31:57 -0500 Organization: Emprise Consulting Services, Inc. Message-ID: <32554A0D.211C@maroon.tc.umn.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Does anyone know how to run the indexing project server for PB on another host?? Thanks, JIM -- ___ ___....-----'---'-----....___ ========================================= ___'---..._______...---'___ (___) _|_|_|_ (___) \\____.-'_.---._'-.____// klett002@maroon.tc.umn.edu
From: eric@skatter.USask.Ca Newsgroups: comp.sys.next.programmer Subject: Problem with colors after OPENSTEP conversion Date: 4 Oct 1996 19:04:40 GMT Organization: University of Saskatchewan Message-ID: <533n48$7qk@tribune.usask.ca> I'm busy converting an application from NS 3.3 to OS 4.0. Until now, everything has gone more smoothly than I had expected. The conversion scripts work very well. I've run into a problem, though. Whenever my application tries to print a document on one of the Apple LaserWriters here or on a NeXTstation (running NS 3.3) color printer I get: From the LaserWriter -- Nothing From the NeXTstation -- A complaint about a PostScript error I ran Yap on the NeXTstation and found that the offending command is: setuseprecisecolor Where is this command coming from? Why doesn't the print package define it? Here are the first few lines of my file: %!PS-Adobe-2.0 %%Title: W802-1 %%Creator: FlyingWire %%CreationDate: Fri Oct 4 10:49:56 1996 %%For: eric %%DocumentFonts: (atend) %%Pages: (atend) 1 %%BoundingBox: (atend) %%DocumentPaperSizes: Letter %%Orientation: Landscape %%NXNextStepVersion: 3.0 %%EndComments %%BeginProcSet: /NextLibrary/Frameworks/AppKit.framework/Versions/B/Resources/printPackageJ.p s 3.0 %! % % printPackageJ.ps % Application Kit % Copyright (c) 1988-1996, NeXT Software, Inc. % All rights reserved. % -- Eric Norum eric@skatter.usask.ca Saskatchewan Accelerator Laboratory Phone: (306) 966-6308 University of Saskatchewan FAX: (306) 966-6058 Saskatoon, Canada. NeXTMail accepted.
From: mpaque@pacbell.net (Mike Paquette) Newsgroups: comp.sys.next.programmer Subject: Re: Help .... Date: Fri, 04 Oct 1996 19:05:17 GMT Organization: Electronics Service, Unit No. 16 Message-ID: <533nd4$1t8@news2.snfc21.pacbell.net> References: <52vas6$f56@hptemp1.cc.umr.edu> sanjeev@ee.umr.edu (Sanjeev Agarwal) wrote: >Hi, > I have a software problem. I am working with NextStep 3.3 >and I have some features (Image features) on an Image which I >want to move around using mouse. At present I am using >mouse-down event (double click) to select a feature and then >drag it. On dragging I use the location of the event to find >the image coordinates. However this gives me Image coordinates >in Integers (i.e. non fractions). In order to get exact match >I would like to be able to move the features by a fraction of >the pixel also. How can I do this. Any help please .... >Thank you very much .... While represented as floats, the coordinates passed in the event record are always integer values. The mouse driver doesn't deal in fractional pixel movements. What you might try is letting ALT-drag operations scale the mouse movement down. That is, when your drag loop detects that the alt key is down, save the original coordinates, and compute each new position as a scaled down delta off the original position. A scale of 4, for example, would let you position things to the nearest quarter pixel. This sort of fine position scaling works well for those of us who drink too much coffee, or otherwise have shakey hands. For an existing example, look at the behavior of the scroll bars in Terminal or Edit. Try moving them with the ALT key down and up, and you'll see what I mean. Mike Paquette I don't speak for my employer, and they don't speak for me. mpaque@pacbell.net Personal E-mail mpaque@next.com NeXT business mail only, please
From: alexdn@globalobjects.com (Alex D. Nghiem) Newsgroups: comp.sys.next.programmer Subject: US - TX and Southeast - WebObjects and EOF experts needed Date: Fri, 04 Oct 1996 20:29:16 GMT Organization: Global Objects Inc. Message-ID: <32557341.18065374@news-s01.ny.us.ibm.net> Folks: Global Objects specializes in providing quality solutions to its clients through object technology. We're currently looking for folks with WebObjects expertise to work on various projects in the Southeast and TX. These projects will start almost immediately, and we need you now! Qualifications: - WebObjects expertise (preferred) - EOF expertise as well - NextStep development experience - 3 or more years - Modelling experience with Booch/Rumbaugh/Jacobson use cases We will also consider candidates with a strong Java or Smalltalk background. These positions do provide some hands-on WebObjects training but we prefer candidates that already have this expertise. All work will be done on the client site. Join a team of OO experts and work on the state-of-the art projects. Each of the members on this team (including one PhD and one book author) has over 7 years of OO experience (check out our NextStep development book from Prentice Hall, _NextStep Programming: Concepts and Applications, and our recent article on WebObjects in the July issue of Object magazine). For immediate consideration, pls e-mail your resume to jobs@globalobjects.com (preferrably in Word format) or fax it to 770.457.7333. Thanks, - Alex - Global Objects Object and Web Solutions Experts Phone: 770.235.2036 Fax:770.457.4144
From: francis@az.stratus.com (Francis Hartojo) Newsgroups: comp.lang.perl.misc,comp.sys.next.programmer Subject: Perl 5.003 configuration problem? Followup-To: poster Date: 4 Oct 1996 23:30:14 GMT Organization: Stratus Computer, Inc. Message-ID: <5346m6$i1u@transfer.stratus.com> Summary: Perl, NEXTSTEP, NFS. Hi all, I've just got a quick question regarding Perl configurations and might also involve NFS on NeXT 3.3. Here's the situation: I compiled Perl 5.003 on my workstation that runs NEXTSTEP 3.3 and installed it under my home directory (i.e., ~/bin, ~/lib/perl5, etc.) Everything seems to have gone well at this point. Now let me explain what the network setup looks like. People's home directories on their individual NeXTstations are mounted under the /Net/their_workstation_name/his_or_her_name. I can also access my own home directory on my workstation, which is physically located under /francis, through /Net/my_workstation_name/francis. (/Net/my_workstation_name, in fact, is just a link to / on my machine.) Okay. Now, here's the little problem I'm having: I can run my perl script just fine if I put #!/francis/bin/perl at the top. If I put #!/Net/my_workstation_name/francis/bin/perl, then it wouldn't run. It'd give some errors like: "print: Command not found", or "Badly placed ()'s", etc. The same case if I try to run the script from somebody else's workstation. However, I can run the same script by interactively entering the perl command at the command line: % /Net/my_workstation_name/francis/bin/perl script_name from any machine in the LAN. The reason I want to be able to access perl through the /Net directory is because the web server is run on another host and if I write a perl CGI script, I have to put #!/Net/my_workstation_name/francis/bin/perl at the top. It's not going to be able to access #!/francis/bin/perl. I don't know whether this a security feature, which is fine by me but I need to know it for certain, and if there's a way around it, perhaps some through some compilation settings. I'm wondering if anybody else've encountered the same, or similar, problem? I'd appreciate any tips, suggestions, advise you can give me on this matter. Please send your follow-up to me directly (francis@az.stratus.com). I'll post a summary if there're enough responses and interest. I apologize if this is a FAQ. Thank you very much. -- +------------------------+-------------------------------------------------+ | Francis Hartojo | Ph.: 602-852-3195 [Insert cute | | Stratus Computer, Inc. | 800-294-1344 x3195 disclaimer here] | | francis@az.stratus.com | Fax: 602-852-3093 |
From: aisbell@ix.netcom.com (Art Isbell) Newsgroups: comp.sys.next.programmer Subject: Re: copy vs. retain when unarchiving (OpenStep 4.0) Date: 4 Oct 1996 02:20:23 GMT Organization: Netcom Distribution: world Message-ID: <531s97$81e@dfw-ixnews4.ix.netcom.com> References: <5316nl$t47@tribune.usask.ca> eric@skatter.USask.Ca wrote: > The OPENSTEP 4.0 document ``ImplementingASubclassofNSObject'' gives the > following example for unarchiving a subclass of NSObject: > > - (id)initWithCoder:(NSCoder *)coder > { > name = [[coder decodeObject] copy]; > transportation = [[coder decodeObject] copy]; > hotels = [[coder decodeObject] copy]; > languages = [[coder decodeObject] copy]; > [coder decodeValueOfObjCType:"s" at:&englishSpoken]; > currencyName = [[coder decodeObject] copy]; > [coder decodeValueOfObjCType:"f" at:¤cyRate]; > return self; > } > > 1) Why use `copy' instead of `retain' after decoding all the objects? The only reason I could figure when I read this was consistency. NeXT has suggested that, in general, value instance variables should be independent copies, not shared (i.e., retained) with other objects. So presumably, if setName: were implemented in the above class: - (void)setName:(NSString *)aName { [name autorelease]; name = [aName copy]; } So to be consistent, if aName is copied in setName:, then the decoded object should be copied in initWithCoder:. Note that copying an immutable object can be implemented by merely incrementing its retain count - i.e., just as if it were retained. > 2) Shouldn't this method have a`[super init];' at the beginning? From NSObject's init method description: "The version of the init method defined in the NSObject class does no initialization; it simply returns self." self has already been set by the allocation method, so nothing is accomplished by invoking NSObject's init. However, it might be considered good programming style to always be consistent in object initializers and invoke super's designated initializer which is init in this case. -- Art Isbell NeXT/MIME Mail: aisbell@ix.netcom.com Trego Systems Voice/Fax: +1 408 335 2515 CaseServ: OPENSTEP Voice Mail: +1 408 335 1154 managed care solutions US Mail: Felton, CA 95018-9442
From: dave@turbocat.de (David Wetzel) Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: Re: Printf %g format still broken in OPENSTEP 4.0 Date: 5 Oct 1996 17:13:08 GMT Organization: Turbocat's Development, Germany Message-ID: <5364v4$34r@turbocat.turbocat.de> References: <52p5vp$1sq@tribune.usask.ca> <32546EB0.3CD6@Eng.Sun.COM> Mark Anenberg <marka@Eng.Sun.COM> wrote: (...) > You could always buy a real Operating system like Solaris and use > Solaris Openstep.... (sorry I couldn't resist the jab). > > Here is the output of your program on Solaris 2.5. (...) alice# cc bla.c -o bla alice# ./bla Expect 0.00123: 0.00123 Expect 123: 123 Expect 123.5: 123.5 Expect 1e+03: 1e+03 alice# echo $OSTYPE NetBSD (I got it in some hours via modem) The source can be found on ftp.NetBSD.org and its mirrors. Maybe NeXT manages to port some of the stuff to the NeXT release. _ _ _(_)(_)_ David Wetzel, Turbocat's Development, (_) __ (_) Buchhorster Strasse, D-16567 Muehlenbeck/Berlin, FRG, _/ \_ Phone +49 33056 82151, Fax +49 33056 82152 (______) dave@turbocat.de (NeXTMail,MIME)
From: hugues@precipice.fdn.fr (Hugues RICHARD) Newsgroups: comp.sys.next.programmer Subject: Re: Ghostscript 4.01 or 4.03 anyone? Date: 5 Oct 1996 11:18:37 GMT Organization: Individual - France Message-ID: <535g6d$3up@precipice.fdn.fr> References: <52rsvh$ftd@chinx10.thoughtport.net> bbum@friday.com (Bill Bumgarner) wrote: >Does anyone have Ghostscript 4.03 or 4.01 available for NEXTSTEP? >Without X11 dependencies? > >I have found that Ghostscript under Linux makes an excellent driver >for the HP deskjet series of printers-- I'm wondering if the same >holds true for Ghostscript under Linux... I'm not so much interested >in hooking up the printer to the parallel port (I have a 486 DX/2 >running Linux that makes an excellent lpd server ;-), but am very >much interested in using Ghostscript as the rasterization engine... What such a paradox ! NS has a true PS Level 2 interpreter ready to rasterize for you and people use Ghostscript. What a ressource waste... And why ? Because of undocumentation... ? Don't tell me you care about Adobe DPS rasterizing limitation licence :-) Hugues. -------------------------------------------------------------------- hugues@precipice.fdn.fr - French, English, Italian and a few JP ->OK ------------ NS3.2 ------------ NS3.0J ------------ :-) ------------
From: michal@gortel.phys.ualberta.ca (Michal Jaegermann) Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: Re: Printf %g format still broken in OPENSTEP 4.0 Followup-To: comp.sys.next.programmer,comp.sys.next.bugs Date: 6 Oct 1996 04:17:39 GMT Organization: Disorganized Bits Message-ID: <537bt3$uug@pulp.ucs.ualberta.ca> References: <52p5vp$1sq@tribune.usask.ca> <32546EB0.3CD6@Eng.Sun.COM> Mark Anenberg (marka@Eng.Sun.COM) wrote: : : You could always buy a real Operating system like Solaris and use : Solaris Openstep.... (sorry I couldn't resist the jab). : : Here is the output of your program on Solaris 2.5. : : /tmp:503> cc x.c : /tmp:504> a.out : Expect 0.00123: 0.00123 ... If this is what is needed to become "a real Operating system", then my old Atari had one which was very real. Or cc output has to be called a.out? Sorry, this was not satisfied. --mj
From: cnyap@next (Chih Nam Yap) Newsgroups: comp.sys.next.programmer Subject: Drag a subview out from its superview Date: 6 Oct 1996 15:21:39 GMT Organization: University of Sheffield, UK Message-ID: <538iq3$mrv@bignews.shef.ac.uk> Hi there, Supposing I have a superview and a set of subviews. When the user hold the left mouse button down on one of the subviews, the user can move it around within the superview. However, if the mouse has moved out the superview, the user can move this subview to a bin to dispose it. What is the best way to drag a subview out from its superview ? I understand that subviews can never draw outside their superview, but not quite sure wherether I can treat the superview as the source of a drag-&-drop operation and allow users to drag subviews out from their superview ? Do I need to convert a subview into a .tiff image before I can drag it out ? Thanks.
From: bbum@friday.com (Bill Bumgarner) Newsgroups: comp.sys.next.programmer Subject: Re: Ghostscript 4.01 or 4.03 anyone? Date: 6 Oct 1996 18:59:03 GMT Organization: Friday Software & Consulting Message-ID: <538vhn$8qv@chinx10.thoughtport.net> References: <52rsvh$ftd@chinx10.thoughtport.net> <535g6d$3up@precipice.fdn.fr> Cc: hugues@precipice.fdn.fr While Adobe's DPS 2.0 under NEXTSTEP is a wonderful rasterization engine, it does not contain drivers for spewing forth images (tiff, gif, png, jpeg, whatever) nor does it contain drivers to drive devices like the deskjet. Nor is it portable. Now-- yes-- I could pay more than twice as much as I paid for my printer and purchase a license to Dots. But, considering that the printer is a toy in the first place, I cannot justify that kind of expense. OR I could take the time to take GHOSTSCRIPT's deskjet 550c driver and rip it apart enough to figure out how it converts a bitmap into printable output THEN I could figure out how to make postscript rasterize into an arbitrary image without using any windowserver resources... But the reality is that there are a million projects that I could throw code at-- a throwaway printer driver is not high enough on the list to warrant my time. OR I could take an off-the-shelf freeware solution [ghostscript] and use it 'out of the box' so to speak (it turns out that a pre-compiled copy of Ghostscript already exists). Seeing that ghostscript also supports all X fonts AND all of the standard PostScript fonts, it is a superior solution anyway. As well, this solution is 100% portable-- something that could never be achieved with a DPS/PSII under NEXTSTEP solution. As well, since I have a NeXT Cube (with NeXTDimension, no less) with 40mb of memory and a 2gb system disk (it had been up 42 days the other day when I noticed it had a 1.5 GB swap file. Oops. And that was before I started using it as a raster engine) that is already nothing but a fileserver, NeXT printer rasterization engine, and fax modem driver, making it into a GHOSTSCRIPT rasterization engine seems like a pretty reasonable use of resources to me. b.bum
From: royalta@aol.com (RoyalTA) Newsgroups: comp.sys.next.programmer Subject: Re: Project Builder Crashes System Over and Over Date: 6 Oct 1996 18:50:23 -0400 Organization: America Online, Inc. (1-800-827-6364) Sender: root@newsbf02.news.aol.com Message-ID: <539d3f$a2p@newsbf02.news.aol.com> References: <3254BDBE.61FD@maroon.tc.umn.edu> I find that turning the indexing off in preferences eliminates most crashing. Tony Royal Jenike & Johanson, Inc.
From: Robert La Ferla <Robert_La_Ferla@hot.com> Newsgroups: comp.sys.next.sysadmin,comp.sys.next.programmer Subject: sendmail 8.8 Date: 7 Oct 1996 02:23:31 GMT Organization: Posted via CAIS Internet <info@cais.com> Message-ID: <539pj3$nlo@news.cais.com> I am currently preparing a sendmail 8.8 distribution for NEXTSTEP. I would like to include mail.local as part of this distribution. If anyone has PORTED this to NEXTSTEP, please e-mail me and I will include it (and credit you...) Robert La Ferla Registered OPENSTEP Consultant -- Robert La Ferla Registered NEXTSTEP/OPENSTEP Consultant HTI Boston, MA + 1 (617) 252-0088
From: Constantin Szallies <szallies@energotec.de> Newsgroups: comp.sys.next.programmer Subject: Re: DO problem under NEXTSTEP 3.3 "[NXConnection run] - tossing received reply msg" Date: 7 Oct 1996 09:00:36 GMT Organization: Tech Net GmbH Message-ID: <53agrk$jc8@ddfservb.technet.net> References: <52ot2m$gb1@ddfservb.technet.net> <532qjm$ctv@baldwin.rat.se> Joakim Johansson <jocke@rat.se> wrote: >Constantin Szallies <szallies@energotec.de> writes >> What happens if you send a message and the sender wants a proxy >> and the server returns a copy (or the other way around)? > >Could you please rephrase the question? A and B live in different processes. A calls B someObject=[B getFoo] and B returns a proxy (a copy of Foo) but A expects a copy of Foo (a proxy) >> I guess one should view >> >(that is: "@interface X:Object // <protocolXDO> oops!") >> >> as a bug in the application program. > >I dont think so, I haven't seen any references in the docs >(those that exist... ;-) that the receiving proxy actually >has to explicitly conform to the protocol. Client and server have a different message signature and it's not the fault of the PDO system. I would consider this a bug of the application :-) Well let's say it's a bug in the PDO documentation... >Anyway, it's not a serious problem - it was just a suggestion >on a reason why <dennis.glatting@plaintalk.bellevue.wa.us> might >get that warning message. (which can be quite annoying until >you find out what causes it...) -- Constantin Szallies, Energotec GmbH szallies@energotec.de 49211-9144018
From: dietzsch@rmhs2.urz.tu-dresden.de (Andreas Dietzsch) Newsgroups: comp.sys.next.programmer Subject: Messages to a Main-Bundle Date: 7 Oct 1996 09:40:49 GMT Organization: TU Dresden (URZ) Message-ID: <53aj71$sdq@rks1.urz.tu-dresden.de> How can I send a message from a bundle to its main-bundle? Thanks Andreas
From: neuss@isa.informatik.th-darmstadt.de.NOSPAM (Christian Neuss) Newsgroups: comp.sys.next.software,comp.sys.next.programmer,comp.sys.next.misc Subject: Re: Does Enterprise Objects Framework 2.0 support Java? Followup-To: comp.sys.next.software,comp.sys.next.programmer,comp.sys.next.misc Date: 7 Oct 1996 13:55:40 GMT Organization: Fachbereich Informatik, TH Darmstadt, Deutschland Message-ID: <53b24s$uku@rs18.hrz.th-darmstadt.de> References: <32532460.1699515@news.io.org> Andrew Tang (atang@io.org) wrote: > In WebObjects Enterprise 3.0/Enterprise Objects Framework 2.0, can I > write ALL (100%) of my business objects in Java? Basically, we'd like > to do all of our implementation in Java; we do not want to write even > one line of Objective-C code. Is this possible with WebObjects > Enterprise 3.0? Hm, is WO 3.0 released yet? If so, does it contain a Java compiler? :-) If the latter is the case, the answer probably is yes, otherwise it's doubtful. Now if you were talking about the client side of a WebObjects based Weblication, that'd be a completely different story. Let's wait until friday. Yours truely, Chris -- // Christian Neuss "I ride tandem with a random.." // http://www.informatik.th-darmstadt.de/~neuss/ // fax: (+49) 6151 16 5472
From: arunc@cmp-rs.music.uiuc.edu (Arun Chandra) Newsgroups: comp.sys.next.programmer,comp.sys.next.sysadmin Subject: Compiling tcl/tk for NS3.1 and NS3.2 Date: 7 Oct 1996 15:11:40 GMT Organization: University of Illinois at Urbana Message-ID: <53b6jc$gmm@vixen.cso.uiuc.edu> We're trying to compile Tcl7.5 and Tk4.1 for black NeXTs running NS3.1 and NS3.2. Both libraries (libtk4.1.a and libtcl7.5.a) compile correctly, as does the "tclsh" program. We cannot get the "wish" program to complete its compile. It breaks down during the linking stage, complaining of an undefined symbol "panic". This symbol IS defined and is present in the library libtcl7.5.a, which is listed in the linking process. Why can't the linker find this symbol? (We're using gcc 2.7.0, (on NS3.1) and gcc 2.7.2 (on NS3.2) as the compiler, and gnu make 3.75 on both systems.) gcc tkAppInit.o -L/disk2/XApps/tkl/tk4.1/unix -ltk4.1 \ -L/disk2/XApps/tkl/tcl7.5/unix -ltcl7.5 -L/usr/lib -lX11 -o wish ld: Undefined symbols: _panic make: *** [wish] Error 1 Thanks in advance for any suggestions. -- Arun Chandra cmp-rs.music.uiuc.edu
From: Denis Mulder <denis@fygir.nl> Newsgroups: comp.sys.next.programmer Subject: Nextstep dev. wanted, Manufacturing systems, The Hague, The Netherlands Date: 7 Oct 1996 15:26:41 GMT Organization: Fygir inc. Message-ID: <53b7fh$sij@news.xs4all.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Fygir is a rapidly growing international company that focuses on the development, sale and implementation of state-of-the-art planning and scheduling software for manufacturers. From offices in Boston (USA), The Hague (The Netherlands) and London (UK), Fygir works for many multinational companies. Fygir has distinguished itself in this market by applying new logistic principles and modern information technology within their innovative software. Our developers use the modern OO NEXTSTEP / Openstep environment. Due to the development of a new standard software package, we have an immediate opportunity in The Hague for SOFTWARE ENGINEERS who will independently design and implement object-oriented software. In particular you should possess the following qualities: object oriented design and programming experience; knowledge of NEXTSTEP/Openstep; excellent communicative capabilities, both verbally and written (English) In return, we offer challenging work in a stimulating environment, an excellent renumeration package and very good possibilities for personal development. An application letter with resumé should be addressed for the attention of : Denis Mulder Fygir logistic information systems Polakweg 23 2288 GG Rijswijk The Netherlands Telephone: 31 70 336 8330, email: denis@fygir.nl
Message-ID: <32592C7D.272E@dsm.fordham.edu> Date: Mon, 07 Oct 1996 12:14:53 -0400 From: Ellen <hliu@dsm.fordham.edu> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: GCC 2.7.2.1 on NextStep Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I need either next step 3.3 gcc binaries or libgcc.a so that I could compile a gcc compiler for my next step computer. I would appreciate any help anyone can give me. Please post information on sites that have the binaries or libgcc.a. Thank You.
Message-ID: <32592E9E.5BB3@dsm.fordham.edu> Date: Mon, 07 Oct 1996 12:23:58 -0400 From: Ellen <hliu@dsm.fordham.edu> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Re: C++ rot 4.0 OpenSTEP References: <32542705.7BCF@samuel.cpsc.sru.edu> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sam R. Thangiah wrote: > > I am trying to get the C++ compiler running on the Openstep (4.0) but it > does not seem to like either the c++ command or the file with .c > extension. > > Could somebody give me some pointers on how to get C++ running for the > 4.0 > > Thanks, > > Sam > > sam.thangiah@sru.edu or > thangiah@samuel.cpsc.sru.edu c++ file must have a .cpp or .cc extension
Message-ID: <32592E54.4785@dsm.fordham.edu> Date: Mon, 07 Oct 1996 12:22:44 -0400 From: Ellen <hliu@dsm.fordham.edu> MIME-Version: 1.0 Newsgroups: comp.sys.next.programmer Subject: Re: Which GCC version under 4.0 References: <52h84k$72o@news2.snfc21.pacbell.net> <533cjf$1fa@news.next.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Samuel Figueroa wrote: > > Mike Paquette writes > > edx@cc.usu.edu wrote: > > >The Release notes for 4.0 say that the compiler is based > > >on gcc version 2.5.8 > > >I guess NeXT isn't in the compiler business any more either. > > Just as a point of information, if one were going to redo the code > > generator of a compiler, to , oh say, generate highly optimized > > position independent code for several architectures, and one knew that > > several million lines of code would have to be compiled correctly by > > that compiler, then one might want to stop adding random changes from > > third parties some time before the compiler was needed for production > > work, and concentrate on the tasks at hand. > > Another point: Just because the gcc version says 2.5.8 doesn't mean > > that later gcc bug fixes wouldn't be incorporated. Many fixes are > > incorporated where they make sense. > > I'm not an official spokesperson for NeXT, but I agree with Mike. NeXT > has shipped compilers based on gcc 2.7.2 for some of our platforms, but > these are buggier than we would like (though still perfectly usable). On > Mach, we can't afford this level of instability, so we will not ship a gcc > 2.7.2-based compiler for Mach until it is ready. > > - Sam Figueroa (figueroa@next.com) I am looking to compile gcc for my next step 3.3 computer. Unfortionately this requires either a compiler on my next step, which I don't have, or libgcc.a + the include files from a next step computer, which again I don't have. If you could post either of these items I would be very greatfull. Thank You.
Message-ID: <32592ED6.7CD6@dsm.fordham.edu> Date: Mon, 07 Oct 1996 12:24:54 -0400 From: Ellen <hliu@dsm.fordham.edu> MIME-Version: 1.0 Newsgroups: comp.sys.next.sysadmin,comp.sys.next.programmer Subject: Re: Compile problems under 3.3 References: <52vetp$g4n@jolt.pagesat.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Paul R. Wellin wrote: > > I am having some troubles compiling certain apps and the problems seem to > point to the same source. On my black turbo with 3.3, during a compile > (of both Gatekeeper and samba) I get the following error: > > django:6# make > Using CFLAGS = -O -DSMBLOGFILE="/usr/local/samba/var/log.smb" > ... stuff... > Compiling server.c > /NextDeveloper/Headers/netdb.h:66: header file 'sys/bitypes.h' not found > includes.h:591: warning: could not use precompiled header > '/NextDeveloper/Headers/bsd/libc.p', because: > includes.h:591: warning: header 'arpa/inet.h' overridden, > includes.h:591: warning: /NextDeveloper/Headers/arpa/inet.h vs. > /NextDeveloper/Headers/bsd/arpa/inet.h (within the precomp) > /NextDeveloper/Headers/arpa/inet.h:68: header file 'sys/bitypes.h' not found > *** Exit 1 > Stop. > > I can find the file bitypes.h nowhere on my system and during each of > my two recent compiles, this is where the Exit occurs. > > Am I doing something rather unintelligent here? Has anyone seen this > before or have a clue what I can do to prevent this? > > Thanks, > > Paul Wellin I am looking to compile gcc for my next step 3.3 computer. Unfortionately this requires either a compiler on my next step, which I don't have, or libgcc.a + the include files from a next step computer, which again I don't have. If you could post either of these items I would be very greatfull. Thank You.
From: "SSDS, Inc" <wej@ssds.com> Newsgroups: comp.sys.next.programmer Subject: Sr C++ OO Leader Needed Date: 7 Oct 1996 20:58:17 GMT Organization: SSDS Distribution: inet Message-ID: <01bbb492$93c66be0$580e7f86@pc_bos_nec.ssds.com> SSDS, Inc., Boston, MA office, has a need for a Senior C++ OO Specialist to lead development of several OO development projects. Projects are being developed in house and in a Microsoft centric environment using C++, VB, NT4.0, SQL Server 6.5 and Rational Rose OMT plus related OO technology and methods. SSDS Inc. is an equal opportunity employer with very competitive salary and benefits. Interested parties please check out or web site http:\\www.ssds.com and respond with email, resume or phone call to discuss opportunities further. Bill Jackson 1-800-617-7737 WEJ@SSDS.com
From: eric@skatter.USask.Ca Newsgroups: comp.sys.next.programmer Subject: setuseprecisecolor Date: 7 Oct 1996 22:00:08 GMT Organization: University of Saskatchewan Message-ID: <53buh8$e1b@tribune.usask.ca> I received no replies to my previous posting about the problems I've been having with OS/4.0 and printing. Does this mean that everyone else can print to non-NeXT printers without any problems? To summarize the problem I'm having: The `print' files from my application (produced by sending a `print:' message to the applications main window) contain the following code: %%NXBeginNamedColor: (System) (controlColor) 0.113554 0.007689 0.079487 0 0.666667 0.666667 0.666667 _NXSetCMYKOrRGB %%NXEndNamedColor 0 0 197.5 18 rectfill 0.5 setgray false setuseprecisecolor 1 1 195.5 16 rectfill . . . After poking around with Yap for a while, it seems to me that this code is putting out the background rectangle for a slider. The problem is that the command `setuseprecisecolor' doesn't seem to be defined anywhere. I tried getting around this by adding /setuseprecisecolor {pop} def near the beginning of the `.ps' file. With this hack in place I was able to get the file to print on a NEXT color printer. The Apple LaserWriters still produce no output, though. So, am I the only person in the world with this problem? Does no one else see these `setuseprecisecolor' commands in their program output? Can everyone else print to Apple LaserWriters? -- Eric Norum eric@skatter.usask.ca Saskatchewan Accelerator Laboratory Phone: (306) 966-6308 University of Saskatchewan FAX: (306) 966-6058 Saskatoon, Canada. NeXTMail accepted.
From: eric@skatter.USask.Ca Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: Openstep 4.0 printing problems diagnosed Date: 7 Oct 1996 22:56:00 GMT Organization: University of Saskatchewan Message-ID: <53c1q0$e1b@tribune.usask.ca> I figured out why the LaserWriters here were unable to print the PostScript produced by my application. Here's the offending code: %%NXBeginNamedColor: (System) (controlColor) 0.113554 0.007689 0.079487 0 0.666667 0.666667 0.666667 _NXSetCMYKOrRGB %%NXEndNamedColor 2 2 249.5 213 rectfill 0 setgray 2 195 248.5 1 rectfill grestore gsave 0 0 sethalftonephase 2 2 249 193 rectclip [1 0 0 -1 -90 272] concat gsave sethalftonephase is a Display PostScript operator! No wonder the printers choke on it! Like my previous problem (setuseprecisecolors) this one seems to be part of a scrollbar-drawing routine (i.e. the offending commands are from NeXT code). For now, I am working around both problems by adding: /setuseprecisecolor {pop} bind def /sethalftonephase {pop pop} bind def near the beginning of my files before sending them to the printer. This is hardly an adequate solution, though. Anyone out there got a *real* fix for this? Oh yeah, one other thing. My printer log file shows: GothicBBB-Medium.Hankaku not found, using Courier. HeiseiMin-W3.Hankaku not found, using Courier. each time my application prints. What the heck are these and what is requesting them? Certainly nothing in my code. -- Eric Norum eric@skatter.usask.ca Saskatchewan Accelerator Laboratory Phone: (306) 966-6308 University of Saskatchewan FAX: (306) 966-6058 Saskatoon, Canada. NeXTMail accepted.
From: anch@logiball.de (Andreas Christiani) Newsgroups: comp.sys.next.programmer Subject: BLOBs and OpenBase Date: 8 Oct 1996 08:06:26 GMT Organization: Customer of EUnet Germany; Info: info@Germany.EU.net Message-ID: <53d222$cbh@news.Dortmund.Germany.EU.net> Hi all, I need some advice on using BLOBSs with OpenBase and EOF. Is there anyone who has got experience with this stuff ? Thank you for any help, A.C. --- ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ LogiBall gGmbH * Innovationszentrum Herne * Westring 303 * 44629 Herne Andreas Christiani * christiani@logiball.de * http://www.logiball.de Tel.: 02323 / 925 577 * Fax : 02323 / 925 551 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
From: jdubois@mathcs.carleton.edu (Josh DuBois) Newsgroups: comp.sys.next.programmer Subject: NSArray deep copy? Date: 8 Oct 1996 03:49:59 GMT Organization: Carleton College, Northfield, MN, USA Message-ID: <53cj17$gc5@zorak.acns.carleton.edu> Howdy, all- NSArray seems to do a deep copy. Why? Seems like this is prone to lots of errors, as you don't know until run time whether or not items in the array will be able to copy themselves. Any ideas as to why this was done? Just curious, Josh.
From: bmw@visgen.com (Bruce Walker) Newsgroups: comp.sys.next.programmer Subject: Re: Ghostscript 4.01 or 4.03 anyone? Date: 7 Oct 1996 20:11:37 -0400 Organization: Visible Genetics Inc. Message-ID: <53c67p$b7u@ampere.visgen.com> References: <52rsvh$ftd@chinx10.thoughtport.net> <535g6d$3up@precipice.fdn.fr> In article <535g6d$3up@precipice.fdn.fr>, >bbum@friday.com (Bill Bumgarner) 1st wrote: >Does anyone have Ghostscript 4.03 or 4.01 available for NEXTSTEP? >Without X11 dependencies? It compiled cleanly for me. I just commented-out the X-lines, used the unix-gcc.mak Makefile and tweaked a couple of things per make.txt (search for "NeXT" and use the 3.2 notes). I compiled it for both m68k and i386. I using it with mgetty+sendfax-0.99. Hugues RICHARD <hugues@precipice.fdn.fr> then wrote: >What such a paradox ! >NS has a true PS Level 2 interpreter ready to rasterize for you and >people use Ghostscript. What a ressource waste... >And why ? Because of undocumentation... ? For me, it's because there's no way to use the DPS calls unless you are logged-in. That spoils the NeXT DSP interpreter as far as being used in a "server" application goes. >Don't tell me you care about Adobe DPS rasterizing limitation >licence :-) And why so flip about a license restriction? In a public forum? With lawyers present? ;-) -- Bruce M. Walker | Visible Genetics Inc. | bmw@visgen.com | Suite 1000, Box 333 | Phone: 416-813-3282 | 700 Bay Street | FAX: 416-813-3250 | Toronto M5G 1Z6 |
From: David Hinz <David.Hinz@mci.com> Newsgroups: comp.sys.next.programmer Subject: Q: Java compiler for NS 3.3 Date: Tue, 08 Oct 1996 05:19:44 -0600 Organization: MCI Message-ID: <325A38D0.4B7D@mci.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Is there a Java compiler available for NextStep 3.3? Specifically one that is compatible with the JDK 1.02 specs. I've seen the guavac compiler on some of the ftp sites but it seems to be based on JDK 1.0 and isn't complete. Is there anything else out there? Thanks for any info. dave. David.Hinz@mci.com
From: ici@ijexa.osk.threewebnet.or.jp (Toshinao Ishii) Newsgroups: comp.sys.next.programmer Subject: Re: Openstep 4.1? (was: Which GCC version under 4.0) Date: 06 Oct 1996 15:53:59 GMT Organization: 3WEB corp. Message-ID: <ICI.96Oct7005359@ijexa.osk.threewebnet.or.jp> References: <52bq99$ir9@Holly.aa.net> <1996Sep26.101005.85623@cc.usu.edu> <52ff6o$srt@sjx-ixn2.ix.netcom.com> <01bbacc0$eefa9b00$3e031281@bananajr> <52uavj$fvi@news.doit.wisc.edu> In-reply-to: jessica@whitewater.chem.wisc.edu's message of 2 Oct 1996 18:06:43 GMT Hi. In article <52uavj$fvi@news.doit.wisc.edu> jessica@whitewater.chem.wisc.edu (Jessica Severin) writes: > Is Openstep 4.1 already in Beta?!?! > Is there any projected release date for it? > 6months? > 1year? NEXTSOFTWARE Japan is announcing that they will release OPENSTEP for MACH 4.1J in late November. -=-=-=-=-= Toshinao Ishii email: ici@osk.threewebnet.or.jp
From: zander@conextions.com (Aleksey Sudakov) Newsgroups: comp.sys.next.programmer Subject: Re: Q: Java compiler for NS 3.3 Date: 8 Oct 1996 13:01:26 GMT Organization: The Internet Access Company, Inc. Message-ID: <53djb6$6qg@news-central.tiac.net> References: <325A38D0.4B7D@mci.com> David Hinz <David.Hinz@mci.com> wrote: >Is there a Java compiler available for NextStep 3.3? Specifically one >that is compatible with the JDK 1.02 specs. > >I've seen the guavac compiler on some of the ftp sites but it seems to >be based on JDK 1.0 and isn't complete. Is there anything else out >there? How about kaffe - JVM and sun native javac?
From: Matthew Hagerty <matthew@internet1.net> Newsgroups: gnu.gcc.help,comp.sys.next.sysadmin,comp.sys.next.programmer Subject: GCC on a NeXT Date: Mon, 07 Oct 1996 11:46:30 -0700 Organization: internet1 Message-ID: <32595006.6EBA@internet1.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Greetings, I am trying to compile gcc 2.7.2 on a NeXTstationTC (pizza box 68040) running NeXTStep 3.2 and was wondering what I should use for the machine type? Reading the docs indicate that NeXT, m68K, and Mach are all supported, but this is what I get when I try to run configure: eer-next1:# cd /usr/local/gcc-2.7.2.1 eer-next1:# ./configure Config.guess failed to determine the host type. You need to specify one. Usage: configure [--host=HOST] [--build=BUILD] [--prefix=DIR] [--gxx-include-dir=DIR] [--local-pref=DIR] [--exec-pref=DIR] [--with-gnu-as] [--with-gnu-ld] [--with-stabs] [--with-elf] [--nfp] TARGET Where HOST, TARGET and BUILD are three-part configuration names eer-next1:# hostinfo Mach kernel version: NeXT Mach 3.2: Mon Oct 18 21:57:41 PDT 1993; root(rcbuilder):mk-149.30.15.obj~2/RC_m68k/RELEASE_M68K Kernel configured for a single processor only. 1 processor is physically available. Processor type: MC680x0 (68040) Processor speed: 33 MHz Processor active: 0 System type: 5 Board revision: 0xf Primary memory available: 24.00 megabytes. Default processor set: 51 tasks, 80 threads, 1 processors Load average: 0.00, Mach factor: 0.99 eer-next1:# ./configure --build=m68k-next-mach3.2 Configuration m68k-next-mach3.2 not supported eer-next1:# ./configure --build=m68k-next-mach Configuration m68k-next-mach not supported Any suggestions, ideas, or input would be greatly appreciated. Thank you Matthew Hagerty
From: Charles William Swiger <cs4w+@andrew.cmu.edu> Newsgroups: comp.sys.next.sysadmin,comp.sys.next.programmer,gnu.gcc.help Subject: Re: GCC on a NeXT Date: Tue, 8 Oct 1996 18:06:00 -0400 Organization: Carnegie Mellon, Pittsburgh, PA Message-ID: <8mKh18O00Uzx02mnoR@andrew.cmu.edu> References: <32595006.6EBA@internet1.net> In-Reply-To: <32595006.6EBA@internet1.net> Excerpts from netnews.comp.sys.next.programmer: 7-Oct-96 GCC on a NeXT by Matthew Hagerty@internet > I am trying to compile gcc 2.7.2 on a NeXTstationTC (pizza box 68040) > running NeXTStep 3.2 and was wondering what I should use for the machine > type? Reading the docs indicate that NeXT, m68K, and Mach are all > supported, but this is what I get when I try to run configure: > > eer-next1:# cd /usr/local/gcc-2.7.2.1 > eer-next1:# ./configure > Config.guess failed to determine the host type. You need to specify > one. That's surprising-- configure has always managed to ID a NEXTSTEP system as far as I can remember. Anyway, the complete target name that you are looking for is 'm68k-next-nextstep3', although just 'next' should work too. -Chuck Charles Swiger | cs4w@andrew.cmu.edu | standard disclaimer ----------------+---------------------+--------------------- I know you're an optimist if you think I'm a pessimist.
From: gimic@best.com (Windows Minion) Newsgroups: comp.sys.next.programmer Subject: Resources on the Web? Date: 9 Oct 1996 01:38:42 GMT Organization: Apple Computer, Inc. Message-ID: <gimic-0809961840470001@17.211.20.52> What are good sites for OpenStep developement?
From: gimic@best.com (Windows Minion) Newsgroups: comp.sys.next.programmer Subject: Recommened Books Date: 9 Oct 1996 01:39:12 GMT Organization: Apple Computer, Inc. Message-ID: <gimic-0809961841160001@17.211.20.52> Also, are there any good books out there for OpenStep developement?
Newsgroups: comp.sys.next.programmer Subject: JAVA sAWT : anybody working on it? From: apl@kcbbs.gen.nz (Andrew Lindesay) Date: 9 Oct 96 10:57:53 GMT Message-ID: <17496282.39473.24773@kcbbs.gen.nz> Organization: Kappa Crucis Unix BBS, Auckland, NZ. Summary: Keywords:
From: "Georg Tuparev" <gtupar@ctp.com> Newsgroups: comp.sys.next.programmer Subject: Re: Resources on the Web? Date: 9 Oct 1996 17:11:00 GMT Organization: Cambridge Technology Partners, Inc. Message-ID: <53gmb4$t25@concorde.ctp.com> References: <gimic-0809961840470001@17.211.20.52> In article <gimic-0809961840470001@17.211.20.52> gimic@best.com (Windows Minion) writes: > What are good sites for OpenStep developement? http://www.nmr.embl-heidelberg.de/GNUstep/ .. and about books: there are still no published books, but some (well at least one) is in preparation ;-) Have fun -- georg --
From: mow@navigator.de (Markus Wenzel) Newsgroups: comp.sys.next.programmer Subject: suidperl with Nextstep Date: 8 Oct 1996 06:48:59 GMT Organization: Navigator Message-ID: <53ctgr$69e@marsu.navigator.de> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Hi, has anybody used suidperl with Nextstep 3.3 yet? If so, how did you do it? Thanks for any answers in advance. -- Navigator Markus Wenzel info@navigator.de IT Consulting & System Administration http://www.navigator.de/
From: mpaque@next.com (Mike Paquette) Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: Re: Openstep 4.0 printing problems diagnosed Date: Wed, 09 Oct 1996 22:06:34 GMT Organization: Electronics Service, Unit No. 16 Message-ID: <53h7tn$cvj@news2.snfc21.pacbell.net> References: <53c1q0$e1b@tribune.usask.ca> eric@skatter.USask.Ca wrote: >Oh yeah, one other thing. My printer log file shows: >GothicBBB-Medium.Hankaku not found, using Courier. >HeiseiMin-W3.Hankaku not found, using Courier. >each time my application prints. What the heck are these and what is >requesting them? Certainly nothing in my code. That's the font substitution gismology in the PostScript packages. It's sniffing about to see if it should apply a Kanji or European style (Courier...) font substitution for unknown fonts. The log statements are harmless but slightly annoying. Mike Paquette I don't speak for my employer, and they don't speak for me. mpaque@next.com NeXT business mail only, please
From: gclem@dannug.dk Newsgroups: comp.sys.next.programmer Subject: Re: BLOBs and OpenBase Date: 9 Oct 1996 06:19:39 GMT Organization: Danish NeXT User Group Distribution: world Message-ID: <53fg5r$bs@snaps.dannug.dk> References: <53d222$cbh@news.Dortmund.Germany.EU.net> In just 12 days the Copenhagen Police will enter into production mode with a new command-and-control center handling all 112/911 calls for Greater Copenhagen. We have developed all the software (NS 3.3 + OpenBase) and use BLOBs extensively, so I guess you could say that we have some experience. Anything in particular you or any other guy out there is interested in? Geert Andreas Christiani writes > Hi all, > > I need some advice on using BLOBSs with OpenBase and EOF. Is there anyone who > has got experience with this stuff ? > > Thank you for any help, > > A.C. > > --- > ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > LogiBall gGmbH * Innovationszentrum Herne * Westring 303 * 44629 Herne > Andreas Christiani * christiani@logiball.de * http://www.logiball.de > Tel.: 02323 / 925 577 * Fax : 02323 / 925 551 > ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
From: kinau@shadow.csufresno.edu (Kin Hung Au) Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin Subject: Can't compile TCP wrapper Date: 7 Oct 1996 22:32:40 GMT Organization: California State University, Fresno Message-ID: <53c0e8$4ui@zimmer.CSUFresno.EDU> Hi Folks, I can't compile TCP wrapper for NeXT black hardware. I get following message: make[1]: Entering directory `/private/Net/math/Users/kinau/tcp_wrappers_7.4' cc -O -DFACILITY=LOG_MAIL -DHOSTS_ACCESS -DPARANOID -DNETGROUP -DGETPEERN AME_BUG -DBROKEN_FGETS -DLIBC_CALLS_STRTOK -DDAEMON_UMASK=022 -DREAL_DAEMON_DI R=\"/usr/local/adm\" -DSEVERITY=LOG_INFO -DRFC931_TIMEOUT=10 -DHOSTS_DE NY=\"/etc/hosts.deny\" -DHOSTS_ALLOW=\"/etc/hosts.allow\" -Dvsyslog=myvsyslog -DALWAYS_HOSTNAME -o tcpdmatch tcpdmatch.o fakelog.o inetcf.o scaffold.o libwrap.a -m68040 ld: multiple definitions of symbol _closelog fakelog.o definition of _closelog in section (__TEXT,__text) /lib/libsys_s.a(syslog.o) definition of absolute _closelog (value 0x50022a6) ld: multiple definitions of symbol _openlog fakelog.o definition of _openlog in section (__TEXT,__text) /lib/libsys_s.a(syslog.o) definition of absolute _openlog (value 0x5002bd0) ld: multiple definitions of symbol _syslog fakelog.o definition of _syslog in section (__TEXT,__text) /lib/libsys_s.a(syslog.o) definition of absolute _syslog (value 0x5003122) ld: multiple definitions of symbol _getenv libwrap.a(environ.o) definition of _getenv in section (__TEXT,__text) /lib/libsys_s.a(getenv.o) definition of absolute _getenv (value 0x50025d0) make[1]: *** [tcpdmatch] Error 1 make[1]: Leaving directory `/private/Net/math/Users/kinau/tcp_wrappers_7.4' make: *** [next] Error 2 Could anyone give some tips? Or, is there any ftp site I can download binary version TCP wrapper for NeXT? Thanks. -Kin ****************************************************************************** Kin Hung Au Internet Address: kinau@csufresno.edu Instructional Computing Consultant California State University, Fresno Tel# 209-278-3915 School of Natural Science FAX# 209-278-7139 Office of Dean, MS #90 Fresno CA 93740 http://maxwell.phys.csufresno.edu:8001/~kinau/ -----BEGIN PGP PUBLIC KEY BLOCK----- Version: 2.6.2 mQBtAzJJkwEAAAEDAMHh5WMewwIXy3D7K46JeFkiFARiSpYu0osPy0gM7dkQqW28 ym7gMft0gAHC9fpvIQQ93BESamn6MuNt+f9yxhMSZX/UGX6LqG7jyEmTgWzyDLkw VxNkugSY/uDi7ZasOQAFEbQfS2luIEguIEF1IDxraW5hdUBjc3VmcmVzbm8uZWR1 Pg== =2ks8 -----END PGP PUBLIC KEY BLOCK----- ******************************************************************************
From: dietzsch@rmhs2.urz.tu-dresden.de (Andreas Dietzsch) Newsgroups: comp.sys.next.programmer Subject: a Filelist for an directory Date: 10 Oct 1996 12:46:16 GMT Organization: TU Dresden (URZ) Message-ID: <53ir6o$1e3@rks1.urz.tu-dresden.de> How can I get a list of all files of a directory? thank You Andreas
From: jlincoln@foresight-sys.com Newsgroups: comp.sys.next.programmer Subject: How to Browse ObjectiveC Classes? Date: 10 Oct 1996 01:58:33 GMT Organization: OnRamp Technologies; ISP; Dallas/Ft Worth/Houston, TX USA Message-ID: <53hl89$1vu@news.onramp.net> I am trying to learn the NeXT Developer environment and need to know how to browse all of the ObjectiveC classes. Thanks, Jason
From: Alex Duong Nghiem <alexdn@globalobjects.com> Newsgroups: comp.object,comp.sys.next.programmer,comp.sys.announce,comp.lang.java,comp.lang.java.programmer,misc.jobs.offered,misc.jobs.entry,misc.jobs.contract,atl.jobs Subject: US - GA - OO developers needed: WebObjects, NextStep, Objective-C, EOF, Java, Web, Smalltalk, C++ Date: Wed, 09 Oct 1996 22:02:29 -0500 Organization: Global Objects Inc. Message-ID: <325C6745.3F5F@globalobjects.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit CC: jobs@globalobjects.com, peram@globalobjects.com Global Objects is a professional software service company that focuses in providing quality solutions through object technology. We specialize in object modelling, Web, Java, NextStep, Smalltalk, and C++ projects in various domains including telecommunications, petroleum, transportation and environmental services. We also provide quality courses on object technology including Java, design patterns, frameworks, object oriented analysis and design, and software development methodologies. We are experiencing rapid growth and are looking for highly candidates, senior and junior ones. Work with a team of work with a team of world-class object developers to deliver state-of-the-art solutions on the Web. Get hands-on-training on exciting new technologies such as WebObjects, Java and electronic commerce. Requirements are as follows: Experience in at least one of the following area (prefer industry experience but will consider strong academic experience): * NextStep / WebObjects / EOF / Objective-C * Java * Any other Web development environment (Cold Fusion, Sapphire, etc.) * Smalltalk * C++ or any other OO language Unix experience helpful. For immediate consideration, please e-mail your resume (preferrably in Word format or in ASCII) to jobs@globalobjects.com or fax it to 770.457.7333. Please include the following information: * availability * for full-time employment, pls specify salary history * for contract employment, pls specify rate
From: jobs@globalobjects.com Newsgroups: comp.sys.next.programmer Subject: US - GA - OO developers needed: WebObjects, NextStep, Objective-C, EOF, Java, Web, Smalltalk, C++ Date: Thu, 10 Oct 1996 04:46:48 GMT Organization: Global Objects Inc. Message-ID: <325c7e45.6685921@news-s01.ny.us.ibm.net> Global Objects is a professional software service company that focuses in providing quality solutions through object technology. We specialize in object modelling, Web, Java, NextStep, Smalltalk, and C++ projects in various domains including telecommunications, petroleum, transportation and environmental services. We also provide quality courses on object technology including Java, design patterns, frameworks, object oriented analysis and design, and software development methodologies. We are experiencing rapid growth and are looking for highly candidates, senior and junior ones. Work with a team of work with a team of world-class object developers to deliver state-of-the-art solutions on the Web. Get hands-on-training on exciting new technologies such as WebObjects, Java and electronic commerce. Requirements are as follows: Experience in at least one of the following area (prefer industry experience but will consider strong academic experience): * NextStep / WebObjects / EOF / Objective-C * Java * Any other Web development environment (Cold Fusion, Sapphire, etc.) * Smalltalk * C++ or any other OO language Unix experience helpful. For immediate consideration, please e-mail your resume (preferrably in Word format or in ASCII) to jobs@globalobjects.com or fax it to 770.457.4144. Please include the following information: * availability * for full-time employment, pls specify salary history * for contract employment, pls specify rate
From: jobs@globalobjects.com Newsgroups: comp.sys.next.programmer Subject: US - GA - OO developers needed: WebObjects, NextStep, Objective-C, EOF, Java, Web, Smalltalk, C++ Date: Thu, 10 Oct 1996 04:55:50 GMT Organization: Global Objects Inc. Message-ID: <325c7f1d.6901894@news-s01.ny.us.ibm.net> Global Objects is a professional software service company that focuses in providing quality solutions through object technology. We specialize in object modelling, Web, Java, NextStep, Smalltalk, and C++ projects in various domains including telecommunications, petroleum, transportation and environmental services. We also provide quality courses on object technology including Java, design patterns, frameworks, object oriented analysis and design, and software development methodologies. We are experiencing rapid growth and are looking for highly candidates, senior and junior ones. Work with a team of work with a team of world-class object developers to deliver state-of-the-art solutions on the Web. Get hands-on-training on exciting new technologies such as WebObjects, Java and electronic commerce. Requirements are as follows: Experience in at least one of the following area (prefer industry experience but will consider strong academic experience): * NextStep / WebObjects / EOF / Objective-C * Java * Any other Web development environment (Cold Fusion, Sapphire, etc.) * Smalltalk * C++ or any other OO language Unix experience helpful. For immediate consideration, please e-mail your resume (preferrably in Word format or in ASCII) to jobs@globalobjects.com or fax it to 770.457.7333. Please include the following information: * availability * for full-time employment, pls specify salary history * for contract employment, pls specify rate
From: dwy@visi.com (David Young) Newsgroups: gnu.gcc.help,comp.sys.next.sysadmin,comp.sys.next.programmer Subject: Re: GCC on a NeXT Followup-To: gnu.gcc.help,comp.sys.next.sysadmin,comp.sys.next.programmer Date: 10 Oct 1996 14:28:15 GMT Organization: Intellitech Corporation Message-ID: <53j15v$eke@alice.walrus.com> References: <32595006.6EBA@internet1.net> Matthew Hagerty (matthew@internet1.net) wrote: : eer-next1:# ./configure --build=m68k-next-mach3.2 : Configuration m68k-next-mach3.2 not supported : : eer-next1:# ./configure --build=m68k-next-mach : Configuration m68k-next-mach not supported You can probably compile uname from the GNU (bin?)utils. Don't know if this will make gcc get the hosttype. I vaguely remember something like m68k-next-nextstep.
From: khlee@snow.snu.ac.kr (Kyung-Hoon Lee) Newsgroups: comp.sys.next.programmer Subject: binutils for NeXT Date: 10 Oct 1996 05:28:10 GMT Organization: Seoul National University, Korea Message-ID: <53i1ha$e6m@snunews.snu.ac.kr> Summary: objdump in binutils for NeXT wanted. Keywords: binutils objdump NeXT next I'm looking for binutils for NEXTSTEP 3.3 for Motorola (the black slab). Especially, objdump. Source is preferred, but binary would be fine. If anyone successed in building objdump on next, drop me a line. Thanks in advance, Lee -- ____ ____ /| Lee, Kyung-Hoon /\ \/\ \ \| / \___\ \___\ /| Arichi Lab \ / / / / \| Department of Computer Engneering \/___/\/___/ /| Seoul National University \| khlee@lab.snu.ac.kr +82-2-880-7296
From: bbum@friday.com (Bill Bumgarner) Newsgroups: comp.sys.next.programmer Subject: Python w/NEXTSTEP news Date: 10 Oct 1996 18:08:42 GMT Organization: Friday Software & Consulting Message-ID: <53je3a$igq@chinx10.thoughtport.net> The Objective-C module for Python has been coming along nicely. It is now mature enough to build full-blown NEXTSTEP applications and it is compatible with the GNU runtime as well. As well, it now includes a fully-featured implementation of NEXTSTEP services in python; allowing one to easily write new services entirely in python. Enclosed is a copy of the message I just sent to the objc-sig in regards to new binaries and source that have recently been made available. b.bum --- From: Bill Bumgarner <bbum@friday.com> Date: Thu, 10 Oct 96 13:04:59 -0500 To: objc-sig@python.org Subject: [PYTHON OBJC-SIG] New binaries/source released Reply-To: bbum@friday.com Sender: owner-objc-sig@python.org I have integrated Lele's version 0.44 release into my project tree; binaries and source are now available at: ftp://ftp.thoughtport.net/pub/next/lang/ Included in this release are: - complete and working services demo (including installation instructions) - slight restructuring; my tree completely subsumes Lele's. (this will make future integrations much easier). - datastream is gone; it is subsumed by Lele's ObjCStream module It now includes support for GNU runtime streams. Bugs: - probably lots... - readline mode does not work in the pyobjc binary! For some reason, it is not enabled.... don't know why. Map to what's on the ftp site: 58949 Oct 10 17:58 PyObjC-0.44.tar.gz Lele's original 0.44 release. 1300480 Oct 7 07:09 Python-bin.pkg.b.SHIN.tar 2467840 Oct 7 07:12 Python.pkg.b.SHIN.tar Installer packages of Python 1.4b3 w/all of Lele (and mine) patches integrated. The smaller package (-bin) contains just the python binary and python library files; none of the libraries are headers necessary to build the PyObjC module are included. The non -bin package contains everything. 331016 Oct 10 17:55 pyobjc-0_44bbum02.b.SHIN.tar.gz 77566 Oct 10 17:59 pyobjc-0_44bbum02.s.tar.gz Source code for release 0.44 of Lele's module with my changes integrated. Includes services ipmlementation. Binary package is quad-fat binary named 'pyobjc'-- it is a statically linked build of the source. 4884689 Apr 15 18:37 python-1.3.b.SHIN.tar.gz 1.3 final relesae built quad fat. 34404 Oct 7 08:01 python-1_4b3-NeXT.patch.gz All of the patches necessary to build the above packages or the following binary archive 2092 Oct 7 03:05 python-1_4b3-README A [likely out of date] README. 3350115 Oct 7 03:09 python-1_4b3.b.SHIN.tar.gz Same as the the full package; contains everything, including libraries, executable, and all headers. enjoy. b.bum ================= OBJC-SIG - SIG on Python bindings for Objective-C/NeXTSTEP/OpenStep send messages to: objc-sig@python.org administrivia to: objc-sig-request@python.org =================
From: far@ix.netcom.com(Felipe A. Rodriguez) Newsgroups: comp.sys.next.programmer Subject: Re: [Q] Compile problem with bind Date: 10 Oct 1996 19:25:38 GMT Organization: Netcom Message-ID: <53jiji$mm8@dfw-ixnews3.ix.netcom.com> References: <53jcu1$pn3@ns1.moran.com> In article <53jcu1$pn3@ns1.moran.com> dboyce@albert (Doug Boyce) writes: >I'm getting the following error when compiling db_dump.c in the 4.9.2 >release. If someone can figure out the problem by just eye-balling the >include file that would be great. Otherwise request more info. > snip It may not be exactly the advice you are looking for, but, I recommend that you not use 4.9.2. For starters there are various known potential security problems which are being addressed. Also, beginning with the 4.9.3 release BIND builds without too much trouble under NeXT. Get the 4.9.4 version at: ftp://ftp.vix.com/pub/bind/release/4.9.4/bind-4.9_4-P1.tar.gz -- Felipe A. Rodriguez # Francesco Sforza became Duke of Milan from Agoura Hills, CA # being a private citizen because he was # armed; his successors, since they avoided far@ix.netcom.com # the inconveniences of arms, became private (NeXTmail preferred) # citizens after having been dukes. (MIMEmail welcome) # --Nicolo Machiavelli
From: zander@conextions.com (Aleksey Sudakov) Newsgroups: comp.sys.next.programmer Subject: Re: How to Browse ObjectiveC Classes? Date: 10 Oct 1996 13:32:45 GMT Organization: The Internet Access Company, Inc. Message-ID: <53ittt$qmk@news-central.tiac.net> References: <53hl89$1vu@news.onramp.net> jlincoln@foresight-sys.com wrote: >I am trying to learn the NeXT Developer environment and need to know how to >browse all of the >ObjectiveC classes. Try to use HeaderViewer.app
From: jmichel@imtn.dsccc.com (James Michel) Newsgroups: comp.sys.next.programmer Subject: Browsing Classes? Date: 11 Oct 1996 00:12:21 GMT Organization: DSC Communications Corporation, Plano, Texas USA Message-ID: <53k3d5$t4m@camelot.dsccc.com> What tool in the NeXTDeveloper is used to browse NeXT Objective C classes? Thanks, James
From: dwy@mcny.com (David Young) Newsgroups: comp.sys.next.programmer Subject: Re: Q: Java compiler for NS 3.3 Date: 10 Oct 1996 16:54:41 GMT Organization: Media Connection of New York, Inc. Message-ID: <53j9oh$ia8@alice.walrus.com> References: <325A38D0.4B7D@mci.com> <53djb6$6qg@news-central.tiac.net> Aleksey Sudakov (zander@conextions.com) wrote: : >I've seen the guavac compiler on some of the ftp sites but it seems to : >be based on JDK 1.0 and isn't complete. Is there anything else out : >there? I'm a little interested in this as well. I saw kaffe/guavac, but like most post-1990 GNU projects, they leave alot to be desired. I tried porting my copy of the jdk 1.02 sources, with minimal success and alot of coredumps. Has anyone gotten any further than this? -- # david young: network engineer+oop developer # net: dwy@ace.net (NeXTmail ok) web: http://www.ace.net/ # vox: 212.686.3845 201.798.5217 fax: 212.686.3845
From: dboyce@albert (Doug Boyce) Newsgroups: comp.sys.next.programmer Subject: [Q] Compile problem with bind Date: 10 Oct 1996 17:48:49 GMT Organization: a Digital Internet AlphaServer Site Message-ID: <53jcu1$pn3@ns1.moran.com> I'm getting the following error when compiling db_dump.c in the 4.9.2 release. If someone can figure out the problem by just eye-balling the include file that would be great. Otherwise request more info. error: -------- dboyce [1:36pm][/mntb/Source/4.9_2-940221] make /mntb/Source/4.9_2-940221/res /mntb/Source/4.9_2-940221/include /mntb/Source/4.9_2-940221/include/arpa /mntb/Source/4.9_2-940221/compat for x in include lib; do (cd $x; make "DESTDIR=" "INCL=../../include" "COMPINCL=../../compat/include" "LIBDIR=/usr/lib" "CC=cc" "INSTALL=install" "RANLIB=ranlib" "LDS=" 'ARPREF=' 'ARSUFF=' all); done for x in sys; do (cd $x; make DESTDIR="" INSTALL="install" all); done /mntb/Source/4.9_2-940221/named cc -g -I../include -I../compat/include -D_PATH_XFER=\"/usr/libexec/named-xfer\" -c db_dump.c ns_func.h:19: illegal function definition, found `send_msg' ns_func.h:126: illegal function definition, found `dovalidate' *** Exit 1 Stop. *** Exit 1 Stop. ------ ns_funcs.h: ------ /* ns_proc.h - process declarations for ns_*.c's externally visible functions * * $Id: ns_func.h,v 1.6 1993/12/06 00:43:02 vixie Exp $ */ /* ++from ns_resp.c++ */ extern void ns_resp __P((u_char *, int)), prime_cache __P((void)), delete_all __P((struct namebuf *, int, int)); extern struct qinfo *sysquery __P((char *, int, int, struct in_addr *, int)); extern int doupdate __P((char *, int, u_char *, int, struct databuf **, int #ifdef CRED , u_int #endif )), send_msg __P((char *, int, struct qinfo *)), findns __P((struct namebuf **, int, struct databuf **, int *, int)), finddata __P((struct namebuf *, int, int, HEADER *, char **, int *, int *)), wanted __P((struct databuf *, int, int)), add_data __P((struct namebuf *, struct databuf **, u_char *, int)); /* --from ns_resp.c-- */ /* ++from ns_req.c++ */ extern void ns_req __P((u_char *, int, int, struct qstream *, struct sockaddr_in *, int)); extern int stale __P((struct databuf *)), make_rr __P((char *, struct databuf *, u_char *, int, int)), doaddinfo __P((HEADER *, u_char *, int)), doaddauth __P((HEADER *, u_char *, int, struct namebuf *, struct databuf *)); /* --from ns_req.c-- */ /* ++from ns_forw.c++ */ extern time_t retrytime __P((struct qinfo *)); extern int ns_forw __P((struct databuf *nsp[], u_char *msg, int msglen, struct sockaddr_in *fp, struct qstream *qsp, int dfd, struct qinfo **qpp, char *dname)), haveComplained __P((char *, char *)), nslookup __P((struct databuf *nsp[], struct qinfo *qp, char *syslogdname, char *sysloginfo)), qcomp __P((struct qserv *, struct qserv *)); extern struct qdatagram *aIsUs __P((struct in_addr)); extern void nslookupComplain __P((char *, char *, char *, char *, struct databuf *)), schedretry __P((struct qinfo *, time_t)), unsched __P((struct qinfo *)), retry __P((struct qinfo *)), qflush __P((void)), qremove __P((struct qinfo *)), qfree __P((struct qinfo *)); extern struct qinfo *qfindid __P((u_int16_t)), #ifdef DMALLOC *qnew_tagged __P((void)); # define qnew() qnew_tagged(__FILE__, __LINE__) #else *qnew(); #endif /* --from ns_forw.c-- */ /* ++from ns_main.c++ */ extern u_int32_t net_mask __P((struct in_addr)); extern void sqrm __P((struct qstream *)), sqflush __P((void)), sq_done __P((struct qstream *)), setproctitle __P((char *, int)), getnetconf __P((void)); extern struct netinfo *findnetinfo __P((struct in_addr)); /* --from ns_main.c-- */ /* ++from ns_maint.c++ */ extern void ns_maint __P((void)), sched_maint __P((void)), remove_zone __P((struct hashbuf *, int)), loadxfer __P((void)), qserial_answer __P((struct qinfo *, u_int32_t)); extern SIG_FN endxfer __P((void)); #ifdef DEBUG extern void printzoneinfo __P((int)); #endif /* --from ns_maint.c-- */ /* ++from ns_sort.c++ */ extern struct netinfo *local __P((struct sockaddr_in *)); extern void sort_response __P((char *, int, struct netinfo *, u_char *)); /* --from ns_sort.c-- */ /* ++from ns_init.c++ */ extern void ns_init __P((char *)); /* --from ns_init.c-- */ /* ++from ns_ncache.c++ */ extern void cache_n_resp __P((u_char *, int)); /* --from ns_ncache.c-- */ /* ++from ns_stats.c++ */ extern void ns_stats __P((void)); /* --from ns_stats.c-- */ /* ++from ns_validate.c++ */ extern int validate __P((char *, struct sockaddr_in *, int, int, char *, int #ifdef NCACHE , int #endif )), dovalidate __P((char *, int, u_char *, int, int, struct sockaddr_in *, int *)), update_msg __P((u_char *, int *, int Vlist[], int)); extern void store_name_addr __P((char *, struct in_addr *, char *, char *)); /* --from ns_validate.c-- */ ---
From: mark@slab.com (Mark Strand) Newsgroups: comp.sys.next.programmer Subject: Re: How to Browse ObjectiveC Classes? Date: 11 Oct 1996 05:26:28 GMT Organization: Internet Navigator, Inc. Message-ID: <53klq4$26j@composer.inav.net> References: <53hl89$1vu@news.onramp.net> In article <53hl89$1vu@news.onramp.net> jlincoln@foresight-sys.com writes: > I am trying to learn the NeXT Developer environment and need to know how to > browse all of the > ObjectiveC classes. > > Thanks, > > Jason HeaderViewer (pre-4.0) or ProjectBuilder's Frameworks entry (4.0+) HeaderViewer is very nice for letting you view classes in many different views, including in the class hierarchy.
From: Roberto Fabbretti 27231 <fabbretti-roberto@diogenes.huge.ch> Newsgroups: comp.sys.next.misc,comp.sys.next.programmer Subject: Java enabled browser for Next Date: Fri, 11 Oct 1996 09:30:39 +0200 Organization: University of Geneva Message-ID: <325DF79F.41C67EA6@diogenes.huge.ch> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I am considering the development of internet applications based on Next products using Java (evtl. JavaScript) but I found no reference about a Java enabled Browser for Next (Intel or Black). OmniWeb seems not to be able to interpret Java. Am I missing something. -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= | Roberto Fabbretti fabbretti-roberto@diogenes.hcuge.ch| =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
From: Bernhard Scholz <scholz@informatik.tu-muenchen.de> Newsgroups: comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Java enabled browser for Next Date: Fri, 11 Oct 1996 10:05:14 +0200 Organization: Technische Universitaet Muenchen, Germany Distribution: world Message-ID: <Pine.HPP.3.95.961011100417.23844C-100000@hphalle0.informatik.tu-muenchen.de> References: <325DF79F.41C67EA6@diogenes.huge.ch> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII In-Reply-To: <325DF79F.41C67EA6@diogenes.huge.ch> On Fri, 11 Oct 1996, Roberto Fabbretti 27231 wrote: > Java enabled Browser for Next (Intel or Black). OmniWeb seems not to be > able to interpret Java. Am I missing something. > No. :( This lack of technology will probably stop the NeXTworld one day... Greetings, Bernhard.
From: frank@this.net (Frank M. Siegert) Newsgroups: comp.sys.next.programmer,comp.sys.next.software,comp.sys.next.sysadmin Subject: Re: Can't compile TCP wrapper Date: 11 Oct 1996 12:08:32 GMT Organization: NO ORGANIZATION, INC. Message-ID: <53ldc0$h07@bias.ipc.uni-tuebingen.de> References: <53c0e8$4ui@zimmer.CSUFresno.EDU> Cc: kinau@shadow.csufresno.edu In <53c0e8$4ui@zimmer.CSUFresno.EDU> Kin Hung Au wrote: > Hi Folks, > > I can't compile TCP wrapper for NeXT black hardware. I get following message: > > ld: multiple definitions of symbol _closelog > ... Add -m to the link options in the Makefile: -m Don't treat multiply defined symbols as a hard error; instead, simply print a warning. The first such symbol is used for linking; its value is used for the symbol in the symbol table. The other symbols by the same name may be used in the resulting output file through local references. This can still produce a resulting output file that is in error. This flag's use is strongly discouraged! . and do not mind the last sentence from the man page, sometimes you need a hammer to get the thing going... Version 7.4 of the TCP wrappers has a special 'NeXTish' section in the Makefile where you should uncomment a few options when compiling on NeXTSTEP. If you are in the process of hardening your system be sure to replace the old portmapper by a wrapped version, the one from ftp://ftp.win.tue.nl/pub/security/index.html (look for portmap.tar.gz) works fine (at least for me) and will make rpc attacks much harder if configured correctly. It needs the TCPWrapper 7.4 library to compile which is also available from above site. -- * Frank M. Siegert [frank@this.net] - Home http://www.this.net * NeXTSTEP, Linux, BeOS & PostScript Guy
From: Scott Mewett <mewett@mpr.ca> Newsgroups: comp.sys.next.programmer Subject: How to adapt PhoneKit for 3.3?? Date: Fri, 11 Oct 1996 11:35:34 -0700 Organization: MPR Teltech Ltd., Burnaby, B.C., Canada Message-ID: <325E9376.3CA6@mpr.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I just recently got a HAYES ISDN Extender. Does anyone know what it would take to get the phone kit to work with NS 3.3. Is it possible at all or is it imposible. Or is there something else I could use to do stuff with the ISDN extender. Any Help would be appreciated. Scott
Date: 12 Oct 1996 06:25:40 GMT From: clewis@ferret.ocunix.on.ca (Chris Lewis) Sender: remove1@juno.com Message-ID: <cancel.53m9qb$and@usenet10.interramp.com> Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <53m9qb$and@usenet10.interramp.com> Control: cancel <53m9qb$and@usenet10.interramp.com> FREEREPORT spam cancelled by clewis@ferret.ocunix.on.ca Original Subject: ******FREE DEMO******
From: bbum@friday.com (Bill Bumgarner) Newsgroups: comp.sys.next.programmer Subject: CVS 1.9 + other gnu tools available Date: 12 Oct 1996 07:52:32 GMT Organization: Friday Software & Consulting Message-ID: <53nio0$aqd@chinx10.thoughtport.net> I have made a quad-fat compile of CVS version 1.9 -- with NEXTSTEP support -- and all supporting tools available on: ftp://ftp.thoughtport.net/pub/next/tools/ It is in the gnu-XXXXX.b.SHIN.tar.gz archive. It replaces any prior builds I maintained on that site... and requires a slightly different installation. Includes latest versions of grep, diff, cvs, rcs, tar and patch. Please let me know if you discover any problems or have any additions. This version of cvs includes support for automating the handling of binary files. b.bum
From: scollarw@cadvision.com (guzzibill) Newsgroups: comp.sys.next.programmer Subject: WWW survey Date: 12 Oct 1996 16:49:52 GMT Organization: CADVision Development Corp. Message-ID: <53oi7g$3geq@elmo.cadvision.com> I "lurk" this group regularly. Here's something applicable to all "Netizens". Check out the URL below with your favourite Web Browser. This year it is Java-enhanced, but for those like me "without" that capability, there are "regular" pages as well. www-survey.cc.gatech.edu results freely available to all. -- Bill Scollard - Scollard Holdings Ltd. Computer Systems : Cradle-to-Grave Calgary, Alberta, Canada
From: dwy@mcny.com (David Young) Newsgroups: comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Java enabled browser for Next Followup-To: comp.sys.next.misc,comp.sys.next.programmer Date: 13 Oct 1996 00:07:02 GMT Organization: Media Connection of New York, Inc. Distribution: world Message-ID: <53pbr6$icg@alice.walrus.com> References: <325DF79F.41C67EA6@diogenes.huge.ch> <Pine.HPP.3.95.961011100417.23844C-100000@hphalle0.informatik.tu-muenchen.de> Bernhard Scholz (scholz@informatik.tu-muenchen.de) wrote: : No. :( : This lack of technology will probably stop the NeXTworld one day... : Greetings, I don't think jack of Java will stop the NeXTworld, as it were. NeXT Marketing will stop the NeXTworld. -- # david young: network engineer+oop developer # net: dwy@ace.net (NeXTmail ok) web: http://www.ace.net/ # vox: 212.686.3845 201.798.5217 fax: 212.686.3845
From: ehutch@hypnos.norden1.com (E. Hutchinson) Newsgroups: comp.sys.next.programmer,comp.sys.next.misc,misc.jobs.contract Subject: NEXTSTEP/Contract/Long term/Va Date: 11 Oct 1996 19:50:31 GMT Organization: Altopia Corp. - Affordable Usenet Access - http://www.alt.net Message-ID: <53m8e7$8ef@tofu.alt.net> Organization: Norden 1 Communications Summary: Keywords: Programmer/analyst NEXTSTEP Objective C EOF--------A plus Commercial experience Virginia Must Be-----US Citizen or Greencard To Be Considered---Fax resume or mail a hard copy. -- ehutch@norden1.com (419) 893-6367 [fax] Omni Search (419) 893-6334 [voice] 1310 Craig Maumee, Ohio 43537
From: Joakim Johansson <jocke@rat.se> Newsgroups: comp.sys.next.programmer Subject: Re: DO problem under NEXTSTEP 3.3 "[NXConnection run] - tossing received reply msg" Date: 12 Oct 1996 15:53:56 GMT Organization: Research & Trade AB Distribution: world Message-ID: <53oeuk$ohf@baldwin.rat.se> References: <53agrk$jc8@ddfservb.technet.net> Constantin Szallies <szallies@energotec.de> writes > >> I guess one should view > >> >(that is: "@interface X:Object // <protocolXDO> oops!") > >> > >> as a bug in the application program. > > > >I dont think so, I haven't seen any references in the docs > >(those that exist... ;-) that the receiving proxy actually > >has to explicitly conform to the protocol. > > Client and server have a different message signature and it's not the fault of the PDO > system. I would consider this a bug of the application :-) > Well let's say it's a bug in the PDO documentation... Would they actually get different method signatures? (the method is declared as (oneway void) both in the header file and in the actual implementation! It's just that the class doesn't explicitly conform to the protocol.. Anyone care to explain exactly how/when the method signature information is gathered at compiletime?) Joakim -- Joakim Johansson Software Developer, Research & Trade jocke@rat.se <NeXTmail, MIME> http://www.rat.se/
From: hugues@precipice.fdn.fr (Hugues RICHARD) Newsgroups: comp.sys.next.programmer Subject: Re: Ghostscript 4.01 or 4.03 anyone? Date: 12 Oct 1996 11:47:33 GMT Organization: Individual - France Message-ID: <53o0gl$2sn@precipice.fdn.fr> References: <52rsvh$ftd@chinx10.thoughtport.net> <535g6d$3up@precipice.fdn.fr> <53c67p$b7u@ampere.visgen.com> bmw@visgen.com (Bruce Walker) wrote: > >Hugues RICHARD <hugues@precipice.fdn.fr> then wrote: > >>What such a paradox ! >>NS has a true PS Level 2 interpreter ready to rasterize for you and >>people use Ghostscript. What a ressource waste... >>And why ? Because of undocumentation... ? > >For me, it's because there's no way to use the DPS calls unless you >are logged-in. That spoils the NeXT DSP interpreter as far as being >used in a "server" application goes. about 8 months ago, Jason Beaver wrote in this group (about an lpd-fax driver) : > In article <4gek7g$3jg@bsdi201.britain.eu.net> >> richard@brainstorm.co.uk. (Richard Frith-Macdonald) writes: >> Q1. How do I get it to have a valid postscript context when >> running as a daemon? >> >You need to launch your daemon from /usr/lib/NextPrinter/lpd.children >to make it a child of lpd. Now, you know... Hugues. -------------------------------------------------------------------- hugues@precipice.fdn.fr - French, English, Italian and a few JP ->OK ------------ NS3.2 ------------ NS3.0J ------------ :-) ------------
From: andydunn@op.net (Andy Dunn) Newsgroups: comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Java enabled browser for Next Date: 11 Oct 1996 22:34:53 GMT Organization: OpNet -- Greater Philadelphia Internet Service Distribution: world Message-ID: <53mi2d$6bk@picasso.op.net> References: <325DF79F.41C67EA6@diogenes.huge.ch> <Pine.HPP.3.95.961011100417.23844C-100000@hphalle0.informatik.tu-muenchen.de> Cc: scholz@informatik.tu-muenchen.de > On Fri, 11 Oct 1996, Roberto Fabbretti 27231 wrote: > > > Java enabled Browser for Next (Intel or Black). OmniWeb seems not to be > > able to interpret Java. Am I missing something. > > > > Try Netsurfer. http://www.netsurfer.com
From: embuck@palmer.cca.rockwell.com (Erik M. Buck) Newsgroups: comp.sys.next.programmer Subject: Re: How to Browse ObjectiveC Classes? Date: 10 Oct 1996 20:32:22 GMT Organization: Rockwell Avionics - Collins Message-ID: <53jmgm$4lp@castor.cca.rockwell.com> References: <53hl89$1vu@news.onramp.net> Cc: jlincoln@foresight-sys.com In <53hl89$1vu@news.onramp.net> jlincoln@foresight-sys.com wrote: > I am trying to learn the NeXT Developer environment and need to know how to > browse all of the > ObjectiveC classes. > > Thanks, > > Jason > > Use HeaderViewer in /NextDeveloper/Apps or Librarian in /NextApps or any of the free Objective-C reverse-engineering tools
Newsgroups: comp.sys.next.programmer From: thomas@gamelan.shnet.org (Thomas Funke) Subject: Re: How to Browse ObjectiveC Classes? Message-ID: <1996Oct12.172416.582@gamelan.shnet.org> Sender: thomas@gamelan.shnet.org (thomas) Organization: Disorganization References: <53hl89$1vu@news.onramp.net> Date: Sat, 12 Oct 1996 17:24:16 GMT In article <53hl89$1vu@news.onramp.net> jlincoln@foresight-sys.com writes: > I am trying to learn the NeXT Developer environment and need to > know how to browse all of the ObjectiveC classes. > > Thanks, If you use the one and only editor, called emacs, then you should get a package 'oobr' from the archives. This is a cool OO browser for all major OO languages (C++, ObjC, Smalltalk, Java, Eiffel). Use archie to find a place to ftp from. (oobr-xxxxx.tar.gz) -- ----- Thomas Funke ----------------------- thomas@gamelan.shnet.org ----- C++ is to C as Lung Cancer is to Lung hosttype. I vaguely remember > something like m68k-next-nextstep. > No need to compile. -------------------------- cut here ------------------------- #!/bin/sh # # uname # # Author: Jim Vlcek, ByteWare Consulting (uunet!molly!vlcek) 3 Dec 1993 # Tweaked-By: Larry Blische (lkba!lkb@uunet.uu.net) # # An attempt to implement a SysV-ish "uname" under NeXTStep 3.0 # # Options # # -s Print the operating system name # -n Print the node name (essentially, the hostname) # -v Print the operating system version # -r Print the operating system release # -p Print the host machine's processor type # -m Print the machine hardware name # -a Print all the above information # # Non-SysV options # # -i Print the host identification number (hostid) # if [ $# = 0 ] then system="NEXTSTEP " fi for arg do case $arg in -s) system="NEXTSTEP " ;; -n) node="`uuname -l` " ;; -r) release="`hostinfo | sed -n 's/.*NeXT Mach \([0-9\.]*\).*/\1/p'` " ;; -m) mach="`hostinfo | sed -n 's/.*Processor type: \([^ ]*\).*/\1/p'` " ;; -p) processor="`hostinfo | sed -n 's/.*Processor type: [^ ]* (\([^)]*\).*/\1/p'` " ;; -v) version="`tail -1 /usr/lib/NextStep/software_version` " ;; -i) hostid="`hostid` " ;; -a) exec $0 -s -n -v -r -p -m ;; *) echo $0: Usage: $0 [-asnvrpm] >&2 ; exit 1 ;; esac done echo $system$node$version$release$processor$mach$hostid | sed 's/ $//' -- ----- Thomas Funke ----------------------- thomas@gamelan.shnet.org ----- C++ is to C as Lung Cancer is to Lung
From: "Eric A. Dubiel" <eadubie@rs6000.cmp.ilstu.edu> Newsgroups: comp.sys.next.misc,comp.sys.next.programmer Subject: Java in NetSurfer or OmniWeb any time soon? WAS:Re: Java enabled browser for Next Date: Sun, 13 Oct 1996 20:20:23 -0500 Organization: Illinois State University- Instructional Technology Services Message-ID: <32619550.2D53@rs6000.cmp.ilstu.edu> References: <325DF79F.41C67EA6@diogenes.huge.ch> <Pine.HPP.3.95.961011100417.23844C-100000@hphalle0.informatik.tu-muenchen.de> <53mi2d$6bk@picasso.op.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: info@netsurfer.com, info@omnigroup.com Any chance of seeing Java in NetSurfer or OmniWeb any time soon? A possibility would be for Omni and Netsurfer to colaborate to bring Java to the NeXT platform! This is the era of cooperation! ;) Thanks! -Eric Andy Dunn wrote: > > > On Fri, 11 Oct 1996, Roberto Fabbretti 27231 wrote: > > > > > Java enabled Browser for Next (Intel or Black). OmniWeb seems not to be > > > able to interpret Java. Am I missing something. > > > > Try Netsurfer. > > http://www.netsurfer.com -- Eric A. Dubiel; http://www.ilstu.edu/~eadubie mailto:eadubie@rs6000.cmp.ilstu.edu ASCII, MIME, SUN, NeXT, PGP Mail ytalk:eadubie@138.87.201.11 Instructional Technology Services- Illinois State University "Intelligence is the ultimate aphrodisiac." - Dr. Timothy Leary ALL VIEWS EXPRESSED REPRESENT MYSELF ONLY
From: bbum@friday.com (Bill Bumgarner) Newsgroups: comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Java in NetSurfer or OmniWeb any time soon? WAS:Re: Java enabled browser for Next Date: 14 Oct 1996 13:27:08 GMT Organization: Friday Software & Consulting Message-ID: <53tf3c$8pd@chinx10.thoughtport.net> References: <325DF79F.41C67EA6@diogenes.huge.ch> <Pine.HPP.3.95.961011100417.23844C-100000@hphalle0.informatik.tu-muenchen.de> <53mi2d$6bk@picasso.op.net> <32619550.2D53@rs6000.cmp.ilstu.edu> Cc: eadubie@rs6000.cmp.ilstu.edu Actually, there has been a NeXTSTEP oriented Java discussion group for some time-- send email to... next-java-request@friday.com ... to subscribe. Unfortunately, porting Java to NEXTSTEP is extremely difficult-- not the runtime, mind you, that was done a long time ago. It's the AWT that causes all the problems. JDK 1.0.2 sorta kinda works under NEXTSTEP-- the UI is far, far from complete. Long ago, the hooks were placed in Netsurfer to support Java-- other embedded languages have been supported for some time (TipTop's Objective-Perl/Tcl basically 'just works'; as do bundles of dynamically loaded obj-c code). I'm sure that-- with the recently documented 'plug-in' API-- OmniWeb would be capable of a similar level of language integration. Seeing as how NS 3.3 based technology is a dieing environment-- and NS 4.0 is very different, there is a whole slew of new and exciting porting issues. If you have the time to invest in dealing with said issues, please subscribe to the above list. If you are not a programmer or do not have the time to invest in porting issues-- don't bother subscribing; the contents of the list will not interest you [trust me]. If you wish to discuss the wonders of Java under NEXTSTEP-- do so here... it will do more for the community than discussing it on a private list. b.bum
Newsgroups: comp.sys.next.programmer Subject: Dictionary by reference :Example needed for DO From: shill@iphysiol.unil.ch (Sean Hill) Message-ID: <3262552a.0@cisun2000.unil.ch> Date: 14 Oct 96 14:58:50 GMT Can anyone send me an example of getting an NSDictionary by _reference_ and examining (even changing!) it's contents using an NXConnection (EOF 1.1 or OS 4.0)? I've tried defining a protocol that insisted the object be by reference but it has an encoding problem. Thanks- Sean shill@iphysiol.unil.ch
Newsgroups: comp.sys.next.programmer From: Mark Lee <mlee@impact.com> Subject: Palatte Problem Under Openstep 4.0 Content-Type: text/plain; charset=us-ascii Message-ID: <32625639.C2@impact.com> Sender: news@pfizer.com (News Admin) Content-Transfer-Encoding: 7bit Cc: mlee@impact.com Organization: Impact Software Mime-Version: 1.0 Date: Mon, 14 Oct 1996 15:03:21 GMT I'm converting a 3.3 app to 4.0. The app uses a palatte. When I try to compile under 4.0 I get multiple header conflict warnings/errors. I then tried to create an empty palatte within a new project under 4.0 and got the same error messages. This appears to be a generic problem with the Framework headers. I was wondering if anyone had some suggestions/workarounds for this problems. BTW, my Postscript code compiles cleanly. Thanks.
From: wilkie@cg.tuwien.ac.at (Alexander Wilkie) Newsgroups: comp.sys.next.programmer Subject: [Q] Workspace mount/unmount for removable media? Date: 14 Oct 1996 14:45:46 GMT Organization: Vienna University of Technology, Austria Message-ID: <53tjmq$4go@news.tuwien.ac.at> Hi newsgroup! What I'm trying to do is to copy the tools that iomega corp. ship with their Jaz and Zip drives. The main objective of these tools is to control the software write protection feature found on these devices. While it is no problem do set/unset the write protection itself (simple SCSI commands; I've uploaded a command-line demo called ziptool to the peak NeXT ftp area), the NeXT Workspace does not seem to particularly like someone changing the type of a mount from "rw" to "ro" (or vice versa) behind its back; apparently it maintains some kind of internal state with respect to mounts that one cannot easily sync with reality from outside the app. All the Workspace request protocol offers is a "unmountAndEject" command. What I need would be a single "unmount" command; that it is missing form the protocol can probably be explained by the absence of software-controlled write protection features when WM 3.3 was hatched (for what else would one need something like this?). The ccommand sequence I'd like to perform would be something like - stall automount daemon (to prevent interference) - unmount disk in question - change write protection - re-mount disk - revive automunter Frustratingly, otool reveals plenty of single "unmounts" within the Workspace.app executable (any way to call these?). I've tried plenty of manual un/re-mount strategies, all of which confuse WM.app to some degree and damage the filesystem of the disks that are being processed. Sooo.... Could anyone with in-depth knowledge of NS 3.3 in general and WM 3.3 in particular comment on this? Someone at NeXT listening? Right now I'm out of ideas on how to get this to work and I'd greatly appreciate any information I can get. ys Alexander Wilkie -- e-mail: wilkie@cg.tuwien.ac.at (NeXTMail preferred, MIME o.k.) www : http://www.cg.tuwien.ac.at/~wilkie/-- e-mail: wilkie@cg.tuwien.ac.at (NeXTMail preferred, MIME o.k.) www : http://www.cg.tuwien.ac.at/~wilkie/
From: shess@shell.one.net (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: [Q] Workspace mount/unmount for removable media? Date: 14 Oct 1996 11:45:58 -0400 Organization: Is a sign of weakness Sender: shess@shell.one.net Message-ID: <veehgnxzgzt.fsf@shell.one.net> References: <53tjmq$4go@news.tuwien.ac.at> In-reply-to: wilkie@cg.tuwien.ac.at's message of 14 Oct 1996 14:45:46 GMT In article <53tjmq$4go@news.tuwien.ac.at>, wilkie@cg.tuwien.ac.at (Alexander Wilkie) writes: Could anyone with in-depth knowledge of NS 3.3 in general and WM 3.3 in particular comment on this? Someone at NeXT listening? You should be able to do "umount /DirName" or "unmount /dev/sd2a" or somesuch from the command-line. Perhaps /dev/rsd2a, I can't recall. This will unmount the disk but _not_ eject it. You can use "disk -e /dev/rsd2a" to eject a disk. Note, though, that this only works if there is nobody currently referencing the disk, just like Workspace Eject. From there, you can do your SCSI magic, and then do "mount /dev/sd2a /DirName" to get things back online. You'll probably need to play around with the mount parameters (see "man 8 mount") to get things exactly right. Note that umount and mount need to be run as root. And if you want to integrate the calls, the mount(2) and umount(2) system calls need to be run as root. That means your program will need to be installed setuid root to function correctly. Workspace _is_ installed this way, because NeXT got to define how systems are built ... Later, -- scott hess <shess@winternet.com> http://www.winternet.com/~shess/ Work: 288 Rampart Court #105, Ft Mitchell, KY, 41017 (606) 578-0412 (Was) 12550 Portland Avenue South #121, Burnsville, MN 55337 (612)895-1208 <address and phone work now, I am already in Cincinnati. No new email, yet.>
From: aeodoux@worldnet.fr (Alexandre Odoux) Newsgroups: comp.sys.next.programmer Subject: Openstep (port|comp)atibility Sun vs NeXT Date: Mon, 14 Oct 1996 17:44:50 GMT Organization: SCT / Worldnet - Internet Provider & Information Exchange - Paris, France Distribution: world Message-ID: <53tu7g$oih@news.sct.fr> Hi, has someone some experience about the compatibility between Openstep 4.x / Openstep for NT and Sun's Openstep. For example, it seems that NSTableView has become part of the Appkit (in 4.0 ?). Is this valid for Sun's Openstep as well ? And what about the source compatibility between Openstep NT and Openstep Mach ? Is it hard to maintain ? (FYI I don't have access to any of the Openstep developer kits). Regards - Alexandre
From: "James P. Klett" <klett002@maroon.tc.umn.edu> Newsgroups: comp.sys.next.programmer Subject: IntraBuilder? Date: Mon, 14 Oct 1996 13:30:05 -0500 Organization: Emprise Consulting Services, Inc. Message-ID: <326286AD.3968@maroon.tc.umn.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Has anyone seen Borland's IntraBuilder? I am wondering how it compares to NeXT's WebObjects. Their Client/Server price is $1995 and the interface building part looks just like a typical form designer from any one of Borlands DB packages. -- ___ ___....-----'---'-----....___ ========================================= ___'---..._______...---'___ (___) _|_|_|_ (___) \\____.-'_.---._'-.____// klett002@maroon.tc.umn.edu
From: Fame@success.net (Walter Brooke) Newsgroups: comp.sys.next.programmer Subject: Make Money Date: Tue, 15 Oct 1996 02:15:18 GMT Organization: Netcom Message-ID: <3262f358.23713058@nntp.ix.netcom.com> >> Hello everyone, >> >> I used to think this was a huge joke until I tried it - the "turn $5 >> into $50,000" idea. But after awhile I started to realize "This >> either works or it's a complete scam - but I intend to find out." >> >> Guess what? It works! In only three weeks, I've received $1827!! I >> don't know anyone who couldn't use a extra $1800 do you? And there is >> certainly lots more to come for me - up to $55,000!!! >> >> What about you? You can do it with only $5 and 5 postage stamps >> (that's only $6.60 in the USA - hardly what I call a risky investment >> when you consider the rewards). Everyone who tries this will make >> much, much more than that, so what have you got to loose? $5.00 plus >> postage? Big Deal! >> >> Hopefully you can see that this really does work and you'll read on. >> >> Follow this simple, perfectly legal 3 step procedure and you can't go >> wrong; >> >> ===================================================== >> First Step - Type or write your name and address on 5 separate pieces >> of paper with the words "PLEASE ADD ME TO YOUR MAILING LIST". Fold a >> $1 bill or money order in each of the pieces of paper, place them in 5 >> envelopes and mail one to each of the following five addresses: >> >> 1- Robert Knowlton >> PO Box 924 >> Concord, NH 03302-0924 >> >> 2- JCL >> 1313 Wake Forest Dr. #125 >> Davis, CA 95616 >> >> 3- RFP >> 9594 1st Ave, NE, #423 >> Seattle, WA 98115 >> >> 4- Bruce Fuller >> 293 Chestnut ST. >> Chillicothe, OH 45601 >> >> 5- Walter Brooke >> 932 Lake Nora South Dr. Apt.C >> Indianapolis, IN 46240 >> USA >> >> Second Step - Now remove the 1st name on the list, move the other 4 >> names up (5 becomes 4, 4 becomes 3 etc.) and put your name and address >> as number 5 on the list. You can do this by re-typing this article or >> simply editing it in a word processor. >> >> Third Step - Now Post your adjusted article to at least 200 Usenet >> news groups (there are over 20,000 of them). >> >> That's it! You are now in the Mail Order Investment Business and you >> will start receiving $1 returns by mail within a week or two. The more >> newsgroups you post to, the bigger your return will be. You may even >> want to rent a P O Box to handle the volume of mail you're likely to >> receive. >> >> If you wish to remain anonymous, you can use a pseudonym such as "The >> Manager" or "The Investor", but make sure your snail mail address is >> correct! >> >> How does this work? >> >> Of every 200 postings I made, I received an average of 5 replies for >> $5. Ok that may not seem like much but read on.... >> >> Each person who sent you $1 also makes - let's say, only 200 >> additional postings WITH YOUR NAME AT NUMBER 4, i.e. 1000 postings. >> >> On average therefore, 50 people will send you $1 with your name at >> number 4. You make $50. >> >> Your 50 new agents make 200 posting each WITH YOUR NAME AT NUMBER 3 or >> 10,000 postings - average return 500 at $1 each is $500. They make 200 >> postings each WITH YOUR NAME AT NUMBER 2 = 100 000 postings = 5 000 >> returns at $1 each = $5 000. >> >> Finally, 5,000 people make 200 postings each WITH YOUR NAME AT NUMBER >> 1 and you get a return of $50,000 before you name drops off the list. >> AND THAT'S IF EVERYONE DOWN THE LINE ONLY MAKES 200 POSTINGS! Total >> income in one cycle = $55,500. >> >> And when you drop off the list - simply start over! >> >> Remember, this only costs Five dollars, 5 postage stamps and some >> online time. So what are you waiting for? >> >> You may be thinking "This has been around a while - What happens if >> the scheme is 'played out" and no one sends me any money"? >> >> Big deal! So you lose $5 - but with hundreds of thousands of new >> people joining the internet every day thats hardly likely to happen. >> >> Remember, follow the instructions and play fair - send those 5 >> one-dollar bills (this process is based on honesty not on cheating on >> the initial investment). Perhaps we'll meet on the beach in Hawaii. >> That is what my $1827 is going towards (I'll have to think about what >> to do with the rest of the $$$$ that is coming my way!)
Newsgroups: comp.sys.next.programmer Control: cancel <3262f358.23713058@nntp.ix.netcom.com> Subject: cmsg cancel <3262f358.23713058@nntp.ix.netcom.com> From: snowhare@xmission.com Sender: Fame@success.net (Walter Brooke) Date: 15 Oct 96 02:47:51 GMT Message-ID: <cancel.3262f358.23713058@nntp.ix.netcom.com> Organization: Devilbunnies Spam Cancelling Division Spam Cancellation. Original Subject: Make Money A copy of the original message can be found at <URL:http://www.xmission.com/~snowhare/spam/udp/mmf-action3/845347671.90>
From: jobs@globalobjects.com Newsgroups: comp.lang.java,comp.lang.smalltalk,misc.jobs.offered,misc.jobs.offered.entry,misc.jobs.contract,comp.sys.next.programmer,atl.jobs Subject: US - GA - OO developers needed: WebObjects, NextStep, Objective-C, EOF, Java, Web, Smalltalk, C++ Date: Tue, 15 Oct 1996 09:58:59 -0500 Organization: Global Objects Inc. Message-ID: <MPG.ccd70c1950961eb989680@news-s01.ny.us.ibm.net> Keywords: OO, WebObjects, EOF, Objective-C, NextStep, Smalltalk, Java, C++ Global Objects is a professional software service company that focuses in providing quality solutions through object technology. We specialize in object modelling, Web, Java, NextStep, Smalltalk, and C++ projects in various domains including telecommunications, petroleum, transportation and environmental services. We also provide quality courses on object technology including Java, design patterns, frameworks, object oriented analysis and design, and software development methodologies. We are experiencing rapid growth and are looking for highly qualified candidates, senior and junior ones. Work with a team of work with a team of world-class object developers to deliver state-of-the-art solutions on the Web. Get hands-on-training on exciting new technologies such as WebObjects, Java and electronic commerce. Requirements are as follows: Experience in at least one of the following area (prefer industry experience but will consider strong academic experience): * NextStep / WebObjects / EOF / Objective-C * Java * Any other Web development environment (Cold Fusion, Sapphire, etc.) * Smalltalk * C++ or any other OO language Unix experience helpful. For immediate consideration, please e-mail your resume (preferrably in Word format or in ASCII) to jobs@globalobjects.com or fax it to 770.457.7333. Please include the following information: * availability * for full-time employment, pls specify salary history * for contract employment, pls specify rate
From: dietzsch@rmhs2.urz.tu-dresden.de (Andreas Dietzsch) Newsgroups: comp.sys.next.programmer Subject: reading NSCalendarDate Date: 15 Oct 1996 13:15:09 GMT Organization: TU Dresden (URZ) Message-ID: <5402ot$r86@rks1.urz.tu-dresden.de> How can I get a NSString from a NSCalendarDate? Thank You Andreas
From: dboyce@albert (Doug Boyce) Newsgroups: comp.sys.next.programmer Subject: Re: [Q] Compile problem with bind Date: 15 Oct 1996 18:29:30 GMT Organization: a Digital Internet AlphaServer Site Message-ID: <540l6a$f1s@ns1.moran.com> References: <53jcu1$pn3@ns1.moran.com> Never mind. I found the problem and it compiles fine. -- /dwb
From: wilkie@cg.tuwien.ac.at (Alexander Wilkie) Newsgroups: comp.sys.next.programmer Subject: Re: [Q] Workspace mount/unmount for removable media? Date: 15 Oct 1996 21:00:23 GMT Organization: Vienna University of Technology, Austria Message-ID: <540u17$hf6@news.tuwien.ac.at> References: <53tjmq$4go@news.tuwien.ac.at> <veehgnxzgzt.fsf@shell.one.net> Cc: shess@shell.one.net Hi newsgroup! Scott Hess wrote: > You should be able to do "umount /DirName" or "unmount /dev/sd2a" or > somesuch from the command-line. Perhaps /dev/rsd2a, I can't recall. > This will unmount the disk but _not_ eject it. .... [snip] Yeah, sure, that's precisely what I'm doing at the moment, and it's working fine, except for the fact that Workspace.app does not take kindly to someone changing the _type_ of a mount (or to just ejecting a disk) behind it's back. Apparently "umount /DirName" etc. performed from the command line do _not_ cause the Workspace to gracefully terminate the mount as far as some internal state is concerned. And since this is NeXTland, I thought I might as well try to get my app to cooperate well with _the_ central file manipulation tool. Not that I'm so keen on performing a sequence like "unmount - change write permissions - remount" (which obviously is somewhat nasty for WM.app to swallow). A simple "unmount - change permissions - eject" would do fine for me, but if you do this, disks that go from "rw" to "ro" sometimes (not always!) are reported to be "damaged" the next time you insert them. Doesn't umount sync the devices it is about to unmount? ys Alexander Wilkie -- e-mail: wilkie@cg.tuwien.ac.at (NeXTMail preferred, MIME o.k.) www : http://www.cg.tuwien.ac.at/~wilkie/
From: yannick@uranus.univ-lr.fr (Yannick Buisson) Newsgroups: comp.sys.next.programmer Subject: mail and files Date: 16 Oct 1996 08:40:47 GMT Organization: Universite de La Rochelle Message-ID: <54272f$628@hpuniv.univ-lr.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Hi all, I want to use an API for sending mail. I want to automaticaly include files in the mail. Is it Possible ? how ? Are there some examples ? thanks for your answer. -- //// (. .) ----oOO--(_)--OOo-------------------------------------------- Yannick BUISSON Centre de Ressources Informatiques Université de La Rochelle tel prof. : 46 45 82 14. fax prof. : 46 45 82 45. Email (NeXTMail , MIME) : -> yannick@cri.univ-lr.fr
From: Bernhard Scholz <scholz@informatik.tu-muenchen.de> Newsgroups: comp.sys.next.programmer Subject: Re: [Q] Workspace mount/unmount for removable media? Date: Wed, 16 Oct 1996 10:59:27 +0200 Organization: Technische Universitaet Muenchen, Germany Distribution: world Message-ID: <Pine.HPP.3.95.961016105805.8143A-100000@hphalle0.informatik.tu-muenchen.de> References: <53tjmq$4go@news.tuwien.ac.at> <veehgnxzgzt.fsf@shell.one.net> <540u17$hf6@news.tuwien.ac.at> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII In-Reply-To: <540u17$hf6@news.tuwien.ac.at> On 15 Oct 1996, Alexander Wilkie wrote: > swallow). A simple "unmount - change permissions - eject" would do fine for > me, but if you do this, disks that go from "rw" to "ro" sometimes (not > always!) are reported to be "damaged" the next time you insert them. Doesn't > umount sync the devices it is about to unmount? > I only experienced this behaviour with OS4.0 and automounted disks. It showed up to me every secend umount/mount. Greetings, Bernhard.
From: Petr Novak <novak@microcomp.de> Newsgroups: comp.sys.next.programmer Subject: Dll library Date: Wed, 16 Oct 1996 13:15:54 +0200 Organization: Microcomp GmbH , Cologne, Germany Message-ID: <3264C3EA.7093@microcomp.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I work on 4.1 Beta for NT. I try bind a DLL library in project. The compiler automatically adds suffixes '.lib' or '.a' to the name of library and cannot find this library. How to change this default behaviour ? Petr
From: Laurent.Champciaux@emn.fr Newsgroups: comp.sys.next.programmer Subject: How to know if a key is pressed ? Date: 16 Oct 1996 12:00:58 GMT Organization: Ecole des Mines de Nantes Message-ID: <542ipq$5er@wfn.emn.fr> Hi, I cannot find how to detect that the alternate key is pressed or not! I don t want to wait for such an event, i.e. the flagsChanged event. Any clue ? I use NEXSTEP 3.3 on a SPARC5. Thanks in advance. Laurent. -- ======================================================= Laurent Champciaux Departement Informatique - Ecole des Mines de Nantes 4, rue A.Kastler 44070 Nantes Cedex 03 Tel: (33)51 85 82 18 (B220) email: Laurent.Champciaux@emn.fr WWW: http://www.emn.fr/dept_info/perso/laurent/
Newsgroups: comp.sys.next.programmer Organization: Antigone Press gateway, San Francisco Return-Path: <gmcfarla@LandP.com> X400-Received: by mta EMX01.LANDP.COM in /PRMD=LANDP/ADMD=ATTMAIL/C=GB/; Relayed; Mon, 14 Oct 1996 15:20:27 +0100 X400-Received: by mta ldnmail in /PRMD=LANDP/ADMD=ATTMAIL/C=GB/; Relayed; Mon, 14 Oct 1996 15:16:11 +0100 X400-Received: by mta chmail in /PRMD=LANDP/ADMD=ATTMAIL/C=GB/; Relayed; Mon, 14 Oct 1996 15:16:11 +0100 X400-Received: by mta ancient in /PRMD=LANDP/ADMD=ATTMAIL/C=GB/; Relayed; Mon, 14 Oct 1996 15:15:53 +0100 X400-Received: by mta NeXT.Mailer in /PRMD=LANDP/ADMD=ATTMAIL/C=GB/; Relayed; Mon, 14 Oct 1996 15:20:28 +0100 X400-Received: by /PRMD=LANDP/ADMD=ATTMAIL/C=GB/; Relayed; Mon, 14 Oct 1996 15:15:50 +0100 Date: Mon, 14 Oct 1996 15:15:50 +0100 X400-Originator: gmcfarla@LandP.com X400-Recipients: comp-sys-next-programmer@antigone.com X400-Mts-Identifier: [/PRMD=LANDP/ADMD=ATTMAIL/C=GB/;0005100001476834000002] X400-Content-Type: P2-1988 (22) Content-Identifier: Reuse library... From: Garrick McFarlane <gmcfarla@LandP.com> Message-ID: <9610141416.AA18933@chmail> Subject: Reuse library sought: Hierarchical Outlining Original-Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 3.3risc v118.3) Hello all As part of an in-house OpenStep development project, we need to build a hierarchical, fully manipulable outline view, such as is found in Concurrence by Lighthouse. (A less featured example is the Bookmarks window in OmniWeb, or perhaps the Outlets/Actions list in Interface Builder.) This would probably be fairly complex to build from scratch. If anyone could point me towards any available reuse resources (source, libraries, palettes, anything...) I would be extremely grateful. Please cc: my email account on any responses as I no longer appear to receive c.s.n.programmer articles (are the antigone digests still working?). Much appreciated. Garrick McFarlane Linklaters and Paines, London gmcfarla@landp.com +44 171 606 7080
From: eric@skatter.USask.Ca Newsgroups: comp.sys.next.programmer Subject: TeX and OpenStep 4.0 Date: 16 Oct 1996 15:55:43 GMT Organization: University of Saskatchewan Message-ID: <5430hv$ran@tribune.usask.ca> LaTeX and TeXview.app are a couple of the tools I use a *lot* on my NEXTSTEP machine. I was very disappointed to see that these have been dropped from the OPENSTEP 4.0 distribution. So far I've been able to get teTeX-0.4 installed and running on the Intel and NeXT machines here. The next tasks are to get TeXview.app converted to OPENSTEP and to add the `kpathsea' code. I started with the conversion when it occurred to me that perhaps someone else has already done this. If so, are the results available on any of the archive sites? -- Eric Norum eric@skatter.usask.ca Saskatchewan Accelerator Laboratory Phone: (306) 966-6308 University of Saskatchewan FAX: (306) 966-6058 Saskatoon, Canada. NeXTMail accepted.
From: Mark_Bessey@next.com (Mark Bessey) Newsgroups: comp.sys.next.programmer Subject: Re: Openstep (port|comp)atibility Sun vs NeXT Date: 16 Oct 1996 18:42:14 GMT Organization: NeXT Software, Inc. Distribution: world Message-ID: <543aa6$n12@news.next.com> References: <53tu7g$oih@news.sct.fr> Alexandre Odoux writes > Hi, > > has someone some experience about the compatibility between Openstep > 4.x / Openstep for NT and Sun's Openstep. I really don't know about Solaris, but compatibility between NeXT's implementations of OpenStep is very good. > For example, it seems that NSTableView has become part of the Appkit > (in 4.0 ?). Is this valid for Sun's Openstep as well ? It doesn't look like it. OPENSTEP/Mach's headers have NSTAbleView as #ifndef STRICT_OPENSTEP So, it looks to be a NeXT-only enhancement. > And what about the source compatibility between Openstep NT and > Openstep Mach ? Is it hard to maintain ? See above. If you're careful to use Foundation where it makes sense to do so, most code is portable between Windows and Mach. > (FYI I don't have access to any of the Openstep developer kits). > > Regards - Alexandre Hope this helps, -Mark -- Mark Bessey NeXT Software, Inc Software Quality Assurance -->I DON'T SPEAK FOR NeXT <--
From: altenber@acpub.duke.edu (Lee Altenberg) Newsgroups: comp.sys.next.programmer Subject: Dealing with 32 views at once Date: 16 Oct 1996 20:19:34 -0400 Organization: Duke University, Durham, NC, USA Message-ID: <543u2m$8rd@news.duke.edu> I am building an application that has 32 small custom views. Well, I made them a bit too small, and want to enlarge them. I would like to be able to do this programmatically, but have run into some problems. I can enlarge the custom views, but each is surrounded by a box, and I don't know how to access them. Each custom view is connected through Interface Builder to an instance variable in my Controller class. But that doesn't allow me to access the box that surrounds each one. I tried using [aCustomView superview];, but that doesn't return the box. Do I need to create 32 new instance variables in Controller and connect each box to one by hand? Should I figure out a programmatic way to create the custom views and their boxes in the first place, and not use Interface Builder to create them? Any tips will be greatly appreciated. ======================================================================= Lee Altenberg, Ph.D. Research Affiliate, University of Hawai`i at Manoa Office: Maui High Performance Computing Center 550 Lipoa Parkway, Suite 100 Kihei, Maui HI 96753 Phone: (808) 879-5077 x 296 (work), (808) 879-5018 (fax) E-mail: altenber@mhpcc.edu <NeXTMail and MIME: altenber@pueo.mhpcc.edu> Web: ¬http://pueo.mhpcc.edu/~altenber/ =======================================================================
From: root@starbase1.gpl.net (Dale Amon as Operator) Newsgroups: comp.sys.next.programmer Subject: Re: cmu_snmp2 Date: 17 Oct 1996 04:59:54 GMT Organization: Genesis Project Ltd Message-ID: <544ega$3c4@divis.gpl.net> References: <544cqr$32t@divis.gpl.net> A PS to that... I've got the client side working, its' just the agent side that looks messy. The biggest secret to it all is the seemingly undocumented -posix flag in the gcc used in 3.2. Cures many ills... But if anyone *has* done anything with the agent side, I'd really appreciate some info. Dale Amon <amon@gpl.com>
From: gvandyk@icon.co.za Newsgroups: comp.sys.next.programmer Subject: Re: Openstep (port|comp)atibility Sun vs NeXT Date: 17 Oct 1996 04:41:28 GMT Organization: E.S. Systems cc (Financial Systems Development) Message-ID: <544ddo$o66@hermes.is.co.za> References: <53tu7g$oih@news.sct.fr> <543aa6$n12@news.next.com> In-Reply-To: <543aa6$n12@news.next.com> On 10/16/96, Mark Bessey wrote: >Alexandre Odoux writes >> For example, it seems that NSTableView has become part of the Appkit >> (in 4.0 ?). Is this valid for Sun's Openstep as well ? > >It doesn't look like it. OPENSTEP/Mach's headers have NSTAbleView as > #ifndef STRICT_OPENSTEP >So, it looks to be a NeXT-only enhancement. > NSTableView is part of Openstep for Sun. If you get the developer you will also get the docs for this. I haven't used it on Sun/Openstep yet, but the headers and docs are all there so I would presume that the classes exist The Palette for NSTableView also exists. I personally think that Sun has made a very good port of Openstep. It looks much better on Sun than it does on Windows/NT. On NT you basically only get the runtime environment, no useful apps etc, etc. On sun you get the FileViewer and Mail as well. -- Regards, Gerrit van Dyk email: gvandyk@icon.co.za (NeXTMail welcome) E.S. Systems cc The OBJECT is the ADVANTAGE
From: frontier@fii.com Newsgroups: comp.sys.next.programmer Subject: 2Mx32 60ns $34.00 Date: Wed, 16 Oct 1996 23:08:56 Organization: Delta Internet Services, Anaheim, CA Message-ID: <543pik$cok@news05.deltanet.com> Hi there, I am the Marketing Manager of Frontier Industrial, I would like to let you know of special that we have on 8 MB SIMMS. I can offer you 2MX32 60ns modules @ $34.00 per piece. VARS, DEALERS, DISTRIBUTORS, & SYSTEM INTEGRATORS ONLY PLEASE Frontier Industrial http://www.fii.com
From: vadim@physics.utexas.edu Newsgroups: comp.sys.next.programmer Subject: Re: TeX and OpenStep 4.0 Date: 17 Oct 1996 06:39:28 GMT Organization: The University of Texas at Austin, Austin, Texas Message-ID: <544kb0$d8j@geraldo.cc.utexas.edu> References: <5430hv$ran@tribune.usask.ca> eric@skatter.USask.Ca wrote: > LaTeX and TeXview.app are a couple of the tools I use a *lot* on my NEXTSTEP > machine. I was very disappointed to see that these have been dropped from > the OPENSTEP 4.0 distribution. > > So far I've been able to get teTeX-0.4 installed and running on the Intel and > NeXT machines here. The next tasks are to get TeXview.app converted to > OPENSTEP and to add the `kpathsea' code. I started with the conversion when > it occurred to me that perhaps someone else has already done this. If so, > are the results available on any of the archive sites? > I had exactly the same problem, so I got the "NeXTified" teTeX (unfortunately, v-0.35 rather than v-0.4; I dunno if the difference is important) and the kp-compatible TeXview.app and HyperTeXview.app from the heidelberg archive: ftp://zarquon.mathi.uni-heidelberg.de/pub/NeXT/TeX/ Installation was completely straightforward. teTeX needs installing two .tar.gz packages (binaries & libraries) into /usr/local/teTeX, making a couple of symlinks for legacy's sake, adding /usr/local/teTeX/bin to the shells' execution path and running texconfig to set up printers' modes. That's all, unless you upgrade from earlier NeXTTeX, in which case each existing user needs to update the path in his/her .cshrc (or PATH in .profile) and replace old-style TEXINPUTS env.variables with kpathsea versions. TeXview.app & HyperTeXview.app work "straight out of the box", no hacking needed. Set preferences to "generate fonts as needed", then TeXview you favorite .dvi files to generate the .pk fonts. Dmitri Linde's InstantTeX (TeXEdit, TeXWorks, etc.) work with teTeX just fine, but you must modify the old-style default TEX... paths in the TeXWorks preferences. (On a multi-user system, each user has to do it for him/herself). Specifically, set TEXPOOLS to *nothing* (blank); a safe default is to set all paths to nothing. Users with private macros should set TEXINPUTS to something like .:$HOME/TeX/macros//: (the trailing colon appends the teTeX defaults). Have fun. -- Vadim. ***************************************************************************** Vadim S. Kaplunovsky, | vadim@physics.utexas.edu (NextMail OK) Associate Professor of Physics, | #include <std_disclaimer.h> University of Texas at Austin. | #excuse bad_typing.
From: root@starbase1.gpl.net (Dale Amon as Operator) Newsgroups: comp.sys.next.programmer Subject: cmu_snmp2 Date: 17 Oct 1996 04:31:23 GMT Organization: Genesis Project Ltd Message-ID: <544cqr$32t@divis.gpl.net> Has anyone ported cmu_snmp2 to NSFIP 3.2? I've gotten as far as the snmp_vars.c module... at which point I got the sinking feeling that this is not an evenings work... Please respond via email to: Dale Amon <amon@gpl.com>
From: ehutch@hypnos.norden1.com (E. Hutchinson) Newsgroups: comp.sys.next.programmer,comp.sys.next.misc,misc.jobs.contract Subject: NEXTSTEP/Contract/Immediate Date: 17 Oct 1996 13:22:35 GMT Organization: Norden 1 Communications Message-ID: <545bur$ajt@tofu.alt.net> Programmer/analyst/developer NEXTSTEP Objective C Contract---Long term Area-------Virginia EOF--------A plus Must Be----US Citizen or Greencard To Be Considered---Fax resume or mail a hard copy. -- ehutch@norden1.com (419) 893-6367 [fax] Omni Search (419) 893-6334 [voice] 1310 Craig Maumee, Ohio 43537
From: 1kyriaki@swt1.informatik.uni-hamburg.de (Valentino Kyriakides) Newsgroups: comp.sys.next.programmer Subject: H: add/del subview with dynam. winresize Date: 17 Oct 1996 13:47:11 GMT Organization: University of Hamburg -- Germany Message-ID: <545dd0$5d7@rzsun02.rrz.uni-hamburg.de> Hi, can somebody tell me where I can find some sample source or how to implement the dynamic adding and removing of a subview with automatic window resizing (growing/shrinking)? What I exactly need is what NeXT's "Sound.app" does if one presses the 'Edit (arrow down)' button. There the window resizes from top to bottom and adds a view, when the button is pressed again, the window resizes back from bottom to top and removes the view! It would also be interesting to know, how to cover/discover portions of a window dynamically from a window without adding/removing subviews! However, it would be a great help for me, if somebody could tell me how to do such things (small examples are wellcome) as "Sound.app" does. Or if somebody can point me to an MiniExample showing this! Thanks in advance for your help Valentino -- Valentino Kyriakides Email: 1kyriaki@informatik.uni-hamburg.de (NO NeXTmail available here)
Newsgroups: comp.sys.next.programmer From: dsr@lns598.lns.cornell.edu Sender: frontier@fii.com Date: 17 Oct 1996 10:02:13 EDT Control: cancel <543pik$cok@news05.deltanet.com> Subject: cmsg cancel <543pik$cok@news05.deltanet.com> no reply ignore Message-ID: <cancel.543pik$cok@news05.deltanet.com> Spam/MMF cancelled by dsr@lns598.lns.cornell.edu original subject was 2Mx32 60ns $34.00
From: bbum@friday.com (Bill Bumgarner) Newsgroups: comp.sys.next.programmer Subject: Re: Dealing with 32 views at once Date: 17 Oct 1996 15:39:53 GMT Organization: Friday Software & Consulting Message-ID: <545k09$fha@chinx10.thoughtport.net> References: <543u2m$8rd@news.duke.edu> Cc: altenber@acpub.duke.edu There are two answers to this question; 1) Assuming you continue to use 32 sub-views, if you ask for one of the view's -superview, it should return the id of the box it is contained within... Actually; it'll likely return the Box's contentview. So, you could add a method like the following in your custom view category: - getContainingBox { id sv; sv = [self superview]; while (sv && (![sv isKindOf:[Box class]])) sv = [sv superview]; return sv; } This will walk up the view hierarchy unitl it encounters a Box instance and return it [or nil]. BUT: 2) There is a lot of overhead associated with Views-- particularly when switchiing focus. Cells are much, MUCH more effecient for drawing... I would suggest a solution something like: Create an abstract cell subclass (of the appropriate Cell class-- there are several) that contains all functionality common to all your custom cell classes. Then create subclasses of your abstract cell that implements the unique characteristics of each of your 32 views; that is-- if the 32 views are all the same, you only need one cell subclass... but, if any of 'em are different, breaking out the commonalities in an abstract super with the custom pieces in concrete subclasses of the abstract super makes for extremely easy expansion/extension of the system. Create a custom view whose purpose is to: - maintain a collection of 32 cells (from above) - maintain the physical layout/relationships of the cells - forward all mouse and other events to the appropriate cell (event tracking) - (from -drawSelf::) ask the appropriate cell's to draw when necesary The big win here is the -drawSelf:: -- by doing some easy calculations to determine exactly which cells need to be redrawn (by determining which cell's frames intersect with the update rects passed to -drawSelf::), you can reduce drawing to just the cell's that are invalid. As well-- the only requirement when drawing a cell is that they don't paint outside their frame rectangle, you eliminate all -lockFocus/-unlockFocus's when drawing individual cells... it is all handled by the custom view [BTW: one common trap of NEXTSTEP display programming is using a display model that requires explicit calls to -lockFocus/-unlockFocus. The design of the AppKit's drawing system is such that you really shouldn't have to call these methods throughout the 'normal' execution of an application. Instead, use the -display methods and use invalidation rectangles... the Appkit's internals to support effecient drawing are actually quite good-- as long as you don't break the rules! Of course, sometimes the rules are unwritten and require a bit of creative experimentation to determine]. b.bum In <543u2m$8rd@news.duke.edu> Lee Altenberg wrote: # I am building an application that has 32 small custom views. Well, I made # them a bit too small, and want to enlarge them. I would like to be able # to do this programmatically, but have run into some problems. I can # enlarge the custom views, but each is surrounded by a box, and I don't # know how to access them. Each custom view is connected through Interface # Builder to an instance variable in my Controller class. But that doesn't # allow me to access the box that surrounds each one. I tried using # [aCustomView superview];, but that doesn't return the box. # # Do I need to create 32 new instance variables in Controller and connect # each box to one by hand? Should I figure out a programmatic way to # create the custom views and their boxes in the first place, and not # use Interface Builder to create them? Any tips will be greatly # appreciated. # ======================================================================= # Lee Altenberg, Ph.D. # # Research Affiliate, University of Hawai`i at Manoa # Office: Maui High Performance Computing Center # 550 Lipoa Parkway, Suite 100 # Kihei, Maui HI 96753 # Phone: (808) 879-5077 x 296 (work), (808) 879-5018 (fax) # E-mail: altenber@mhpcc.edu <NeXTMail and MIME: altenber@pueo.mhpcc.edu> # Web: ¬http://pueo.mhpcc.edu/~altenber/ # ======================================================================= #
From: sanjeev@ee.umr.edu (Sanjeev Agarwal) Newsgroups: comp.sys.next.programmer Subject: Help .... Date: 17 Oct 1996 20:57:37 GMT Organization: UMR Missouri's Technological University Message-ID: <5466k1$lnt@hptemp1.cc.umr.edu> Hi Everyone, I had to read/write to the Parallel Port (LPT1) on my Intel 486 machine running NextStep 3.3. Could you please tell me how do I do it. Thank you very much ... Sanjeev Agarwal sanjeev@isc.umr.edu
From: v22osprey@aol.com (V22Osprey) Newsgroups: comp.sys.next.programmer Subject: OCR Available Date: 17 Oct 1996 22:13:58 -0400 Organization: America Online, Inc. (1-800-827-6364) Sender: root@newsbf02.news.aol.com Message-ID: <546p56$t31@newsbf02.news.aol.com> Does anyone know of an OCR software for NeXT 3.2?
From: "Chris Trimble" <trimble@walrus.com> Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 18 Oct 1996 02:47:46 GMT Organization: I am the walrus Message-ID: <01bbbc9e$250b8ff0$761018ce@barrington> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit To all at NeXT: Most great applications don't start in the annals of consulting work or corporate development. As you know, they often start at colleges -- and I think marketing to academia is still a great idea. But they also start at home... some guy with a bright idea hacking away late at night. Most of the great Mac applications have started exactly this way. The only thing holding NeXT back from getting those kind of developers and homegrown killer apps is price. I just checked the price for the OpenSTEP/Mach developer license at Optimal Object -- it's $5K. The OS wars are NOT over; people still deserve better than NT 4.0. With a reasonably priced user license and by pricing its development tools like Delphi and Powerbuilder, with single user development being $300-$400, NeXT can break the home market wide open for its OS _and_ still gain in the enterprise development market. If NeXT brings home its development kit and OS, I think the range of apps that came out within a year would be frighteningly expansive. Right now, the only alternative OS is MacOS -- it will soon probably be BeOS; unless NeXT steps in, at-home developers and users will be building bridges to the 21st Century on top of second-class OSes. - Chris
From: rwakeman@thoughtport.com Newsgroups: comp.sys.next.programmer Subject: Installing OpenStep 4 on a 200 Mhz intel with NT4 Date: 18 Oct 1996 02:24:43 GMT Organization: The ThoughtPort Authority, Inc. Message-ID: <546ppb$1ho@chinx10.thoughtport.net> Has anybody installed OpenStep 4. on a 200mhz intel with NT4.0, via cd rom drive? From NextAnswers, it seems that it's easier in a pc with win95. If you have had success, please send me the step by step procedure. My machine can't read the cd disk. I'm one of those people who's used to black hardware just working. Thanks, Robert Wakeman rwakeman@thoughtport.com
From: anch@logiball.de (Andreas Christiani) Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 18 Oct 1996 08:58:45 GMT Organization: Customer of EUnet Germany; Info: info@Germany.EU.net Message-ID: <547gs6$131@news.Dortmund.Germany.EU.net> References: <01bbbc9e$250b8ff0$761018ce@barrington> Chris Timble wrote : > To all at NeXT: > > Most great applications don't start in the annals of consulting work or corporate development. As > you know, they often start at colleges -- and I think marketing to academia is still a great idea. But > they also start at home... some guy with a bright idea hacking away late at night. Most of the great > Mac applications have started exactly this way. The only thing holding NeXT back from getting those > kind of developers and homegrown killer apps is price. I just checked the price for the OpenSTEP/Mach > developer license at Optimal Object -- it's $5K. > > The OS wars are NOT over; people still deserve better than NT 4.0. With a reasonably priced user > license and by pricing its development tools like Delphi and Powerbuilder, with single user > development being $300-$400, NeXT can break the home market wide open for its OS _and_ still gain in > the enterprise development market. If NeXT brings home its development kit and OS, I think the range > of apps that came out within a year would be frighteningly expansive. Right now, the only alternative > OS is MacOS -- it will soon probably be BeOS; unless NeXT steps in, at-home developers and users will > be building bridges to the 21st Century on top of second-class OSes. > > - Chris > BIG APPLAUSE !!!!!!! A.C. --- ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ LogiBall gGmbH * Innovationszentrum Herne * Westring 303 * 44629 Herne Andreas Christiani * christiani@logiball.de * http://www.logiball.de Tel.: 02323 / 925 577 * Fax : 02323 / 925 551 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
From: headi@now.ch (Daniel Scheidegger) Newsgroups: comp.sys.next.programmer Subject: Submenu from a NSMenuItem? Date: Thu, 17 Oct 1996 18:45:57 GMT Organization: NOW GmbH, Zuerich, Switzerland Message-ID: <DzFo4L.Eyv.0.astra@now.ch> Hi all, how can i get the submenu from a NSMenuItem? I can call [menuItem hasSubmenu] to find out that it has one, but I don't know how to get the actual submenu. Any hints? Thanks a lot, Daniel -- Daniel Scheidegger Software Engineer, System Administrator NOW GmbH, Scheideggstr. 73, CH-8038 Zuerich ++41-1-2898025 / dscheide@now.ch
Newsgroups: comp.sys.next.programmer Organization: Antigone Press gateway, San Francisco Return-Path: <cs@hal.kph.tuwien.ac.at> Message-ID: <9610172101.AA03550@zaphod> Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 4.0 v146.2) From: Christian Starkjohann <cs@hal.kph.tuwien.ac.at> Date: Thu, 17 Oct 96 23:01:05 +0200 Subject: Re: Developing 3.x applications on OS4.0 Following up my own posting: > In article <1996Sep17.120417.85069@cc.usu.edu> edx@cc.usu.edu writes: > > [quoted stuff deleted] > > This script does -NOT- work with NS Dev 3.3 for the same reason > > you can't install 3.3 Developer directly on a 4.0 system. Notice > > that the real work is done in the "for i in $FILES" loop where the > > 4.0 working files are moved to a directory called /Dev-4.0. Now > > as soon as /usr/lib/dyld is moved, your machine is hosed. Half > > the binaries, especially things like mv, cp, rm, etc suddenly don't > > work anymore. > > > > I would highly advise not using this script on with a 3.3 Developer > > unless this is changed. > > I have never bought 3.3, so I can't comment on this. But: as I > understand, the /usr/lib/dyld in 4.0 is compatible with 3.3 > executables. The easiest workaround would be to delete > /Dev-3.3/usr/lib/dyld before you run the script. This would leave > the /usr/lib/dyld of 4.0 for both developer packages. > > I would like to hear about the results if anyone dares to try this. I received the following by e-mail from George Tourlakis <george@csm03.atkinson.yorku.ca>: > You are right, it works (at least, it works for me). I kept the > 4.0 dyld but I also kept all the 4.0 execuatbles in /usr/lib that > are common in both 3.3 and 4.0 (such as "fastcp") and did _not_ > involve the /usr/ucb at all in the linking. > > By "works" I mean this (_limited testing_). Under 3.3 I successfully > compiled from Examples (3.3) Backspace.app, Adaptec 154xB, ATI. > I also checked that the Mux1.7 compiles (BTW, OS 4.0 apparently > does not like Mux). Switching to 4.0 I also ensured that several > projects compile correctly. Moreover, no crashes (so far) and I > can boot under either 3.3 or 4.0 as "current" version of Developer > without any problem. Hope this helps. /usr/ucb only contains gprof and George found out that the 4.0 gprof works with 3.3 gmon.out. The executables in /usr/lib should do the same in 3.2, 3.3 and 4.0, so it might be a good idea to keep the 4.0 versions. If you have missed the posting containing the script and want a copy, mail me. If there is interest I can place the scripts on an ftp server. Bye, Christian. -- Christian Starkjohann <cs@hal.kph.tuwien.ac.at> or <cs@ds1.kph.tuwien.ac.at>, finger for PGP Public Key. PGP fingerprint: DF FD 40 60 91 6A 14 1C CD 2C E9 07 38 AE CB 4E
From: "Georg Tuparev" <gtupar@ctp.com> Newsgroups: comp.sys.next.programmer,comp.sys.next.hardware Subject: Need help for /etc/disktab Date: 18 Oct 1996 10:09:23 GMT Organization: Cambridge Technology Partners, Inc. Message-ID: <547l0j$1dj@concorde.ctp.com> Hi NeXTers, I have an old 2GB Fujitsu disk (Model: M2654SA) but without disk info block (or what the hell the stuff is called). Could somebody send me either the disktab entry or the disk information. Thanks alot ------- /\/\ Georg Tuparev <georg_tuparev@ctp.com> / /_ \ Cambridge Technology Partners \ / / Apollo House, Apollolaan 15 \/\/ 1077 AB Amsterdam, The Netherlands Tel: +31(20)575-0492 Fax: +31(20)575-0500
From: devan2m@imap2.asu.edu Newsgroups: comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Java enabled browser for Next Date: 18 Oct 1996 10:48:46 GMT Organization: Arizona State University Distribution: world Message-ID: <547nae$96o@news.asu.edu> References: <325DF79F.41C67EA6@diogenes.huge.ch> <Pine.HPP.3.95.961011100417.23844C-100000@hphalle0.informatik.tu-muenchen.de> <53pbr6$icg@alice.walrus.com> Hello David, Would you care to provide specifics about what NeXT, someone, or some entity independent of NeXT ( a consortium of advocates) acting on behalf of the NEXTSTEP community's and the technology might do from a marketing standpoint to change this course? This would all be based on the premise that despite NEXTSTEP 3.3 not having garnered the market share and following everyone who favors it would like or having a zillion developers and software packages, it still has the best user and developer environments on the planet. To make it interesting or should I say challenging, you would only have $10 M USD to work with and would need to deliver a principle and return of $15 M on this investment in the NeXT community within 3 years. Furthermore, you have only six months in which to engineer and produce measurable results of an effective longer-term master plan. A good barometer of just how effective and well your master marketing plan would be in the requisite time frame would be if there were 1-2 million new U.S. based NEXTSTEP end-users were to purchase the software in this period. This is all hypothetically speaking of course for the moment. I would love to hear what type of course, time frame, monetary and human resources you would say it would take to make a significant positive impact. Or do you think the war is lost and there is nothing that can be done barring MS falling over dead or incorporating some major catastrophic bug/virus in Win '95 which wipes out every hard drive and system it is loaded on? Here's a chance for all those who fancy themselves "armchair assassins" and "marketing masterminds" to stage what they believe to be a viable campaign and "coup de grace" which would put the NeXT world and OS environment on a track to evolve as it could have, should have, and still can and not as some Windows or Sun chameleon even though the Trojan Horse approach and idea is not a bad one if only we didn't have to go through the Windoze digression. My belief it that there would be nothing stronger than about 1-5 million new end users of NEXTSTEP 3.3 or so to compel NeXT to do some of the things that many in the community would like to see done. Is this a legitimate marketing challenge and pursuit or simply insane and inane exercise? Regards- David Young (dwy@mcny.com) wrote: : Bernhard Scholz (scholz@informatik.tu-muenchen.de) wrote: : : No. :( : : This lack of technology will probably stop the NeXTworld one day... : : Greetings, : I don't think jack of Java will stop the NeXTworld, as it were. : NeXT Marketing will stop the NeXTworld. : -- : # david young: network engineer+oop developer : # net: dwy@ace.net (NeXTmail ok) web: http://www.ace.net/ : # vox: 212.686.3845 201.798.5217 fax: 212.686.3845
From: flight@mathi.uni-heidelberg.de (Gregor Hoffleit) Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Followup-To: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Date: 18 Oct 1996 10:46:16 GMT Organization: University of Heidelberg, Germany Message-ID: <547n5o$310@sun0.urz.uni-heidelberg.de> References: <01bbbc9e$250b8ff0$761018ce@barrington> Chris Trimble (trimble@walrus.com) wrote: : The OS wars are NOT over; people still deserve better than NT 4.0. With : a reasonably priced user license and by pricing its development tools like : Delphi and Powerbuilder, with single user development being $300-$400, : NeXT can break the home market wide open for its OS _and_ still gain in : the enterprise development market. If NeXT brings home its development : kit and OS, I think the range of apps that came out within a year would be : frighteningly expansive. Right now, the only alternative OS is MacOS -- : it will soon probably be BeOS; unless NeXT steps in, at-home developers : and users will be building bridges to the 21st Century on top of : second-class OSes. Wrong. The first question is if the OS war is really of any interest any more, given the fact that quite a few `home-brew' projects are really cross-OS nowadays, and do not really depend on very specific OS features. The fight is at least one level higher. Then, you didn't mention Unix at all. Correct, all the commercial Unices (at least the PC-based) seem to depart quietly, but there are still (at least in the SOHO area with S=scientific) alternative platforms: FreeBSD, NetBSD, the Hurd, Linux to mention just a few. IMHO, NeXT (and probably most other companies, too), just has no chance to compete with those massive development teams on the OS level, as they have no choice to compete with MS in terms of support on such a diverse platform as Intel-based PCs. Therefore, IMHO it's a strong move by NeXT to stay out of the OS wars (maybe quietly keeping on development of new OS technologies in small teams... Mecca ?), wait for something to settle down, and then, build upon this. Not very avantgardistic, but you can make a life of it. IMHO, their only current advantage is their working OO knowledge. If they succeed in pushing and pulling all those prominent OO streams in the direction of their Object model, they have an market advantage that may help them to install some NeXTish ideas through the backdoor. Anyway, you'll certainly loose the NeXTstep GUI advantage, but then who hasn't noticed the MS is strongly moving towards what NS had a few years ago. So - try to see it with the eyes of NeXT Software Inc, the company - they don't have an OS that can compete with say Linux in terms of hw support or with Win95 in terms of marketing - why then bother with those issues when you can make your money with Fortune500 companies and WO, where you don't have to care about a broad and yet stable hardware support. It has been stated quite a few times: NeXT has no interest in the mass market. The only interesting scenario I could think of (still wondered if this is really so absurd as it sounds) would be if NeXT supported a free implementation of their OpenStep specification on top of -say again- Linux. Telling from the response to the GNUstep project, there's a high degree of interest and respect for NEXTSTEP in this community, and a free -or modestly-priced- implementation might help NeXT to establish the NeXT Object model as a quasi-standard among OO projects - which I hope is their interest. This is the only way I see for NeXT to get into the mass market in a way. Provided this won't happen, we can only hope for GNUstep ;-) and buy Win95. Gregor -- | Gregor Hoffleit Mathematisches Institut, Uni HD | | flight@mathi.uni-heidelberg.de INF 288, 69120 Heidelberg, Germany | | (NeXTmail, MIME) (49)6221 54-5771 fax 54-8312 | | PGP Key fingerprint = 23 8F B3 38 A3 39 A6 01 5B 99 91 D6 F2 AC CD C7 |
From: devan2m@imap2.asu.edu Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Followup-To: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Date: 18 Oct 1996 11:49:25 GMT Organization: Arizona State University Message-ID: <547qs5$96o@news.asu.edu> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547gs6$131@news.Dortmund.Germany.EU.net> Hello David, Would you care to provide specifics about what NeXT, someone, or some entity independent of NeXT ( a consortium of advocates) acting on behalf of the NEXTSTEP community's and the technology might do from a marketing standpoint to change this course? This would all be based on the premise that despite NEXTSTEP 3.3* not having garnered the market share and following everyone who favors it would like or having a zillion developers and software packages, it still has the best user and developer environments on the planet. To make it interesting or should I say challenging, you would only have $10 M USD to work with and would need to deliver a principle and return of $15 M on this investment in the NeXT community within 3 years. Furthermore, you have only six months in which to engineer and produce measurable results of an effective longer-term master plan. A good barometer of just how effective and well your master marketing plan would be in the requisite time frame would be if there were 1-2 million new U.S. based NEXTSTEP* end-users were to purchase the software in this period. This is all hypothetically speaking of course for the moment. I would love to hear what type of course, time frame, monetary and human resources you would say it would take to make a significant positive impact. Or do you think the war is lost and there is nothing that can be done barring MS falling over dead or incorporating some major catastrophic bug/virus in Win '95 which wipes out every hard drive and system it is loaded on? Here's a chance for all those who fancy themselves "armchair assassins" and "marketing masterminds" to stage what they believe to be a viable campaign and "coup de grace" which would put the NeXT world and OS environment on a track to evolve as it could have, should have, and still can and not as some Windows or Sun chameleon even though the Trojan Horse approach and idea is not a bad one if only we didn't have to go through the Windoze digression. My belief it that there would be nothing stronger than about 1-5 million new end users of NEXTSTEP 3.3* or so to compel NeXT to do some of the things that many in the community would like to see done. Is this a legitimate marketing challenge and pursuit or simply insane and inane exercise? * Or OPENSTEP 4.x for MACH as long as it is an environment which gets back on track with the evolution as it should be and not as some Windoze/Solaris looking chameleon or derivative. At least this is my preference and what I would to continue using. Mecca anyone? Kind Regards- Andreas Christiani (anch@logiball.de) wrote: : Chris Timble wrote : : > To all at NeXT: : > : > Most great applications don't start in the annals of consulting work or corporate development. As : > you know, they often start at colleges -- and I think marketing to academia is still a great idea. But : > they also start at home... some guy with a bright idea hacking away late at night. Most of the great : > Mac applications have started exactly this way. The only thing holding NeXT back from getting those : > kind of developers and homegrown killer apps is price. I just checked the price for the OpenSTEP/Mach : > developer license at Optimal Object -- it's $5K. : > : > The OS wars are NOT over; people still deserve better than NT 4.0. With a reasonably priced user : > license and by pricing its development tools like Delphi and Powerbuilder, with single user : > development being $300-$400, NeXT can break the home market wide open for its OS _and_ still gain in : > the enterprise development market. If NeXT brings home its development kit and OS, I think the range : > of apps that came out within a year would be frighteningly expansive. Right now, the only alternative : > OS is MacOS -- it will soon probably be BeOS; unless NeXT steps in, at-home developers and users will : > be building bridges to the 21st Century on top of second-class OSes. : > : > - Chris : > : BIG APPLAUSE !!!!!!! : A.C. : --- : ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ : LogiBall gGmbH * Innovationszentrum Herne * Westring 303 * 44629 Herne : Andreas Christiani * christiani@logiball.de * http://www.logiball.de : Tel.: 02323 / 925 577 * Fax : 02323 / 925 551 : ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
From: sk68@cornell.edu (Sung Ho Kim) Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 18 Oct 1996 12:20:21 GMT Organization: Cornell University Sender: sk68@cornell.edu (Verified) Message-ID: <sk68-0101040021590001@cu-dialup-0017.cit.cornell.edu> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547n5o$310@sun0.urz.uni-heidelberg.de> Hi I can't be but a little disappointed by Mr.Hoffleit's comments regarding his views about NeXT and where it should go. First and foremost, I am not a programmer or a fanatic; I'm just a normal user. But what caught my fancy about NeXT and its products was what it stood for in this industry of also-rans. (I consider Linux & BeOS to be in the also-ran department) It took a different approach to computing that will hopefully be rewarded in the annals of computing history. In a market full or market-share mongrels, it tried to produce something that was no-compromise, first-rate and innovative. I don't think anybody can doubt that they did in fact produce something special. I think this is what makes NeXT special. If NeXT had started off with the kind of philosophy Mr. Hoffleit thinks NeXT should take now, I think NeXT as we know it now (albeit its small niche position) would not exist. If in fact NeXT thought in such calculated and market oriented way in the beginning, it would have ended up in the also-ran department of computer history. But in fact it didn't. People can laugh all they want about the market position NeXT is in right now, and they can have their theories about doing this and that with NeXT, but in the end it is a fact that many of the NeXT users have fallen in love with this innovative side of NeXT and not the market oriented side of NeXT. And as Byte magazine once said of NeXT, "NeXTSTEP is probably the most respected software in the planet." Yep. Well, I hope I didn't make too many grammar mistakes. --------------------------- Sung Ho Kim Cornell University sk68@cornell.edu --------------------------- In article <547n5o$310@sun0.urz.uni-heidelberg.de>, flight@mathi.uni-heidelberg.de (Gregor Hoffleit) wrote: > Chris Trimble (trimble@walrus.com) wrote: > > : The OS wars are NOT over; people still deserve better than NT 4.0. With > : a reasonably priced user license and by pricing its development tools like > : Delphi and Powerbuilder, with single user development being $300-$400, > : NeXT can break the home market wide open for its OS _and_ still gain in > : the enterprise development market. If NeXT brings home its development > : kit and OS, I think the range of apps that came out within a year would be > : frighteningly expansive. Right now, the only alternative OS is MacOS -- > : it will soon probably be BeOS; unless NeXT steps in, at-home developers > : and users will be building bridges to the 21st Century on top of > : second-class OSes. > > Wrong. The first question is if the OS war is really of any interest > any more, given the fact that quite a few `home-brew' projects are > really cross-OS nowadays, and do not really depend on very specific OS > features. The fight is at least one level higher. > > Then, you didn't mention Unix at all. Correct, all the commercial > Unices (at least the PC-based) seem to depart quietly, but there are > still (at least in the SOHO area with S=scientific) alternative > platforms: FreeBSD, NetBSD, the Hurd, Linux to mention just a > few. IMHO, NeXT (and probably most other companies, too), just has no > chance to compete with those massive development teams on the OS > level, as they have no choice to compete with MS in terms of support > on such a diverse platform as Intel-based PCs. > > Therefore, IMHO it's a strong move by NeXT to stay out of the OS wars > (maybe quietly keeping on development of new OS technologies in small > teams... Mecca ?), wait for something to settle down, and then, build > upon this. Not very avantgardistic, but you can make a life of it. > > IMHO, their only current advantage is their working OO knowledge. If > they succeed in pushing and pulling all those prominent OO streams in > the direction of their Object model, they have an market advantage > that may help them to install some NeXTish ideas through the backdoor. > > Anyway, you'll certainly loose the NeXTstep GUI advantage, but then > who hasn't noticed the MS is strongly moving towards what NS had a few > years ago. > > So - try to see it with the eyes of NeXT Software Inc, the company - > they don't have an OS that can compete with say Linux in terms of hw > support or with Win95 in terms of marketing - why then bother with > those issues when you can make your money with Fortune500 companies > and WO, where you don't have to care about a broad and yet stable > hardware support. > > It has been stated quite a few times: NeXT has no interest in the mass > market. The only interesting scenario I could think of (still wondered > if this is really so absurd as it sounds) would be if NeXT supported a > free implementation of their OpenStep specification on top of -say > again- Linux. Telling from the response to the GNUstep project, > there's a high degree of interest and respect for NEXTSTEP in this > community, and a free -or modestly-priced- implementation might help > NeXT to establish the NeXT Object model as a quasi-standard among OO > projects - which I hope is their interest. This is the only way I see > for NeXT to get into the mass market in a way. > > > Provided this won't happen, we can only hope for GNUstep ;-) and buy > Win95. > > Gregor > > > -- > | Gregor Hoffleit Mathematisches Institut, Uni HD | > | flight@mathi.uni-heidelberg.de INF 288, 69120 Heidelberg, Germany | > | (NeXTmail, MIME) (49)6221 54-5771 fax 54-8312 | > | PGP Key fingerprint = 23 8F B3 38 A3 39 A6 01 5B 99 91 D6 F2 AC CD C7 |
From: "Georg Tuparev" <gtupar@ctp.com> Newsgroups: comp.sys.next.programmer Subject: Re: mail and files Date: 18 Oct 1996 13:15:24 GMT Organization: Cambridge Technology Partners, Inc. Message-ID: <547vtc$24b@concorde.ctp.com> References: <54272f$628@hpuniv.univ-lr.fr> In article <54272f$628@hpuniv.univ-lr.fr> yannick@uranus.univ-lr.fr (Yannick Buisson) writes: > I want to use an API for sending mail. > I want to automaticaly include files in the mail. > Is it Possible ? how ? > Are there some examples ? Use MiscKit! There are two classes (MiscMailXXX) you should check. BTW: Where is the description of the new API for Mail.app (NS 4.0)? The EnhanceMail.bundle does not work with 4.0 (but Urlifier.bundle dues !?!?) Have fun ------- /\/\ Georg Tuparev <georg_tuparev@ctp.com> / /_ \ Cambridge Technology Partners \ / / Apollo House, Apollolaan 15 \/\/ 1077 AB Amsterdam, The Netherlands Tel: +31(20)575-0492 Fax: +31(20)575-0500
From: Bernhard Scholz <scholz@informatik.tu-muenchen.de> Newsgroups: comp.sys.next.programmer Subject: Re: mail and files Date: Fri, 18 Oct 1996 15:54:25 +0200 Organization: Technische Universitaet Muenchen, Germany Distribution: world Message-ID: <Pine.HPP.3.95.961018155027.10931A-100000@hphalle0.informatik.tu-muenchen.de> References: <54272f$628@hpuniv.univ-lr.fr> <547vtc$24b@concorde.ctp.com> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII In-Reply-To: <547vtc$24b@concorde.ctp.com> On 18 Oct 1996, Georg Tuparev wrote: > In article <54272f$628@hpuniv.univ-lr.fr> yannick@uranus.univ-lr.fr > > BTW: Where is the description of the new API for Mail.app (NS 4.0)? The > EnhanceMail.bundle does not work with 4.0 (but Urlifier.bundle dues !?!?) > There is a special version of EnhanceMail which works fine with Mail of OS4.0. Check out: ftp://peanuts.leo.org/next/OpenStep/bundles/EnhanceMail* or any mirror: ftp.evolution.com, ftp.fdn.fr,ftp.eunet.ch Greetings, Bernhard.
From: ehutch@hypnos.norden1.com (E. Hutchinson) Newsgroups: comp.sys.next.programmer,misc.jobs.offered,misc.jobs.contract Subject: NEXTSTEP/Contract/Va Date: 18 Oct 1996 13:52:01 GMT Organization: Norden 1 Communications Message-ID: <548221$nta@tofu.alt.net> Programmer/analyst NEXTSTEP----------Commercial experience Objective C-------Commercial experience Contract----------Long term Area--------------Virginia Start Date--------Nov 1996 Must Be-----------US Citizen or Greencard To Be Considered--Fax resume or mail a hard copy. -- ehutch@norden1.com (419) 893-6367 [fax] Omni Search (419) 893-6334 [voice] 1310 Craig Maumee, Ohio 43537
From: jmichel@imtn.dsccc.com (James Michel) Newsgroups: comp.sys.next.programmer Subject: NEWBIE: Polymorphism? Date: 18 Oct 1996 15:11:11 GMT Organization: DSC Communications Corporation, Plano, Texas USA Message-ID: <5486mf$epk@camelot.dsccc.com> I am currently reading an Objective-C book entitled "Objective-C Object Oriented Programming Techniques" by Lewis J. Pinson and Richard S. Weiner. In the section covering the mechanics of messaging I ran across this "All implementations of identically named methods must have the same return type and the same argument types since the messaging function has access to the method implementation only through selectors, and thus treats all methods with the same selector alike." Now this sounds like their is no polymorphism in Objective-C OR I need to take this statement in the context of the mechanics of messaging discussion. In this section(mechanics of messaging) the authors speak about using a C function to dynamically create the proper message statement at runtime. Could it be that if you need this flexibilty in your application then methods must adhere to the rules specified in the above quotation otherwise methods may be overloaded? Also, the example in the book concerns two different classes having the same method name with different argument types. This causes a compile time error. Thanks, James
From: jrudd@cygnus.com (John Rudd) Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 18 Oct 1996 17:06:58 GMT Organization: Cygnus Support Message-ID: <548dfi$73d@majipoor.cygnus.com> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547n5o$310@sun0.urz.uni-heidelberg.de> <sk68-0101040021590001@cu-dialup-0017.cit.cornell.edu> Well, I don't know if I should be posting this or not.. In some ways it's good news for the Gnustep people, and in some ways it's bad news for the Nextstep and Openstep people. At Cygnus we had a new GDB developer start this week. He came to us from NeXT. One of the things he said he hopes to do is bring some of the features of NeXT's GDB to other GDB platforms. I asked me why he didn't stay at NeXT, and his response sort of confirmed my worst fears.. "They say they're going to become a 'Web company'. They don't know what that means yet, but that's what they're going to do." Losing GDB people not only spells bad news for the Nextstep/OS-M community, but also for the general Openstep from NeXT community. It looks like they're not just moving away from the OS market, but away from the developers tools market as well. And worst of all, they don't even know what they're moving toward. I guess we'll just have to see what Afterstep and Gnustep will bring us.. This week I've started taking a closer look at netbsd, BSD-lites, and MkLinux. I somehow get the feeling that's where my future lies. It would be nice if we could get some of those large commercial investors in Nextstep to put up money to get NeXT to release the source to their OS, or something.. But I don't see that happening. -- John "kzin" Rudd jrudd@cygnus.com =========Intel: Putting the backward in backward compatible.============ "And, ironically, that's how the founding fathers expected it to work: either the government stays clean, or the people shoot them." -- ttk
From: dekorte@suite.com (Steve Dekorte) Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 18 Oct 1996 18:18:45 GMT Organization: Suite Software Message-ID: <548hm5$s42@news.onramp.net> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547gs6$131@news.Dortmund.Germany.EU.net> <547qs5$96o@news.asu.edu> Cc: devan2m@imap2.asu.edu devan2m@imap2.asu.edu wrote: > Would you care to provide specifics about what NeXT, ... might do from a > marketing standpoint to change this course? ... > To make it interesting or should I say challenging, you would > only have $10 M USD to work with and would need to deliver a principle > and return of $15 M on this investment in the NeXT community within 3 > years... Here's my shot: Give NeXTstep/OpenStep academic away. NeXT can't be making even $1M from this market, yet doing this would do far more than 10 times as much spent on marketing. How's that for a return on investment? -- Steve Dekorte - OpenStep Developer - Anaheim, CA "Fundamentalism isn't about religion. It's about power." - S. Rushdie
From: Mark_Bessey@next.com (Mark Bessey) Newsgroups: comp.sys.next.programmer Subject: Re: NEWBIE: Polymorphism? Date: 18 Oct 1996 18:43:46 GMT Organization: NeXT Software, Inc. Message-ID: <548j52$7ha@news.next.com> References: <5486mf$epk@camelot.dsccc.com> James Michel writes > I am currently reading an Objective-C book entitled "Objective-C > Object Oriented Programming Techniques" by Lewis J. Pinson and Richard > S. Weiner. In the section covering the mechanics of messaging I ran > across this > > "All implementations of identically named methods must have the > same return type and the same argument types since the messaging > function has access to the method implementation only through > selectors, and thus treats all methods with the same selector > alike." > Actually, the messaging function treats *ALL* messages exactly the same. Well, at least NeXT's implementation does... > Now this sounds like their is no polymorphism in Objective-C OR I need > to take this statement in the context of the mechanics of messaging > discussion. In this section(mechanics of messaging) the authors speak > about using a C function to dynamically create the proper message > statement at runtime. > Could it be that if you need this flexibilty in your application then > methods must adhere to the rules specified in the above quotation > otherwise methods may be overloaded? You can have methods with the same name that return different types, you just need to be careful about how you use them - the type of the object being messaged must be known at compile time. > Also, the example in the book concerns two different classes having > the same method name with different argument types. This causes a > compile time error. Actually, it's a compile-time WARNING, right? You can in fact push on and get an executable. What it'll do when you run it is difficult to predict :-); > Thanks, > > James The problem is that given two classes that implement the same method with different return types: @implementation foo: Object -(float)bug; @end @implementation bar: Object -(int)bug; @end and a method invocation like this: id a; float b; a = [someObject aMessage]; b = [a bug]; Which version of the "bug" method is being called? The compiler needs to know this so that it can find the return value (floats and ints are returned by different mechanisms). If you use a declaration like: foo *a; or cast the message reciever: b = [(foo *)a bug]; Then you won't get any warnings, and the program will even work correctly. There's a pretty good explanation of all this in the Objective-C documentation that NeXT provides. See the section "Objective C Extensions" in the book "Object-Oriented Programming and the Objective C Language", available on-line with OPENSTEP Developer, or on NeXT's Web Page. Hope this helps, -Mark -- Mark Bessey NeXT Software, Inc Software Quality Assurance -->I DON'T SPEAK FOR NeXT <--
Newsgroups: comp.sys.next.programmer Subject: Re: Reuse library sought: Hierarchical Outlining Message-ID: <3267E5CE.61E8@running-start.com> From: Ralph Zazula <zazula@running-start.com> Date: Fri, 18 Oct 1996 13:17:18 -0700 References: <9610141416.AA18933@chmail> Organization: Running Start, Inc. MIME-Version: 1.0 To: Garrick McFarlane <gmcfarla@LandP.com> CC: zazula@running-start.com Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Garrick McFarlane wrote: > > Hello all > > As part of an in-house OpenStep development project, we need to > build a hierarchical, fully manipulable outline view, such as is > found in Concurrence by Lighthouse. (A less featured example is the > Bookmarks window in OmniWeb, or perhaps the Outlets/Actions list in > Interface Builder.) > Hi - We have developed a framework for displaying data in a hierarchical fashion for AppKit and WebObjects based applications which we are licensing to developers. Both implementations use our RSOutline back-end and have different front-ends. We are currently adding JavaScript support to the WebObjects version which makes the outline much speedier since you avoid a round-trip to the server for an expand/collapse. You can see an example of this on our web-site(*) at the following URL: http://www.running-start.com/cgi-bin/submitBug.cgi Just enter your email address and press "Continue", you'll be shown an outline of topics to submit requests against generated by our RSWebOutline component. * - apologies in advance for the slowness of our connection, USWest promises our T1 will be in by the end of the month... Ralph --- Ralph Zazula Running Start, Inc. 520/760-4890 zazula@running-start.com http://www.running-start.com
From: dekorte@suite.com (Steve Dekorte) Newsgroups: comp.sys.next.programmer Subject: NSTableView Formatting Date: 18 Oct 1996 23:56:26 GMT Organization: Suite Software Message-ID: <5495fa$5po@news.onramp.net> Any examples out there? Steve Dekorte
From: khader@katie.vnet.net (R. D. Khader) Newsgroups: comp.sys.next.programmer Subject: Is there an Error Object and Handling? Date: 19 Oct 1996 15:40:45 GMT Organization: Vnet Internet Access, Inc. - Charlotte, NC. (704) 374-0779 Message-ID: <54aspt$lvb@ralph.vnet.net> Hi there, Does any one know about an error object and handling, or even a protocol. If available in C++ it is even better. thanks in advance. khader@vnet.net
From: "Chris Trimble" <trimble@walrus.com> Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 19 Oct 1996 15:56:04 GMT Organization: I am the walrus Message-ID: <01bbbdd5$69724200$5b1018ce@barrington> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547n5o$310@sun0.urz.uni-heidelberg.de> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Gregor Hoffleit <flight@mathi.uni-heidelberg.de> wrote: : Wrong. The first question is if the OS war is really of : any interest any more, given the fact that quite a few : `home-brew' projects are really cross-OS nowadays, : and do not really depend on very specific OS : features. The fight is at least one level higher. Actually, I'd say the fight has moved one level LOWER. Those cross-platform projects you're talking about demonstrate very little platform consistency and typically have weak integration with any one OS. The OS wars are still interesting, and the reason why is because Java or GNU-type portable apps can't not make a dent in the general application development market for some time to come... probably never. They simply do not offer the consistency and features that the users in the world demand. : Then, you didn't mention Unix at all. That's right, I didn't. Wireheads like you and me can deal with UNIX fine, and enjoy it as users because we are technical users. Most people on the planet cannot or don't want to deal with installing and maintaining Linux. One of the main reasons is BECAUSE of its distributed development effort... things get updated so often, that to have the most recent OS means you'll be downloading, compiling and installing megabytes per week. Red Hat and others have made this considerably better, but the problem is still there. Whenever I think about OS things, I think about what a soccer mom would want to use. I cannot imagine a soccer mom trying to link in the proper version of X to use her S3 chipset -- that will not happen in our lifetime, no matter what the Linux pundits think. : IMHO, NeXT (and probably most other companies, too), just has no : chance to compete with those massive development teams on the OS : level, as they have no choice to compete with MS in terms of support : on such a diverse platform as Intel-based PCs. NeXT can get a lot of help from the developer community, as Linux does. Linux is supported almost completely by the developer community, the exceptions being MetroX, Redhat, etc., which is its strength AND a weakpoint. Your typical home user won't know where, when, or how to get updates. But the difference between what's going on on Linux would be that NeXT could be the root of all distribution, making it easier for the soccer moms to get their updates as well. As far as competing with MS goes: yes and no. Most users in the world will probably stick with Windows because they don't know better. If they made these kinds of moves and marketed it well, NeXT would get the mid to high-level users and all sorts of developers in the near future, with long-term potential for moving in on Windows. When the time came around for NeXT to have to start dealing with millions of calls from soccer moms, as MS does, they would by then have the resources to afford it. : Therefore, IMHO it's a strong move by NeXT to stay out of : the OS wars, wait for something to settle down, and then, build : upon this. Not very avantgardistic, but you can make a life of it. I still don't really see any reason for it. In the very short term, I can see NeXT trying to boost revenue as much as they can for the IPO. That explains their doubling the price of NS developer when it became OpenStep. That stuff makes sense, even though it's somewhat, uh, client-unfriendly. After that, however, I can't see any reason not to jump in. WebObjects will only serve NeXT well until someone like MS or Netscape comes out with tools that do the same thing cheaper. I'm sure those two have seen how much NeXT is making off of WebObjects, even in limited sales at exorbitant prices. NeXT's object-oriented design of these things is great and all, but it really doesn't matter much to the bean counters. When the all of the servers are running IIS or SuiteSpot and one of those companies offers a sweet deal for less programmer-friendly tools, WebObjects will be much harder to sell. : So - try to see it with the eyes of NeXT Software Inc, : the company - they don't have an OS that can compete : with say Linux in terms of hw support or with Win95 in : terms of marketing - why then bother with those issues : when you can make your money with Fortune500 companies : and WO, where you don't have to care about a broad and yet stable : hardware support. Because the WO advantage will not last forever. NeXT's OO offers them a huge advantage in a programmer's eyes, but WO's price is, and I finally heard a real figure yesterday, ludicrously high. I'm a programmer, and I can't even fathom paying that much for something like that -- it's cheaper to hire five or six programmers for half a year than to pay the WO startup costs. When Netscape and MS start offering WO-like tools, and they will, the WO advantage will drop considerably in the eyes of all but the programmers (and maybe even them, depending on how well NS and MS do it). As far as the OS goes then: if NeXT is that down on their OS, why don't they sell it? I simply can't believe they're going to sit there and let their OS rot, but from the general response to this message, that seems to be the idea. : The only interesting scenario I could think of (still wondered : if this is really so absurd as it sounds) would be if NeXT : supported a free implementation of their OpenStep : specification on top of -say again- Linux. That certainly would be an interesting scenario. If NeXT then packaged and distributed such a thing, it would be very profitable. Of course, it seems rather silly when they have their own implementation right now. : Telling from the response to the GNUstep project, : there's a high degree of interest and respect for NEXTSTEP : in this community, and a free -or modestly-priced- : implementation might help NeXT to establish the NeXT : Object model as a quasi-standard among OO projects - : which I hope is their interest. This is the only way I see : for NeXT to get into the mass market in a way. You're right, there is a lot of interest; and NeXT is a company that can push out a modestly priced implementation right now -- which is my point. There's a strong grass-roots movement going on right now... I don't quite understand why they should wait until that movement gains a foothold to realize that it's a good market. GNUStep has already shown it will be popular and is within our grasp; why NeXT doesn't capitalize on it boggles my mind and is why I posted the original message. : Provided this won't happen, we can only hope for GNUstep ;-) and buy : Win95. Don't just hope... make it happen. Life's too short to wait. - Chris Setting up GNUstep stuff on his machine as he writes
From: colinj@math.math.unm.edu (Colin Eric Johnson) Newsgroups: comp.sys.next.programmer,comp.sys.next.misc Subject: cc++ troubles with iostream.h Date: 19 Oct 1996 18:48:45 GMT Organization: University of New Mexico, Albuquerque Distribution: world Message-ID: <54b7qd$keh@lynx.unm.edu> I've installed NS 4.0 (OS) User/Developer and I'm trying to get my homework done, no I don't want you to do it, for my C++ course. Below is an attempt to make the last project we had due, it worked on an AIX (4.x) system with xlC. The linker complains when it tries to put the final binary together about istream and ostream not being defined symbols. iostream.h has been included and I'm not clear why this is failing. Is there a known problem with C++ and the installed compilers on NS 4.0? Am I doing something outside of proper C++ that xlC will let me get away with? I'm not feeling like I'm being too clear about this problem, if anyone has any insight please let me know. As I said I can get this to compile w/o complaint on an RS6K running AIX/xlC but I'd like to be able to get some homework done at home. A failed attempt at a make follows: thebrain> ls Makefile point.h renderer.C renderer.o rentest.o str.h point.C point.o renderer.h rentest.C str.C str.o thebrain> make clean rm *.o *~ rentest rm: *~ nonexistent rm: rentest nonexistent *** Exit 1 Stop. thebrain> make cc++ -g -c str.C cc++ -g -c point.C cc++ -g -c renderer.C point.C str.o point.o cc: str.o: linker input file unused since linking not done cc: point.o: linker input file unused since linking not done cc++ -g -c rentest.C renderer.o cc: renderer.o: linker input file unused since linking not done cc++ -g rentest.o renderer.o str.o point.o -o rentest ld: Undefined symbols: ostream::operator<<(ostream &(*)(ostream &)) ostream::operator<<(char) cout endl(ostream &) ostream::operator<<(char const *) istream::get(char *, int, char) thebrain> -- "Now my life is better than an ABBA song" - Muriel, "Muriel's Wedding" Colin E. Johnson | colinj@unm.edu | http://www.unm.edu/~colinj/ NeXTMail, MIMEmail, Textmail, send it all, I'm easy.
From: "Eric A. Dubiel" <eadubie@rs6000.cmp.ilstu.edu> Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: Sat, 19 Oct 1996 18:06:42 -0500 Organization: Illinois State University- Instructional Technology Services Message-ID: <32695DED.7A63@rs6000.cmp.ilstu.edu> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547n5o$310@sun0.urz.uni-heidelberg.de> <01bbbdd5$69724200$5b1018ce@barrington> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Chris Trimble wrote: > You're right, there is a lot of interest; and NeXT is a company that can > push out a modestly priced implementation right now -- which is my point. > There's a strong grass-roots movement going on right now... I don't quite > understand why they should wait until that movement gains a foothold to > realize that it's a good market. GNUStep has already shown it will be > popular and is within our grasp; why NeXT doesn't capitalize on it > boggles my mind and is why I posted the original message. I challenge anyone to find a SINGLE mention of GNU OPENSTEP (GNUstep) on the NeXT web site. Hmm, one would most definitely think that NeXT would want to mention GNU as a PARTNER for OPENSTEP!!! GNU SHOULD be mentioned right next to SUN. Heck, HP and DEC aren't doing ANYTHING with OPENSTEP yet... Having GNU as a partner is INVALUABLE, why does NeXT seem so blind to this? -- Eric A. Dubiel; http://www.ilstu.edu/~eadubie mailto:eadubie@rs6000.cmp.ilstu.edu ASCII, MIME, SUN, NeXT, PGP Mail ytalk:eadubie@138.87.201.11 Instructional Technology Services- Illinois State University "Intelligence is the ultimate aphrodisiac." - Dr. Timothy Leary ALL VIEWS EXPRESSED REPRESENT MYSELF ONLY
From: vazquezr@physics.ucla.edu (Rick Vazquez) Newsgroups: comp.sys.next.programmer Subject: help, with root Date: 19 Oct 1996 23:17:24 GMT Organization: University of California, Los Angeles Message-ID: <54bni4$1jos@uni.library.ucla.edu> I have some problems and I was hoping someone might have the solution. When I try to su root, I get the following message... localhost> pppd localhost> su root Password: su: setgid: Not owner localhost> But I have entered the correct password. On My console I get.... Oct 19 16:11:53 localhost su: SU to root by vazquezr on /dev/ttyp1 so it looks as thought I have su root. An I am part of the wheel. The other problem which may fix the first, is.....When I try to erase a user from my computer I get a message from the user manager which tells me that my netinfo file is read only. Can someone please tell me how to make it writable again? Please help Rick UCLA Physics vazquezr@physics.ucla.edu
From: wilkie@cg.tuwien.ac.at (Alexander Wilkie) Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 20 Oct 1996 11:06:36 GMT Organization: Vienna University of Technology, Austria Message-ID: <54d13s$7cb@news.tuwien.ac.at> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547n5o$310@sun0.urz.uni-heidelberg.de> <01bbbdd5$69724200$5b1018ce@barrington> <32695DED.7A63@rs6000.cmp.ilstu.edu> Cc: eadubie@rs6000.cmp.ilstu.edu > Having GNU as a partner is INVALUABLE, why does NeXT seem so blind to > this? Umm... Because they are (intentionally) blind to almost anything that will promote their current product? It's not like this being an atari/amiga/acorn/[insert favourite cool, but long dead system here] advocacy newsgroup. Openstep/Mach/Intel is a highly productive, stable and up-to-date platform that would just require maintenance and the occasional addition of new features. And what does the "insanely great" management at NeXT Inc. do? They try to (and sure will) kill Mach, just like it were something to be ashamed of. They want to be a "web company" _instead_. Why not both, for heavens sake? As someone who got to admire the elegance and conceptual clarity of NeXTStep (speaking relative to, say, WinNT) through daily use over the last year or two I can just shake my head in puzzlement. Maybe pig farming is an alternative after all... Just my $0.2E-32 Alexander Wilkie -- e-mail: wilkie@cg.tuwien.ac.at (NeXTMail preferred, MIME o.k.) www : http://www.cg.tuwien.ac.at/~wilkie/
From: aisbell@ix.netcom.com (Art Isbell) Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 20 Oct 1996 16:08:46 GMT Organization: Netcom Distribution: world Message-ID: <54diqe$i2j@sjx-ixn3.ix.netcom.com> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547n5o$310@sun0.urz.uni-heidelberg.de> <01bbbdd5$69724200$5b1018ce@barrington> <32695DED.7A63@rs6000.cmp.ilstu.edu> "Eric A. Dubiel" <eadubie@rs6000.cmp.ilstu.edu> wrote: > I challenge anyone to find a SINGLE mention of GNU OPENSTEP (GNUstep) on > the NeXT web site. Hmm, one would most definitely think that NeXT would > want to mention GNU as a PARTNER for OPENSTEP!!! GNU SHOULD be > mentioned right next to SUN. Heck, HP and DEC aren't doing ANYTHING > with OPENSTEP yet... > > Having GNU as a partner is INVALUABLE, why does NeXT seem so blind to > this? I have no way of knowing, but I assume that NeXT's intended audience for its Web site are commercial enterprises, NeXT's potential customers. Most U.S. commercial enterprises don't consider the Free Software Foundation's products to be of "commercial quality", mostly because nationwide support isn't available (I don't believe Cygnus has offices around the U.S.). So mentioning GNU doesn't carry much, if any weight with commercial enterprises. This may not be justifiable, but it is reality at this time. Many vendors package GNU utilities with their products, but the companies that buy them usually don't even know that GNU utilities are included. The vendor provides the support in these cases. -- Art Isbell NeXT/MIME Mail: aisbell@ix.netcom.com Trego Systems Voice/Fax: +1 408 335 2515 CaseServ: OPENSTEP Voice Mail: +1 408 335 1154 managed care solutions US Mail: Felton, CA 95018-9442
From: aisbell@ix.netcom.com (Art Isbell) Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 20 Oct 1996 16:24:24 GMT Organization: Netcom Distribution: world Message-ID: <54djno$i2j@sjx-ixn3.ix.netcom.com> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547n5o$310@sun0.urz.uni-heidelberg.de> <01bbbdd5$69724200$5b1018ce@barrington> <32695DED.7A63@rs6000.cmp.ilstu.edu> <54d13s$7cb@news.tuwien.ac.at> wilkie@cg.tuwien.ac.at (Alexander Wilkie) wrote: > It's not like this being an atari/amiga/acorn/[insert favourite cool, but > long dead system here] advocacy newsgroup. Openstep/Mach/Intel is a highly > productive, stable and up-to-date platform that would just require > maintenance and the occasional addition of new features. This is certainly true, but capitalism isn't fair. The best product for the lowest price isn't guaranteed of success. A mediocre product at a very low price with the best marketing is more likely to succeed. NeXT's choice of Mach seems to have been similar to its choice of Motorola CPUs and Objective-C. All were excellent choices when they were made, but none of them has succeeded. BSD UNIX, Mach's personality, is no longer being developed. Microsoft hired the CMU professor who was a principle Mach designer and I'm not sure what is happening with Mach at CMU. Objective-C lost the object-oriented C war to C++. So when all of this is put together, companies just don't want a NEXTSTEP solution because it's too "different" from everything else they have. Finding NS programmers and system administrators is very difficult. NS's UNIX isn't very similar to System V UNIX used by the major UNIX vendors. NS's hardware requirements are very stringent and expensive. > And what does the "insanely great" management at NeXT Inc. do? They try to > (and sure will) kill Mach, just like it were something to be ashamed of. They > want to be a "web company" _instead_. Why not both, for heavens sake? NeXT isn't killing Mach; the marketplace is. Despite having by far the best product in our market, we have been almost totally unsuccessful selling it for one reason only: it's a NS solution. Customers want a Windows solution and will buy inferior, but less expensive Windows solutions over NS :-( > As someone who got to admire the elegance and conceptual clarity of NeXTStep > (speaking relative to, say, WinNT) through daily use over the last year or > two I can just shake my head in puzzlement. I agree, but NT will run the many Windows apps that companies already have. SoftPC just isn't an acceptable solution. NS is a special-purpose solution, and most companies won't buy a solution that requires dedicated hardare and its own OS for which few products are available. Maybe pig farming is an > alternative after all... Working with NT feels similar :-) -- Art Isbell NeXT/MIME Mail: aisbell@ix.netcom.com Trego Systems Voice/Fax: +1 408 335 2515 CaseServ: OPENSTEP Voice Mail: +1 408 335 1154 managed care solutions US Mail: Felton, CA 95018-9442
From: ts110@pmms.cam.ac.uk (Tomaz Slivnik) Newsgroups: comp.sys.next.programmer Subject: Unable to find class: Application Date: 20 Oct 1996 18:57:00 GMT Organization: University of Cambridge, England Message-ID: <54dsls$nkm@lyra.csx.cam.ac.uk> I've converted one of my 3.0 projects to OpenStep 4.0. The project compiles fine, however, when I try to execute it, I get the following error: Oct 20 19:52:16 Graphicon[12577] Unable to find class: Application, exiting I seem to have tried everything, but I can't find the source of the error. Also, I am surprised that there is a problem finding class Application since all references to the Application class in my project have been changed to NSApplication. Anyone had a similar problem before and worked it out? Thanks for your help! Tomaz
Newsgroups: comp.sys.next.programmer Subject: Re: NEWBIE: Polymorphism? Message-ID: <326A8299.31E1@running-start.com> From: Ralph Zazula <zazula@running-start.com> Date: Sun, 20 Oct 1996 12:50:49 -0700 References: <5486mf$epk@camelot.dsccc.com> Organization: Running Start, Inc. MIME-Version: 1.0 To: James Michel <jmichel@imtn.dsccc.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit James Michel wrote: > > I am currently reading an Objective-C book entitled "Objective-C Object > Oriented Programming Techniques" by Lewis J. Pinson and Richard S. Weiner. > In the section covering the mechanics of messaging I ran across this > > "All implementations of identically named methods must have the > same return type and the same argument types since the messaging > function has access to the method implementation only through > selectors, and thus treats all methods with the same selector > alike." > > Now this sounds like their is no polymorphism in Objective-C OR I need to > take this statement in the context of the mechanics of messaging discussion. > [...] > Hi - This statement is correct and is explained by the fact that Objective-C selectors do not contain information about argument and return types. This does not mean that there is a lack of polymorphism. Polymorphism refers to the ability to send the same message to instances of different classes. This is possible when the classes share a common, identical, interface -- which includes argument and return types. If the methods of these classes took different arguments, or returned different types of values, they probably couldn't be used interchangably. Ralph --- Ralph Zazula Running Start, Inc. http://www.running-start.com zazula@running-start.com
Newsgroups: comp.sys.next.programmer Subject: Re: NEWBIE: Dynamically created class instances? Message-ID: <326A835F.905@running-start.com> From: Ralph Zazula <zazula@running-start.com> Date: Sun, 20 Oct 1996 12:54:07 -0700 References: <5486rb$epk@camelot.dsccc.com> Organization: Running Start, Inc. MIME-Version: 1.0 To: James Michel <jmichel@imtn.dsccc.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit James Michel wrote: > > I am now beginning my study of Objective-C and have a question which > the Objective-C material I have does not answer. How do I create > instances of classes dynamically? If I have a NeXTSTEP application > that uses an unknown amount of instances of a special Window class > how can I create and keep a handle on the multiple instances? If I do > something like > > id myWindow; > > myWindow = [SpecialWindow new]; > > and connect this method to a "new" button or menu item then I will get > one new window instance, myWindow. What if the user needs another > instance? Do I have to use pointers like in C to accomplish this? > Hi - You can certainly create mutliple variables of type 'id' (object pointer) and use them to store references to your dynamically created objects. Another option is to use a "collection" class such as FoundationKit's NSArray to store mutliple instances. Ralph --- Ralph Zazula Running Start, Inc. http://www.running-start.com zazula@running-start.com
From: ts110@pmms.cam.ac.uk (Tomaz Slivnik) Newsgroups: comp.sys.next.programmer Subject: cmsg cancel <54dsls$nkm@lyra.csx.cam.ac.uk> Control: cancel <54dsls$nkm@lyra.csx.cam.ac.uk> Date: 21 Oct 1996 00:14:28 GMT Organization: Cambridge University, Pure Mathematics and Mathematical Statistics Message-ID: <54ef94$4a2@lyra.csx.cam.ac.uk> Article cancelled from within tin [v1.2 PL2]
From: k_harbour@mail.bogo.co.uk (Karl Harbour) Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: Mon, 21 Oct 1996 00:22:44 GMT Organization: Orbital Computer Consultancy Message-ID: <3273c1c4.38219381@news.plsys.co.uk> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547n5o$310@sun0.urz.uni-heidelberg.de> <01bbbdd5$69724200$5b1018ce@barrington> <32695DED.7A63@rs6000.cmp.ilstu.edu> <54d13s$7cb@news.tuwien.ac.at> <54djno$i2j@sjx-ixn3.ix.netcom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit aisbell@ix.netcom.com (Art Isbell) wrote: <-- Stuff I agree with --> > NeXT isn't killing Mach; the marketplace is. Slight disagreement here. Next are partly responsible for the state of the marketplace. They could still have a bigger niche with proper NextStep today, I think, if they'd played things differently over the last 3 to 4 years. >Despite having by far the >best product in our market, we have been almost totally unsuccessful selling >it for one reason only: it's a NS solution. Customers want a Windows >solution and will buy inferior, but less expensive Windows solutions over NS >:-( > A lot of people would say, "Daft customers,don't know what they're missing". But I say, if Next had encouraged a third party software industry around NextStep over the last 8 years, there would be a viable Next alternative to Office. Customers want Windows because they want Excel and Word. >> As someone who got to admire the elegance and conceptual clarity of >NeXTStep >> (speaking relative to, say, WinNT) through daily use over the last year or >> two I can just shake my head in puzzlement. > > I agree, but NT will run the many Windows apps that companies already >have. SoftPC just isn't an acceptable solution. NS is a special-purpose >solution, and most companies won't buy a solution that requires dedicated >hardare and its own OS for which few products are available. > Yep. >Maybe pig farming is an >> alternative after all... > > Working with NT feels similar :-) >-- >Art Isbell NeXT/MIME Mail: aisbell@ix.netcom.com >Trego Systems Voice/Fax: +1 408 335 2515 >CaseServ: OPENSTEP Voice Mail: +1 408 335 1154 > managed care solutions US Mail: Felton, CA 95018-9442 puzzlement. Maybe pig farming is an >alternative after all... >
From: "Eric A. Dubiel" <eadubie@rs6000.cmp.ilstu.edu> Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: Mon, 21 Oct 1996 01:19:20 -0500 Organization: Illinois State University- Instructional Technology Services Message-ID: <326B157D.65C0@rs6000.cmp.ilstu.edu> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547n5o$310@sun0.urz.uni-heidelberg.de> <01bbbdd5$69724200$5b1018ce@barrington> <32695DED.7A63@rs6000.cmp.ilstu.edu> <54diqe$i2j@sjx-ixn3.ix.netcom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Art Isbell <aisbell@ix.netcom.com> Art Isbell wrote: > I have no way of knowing, but I assume that NeXT's intended audience for > its Web site are commercial enterprises, NeXT's potential customers. Most > U.S. commercial enterprises don't consider the Free Software Foundation's > products to be of "commercial quality", mostly because nationwide support > isn't available (I don't believe Cygnus has offices around the U.S.). So > mentioning GNU doesn't carry much, if any weight with commercial enterprises. > This may not be justifiable, but it is reality at this time. I think it DEFINITELY DOES! It's the truest form of a standard, it's FREE and OPEN if it's GNU!!!! GNUstep is COMMERCIALLY SUPPORTED TODAY FROM NET-COMMUNITY http://www.net-community.com GNUstep is also FREE AND AVAILABLE TODAY in an alpha version from http://www.gnustep.org That says something! So does this from the Net-Community pages: WSC Technologies http://www.wsc.com WSC Technologies is a New York based firm that uses object-oriented tools to build complex financial systems for the banking and brokerage community. Paul Murphy, Director of Technology, says "GNUstep helps us provide timely solutions to our clients. Because the source code is available for us to review and modify, we can provide better and more cost-effective support." CD purchase Price: $30 The purchase of the CD-ROM comes with installation support; resolution of problems concerning installation of the software from the CD onto the user's system as well as help with basic setup tasks comes free of charge. GNUstep developer Price: $50/year; includes latest CD release. This support package is geared towards the individual developer, who is using the GNUstep and MediaBook software as a basis for their own software. It provides for a number of additional features beyond the basic CD purchase, plus it has the nice feature of offering support on weekends for individuals who may be programming in their spare time. Single customer contact. E-mail/Fax support with maximum 3 business day response time. 9am to 6pm EST; 7 days a week support. Automatic notification of known bugs, workarounds, and patches. 50% discount on upto one (1) copy of additional CD releases. GNUstep corporate Price: $200/year; $50/year for each additional developer. This support package is geared towards the small software development business that has multiple developers using the GNUstep and MediaBook software as a basis for their own software; it goes beyond the GNUstep developer plan by offering a quicker turnaround time and free CD releases, but it is restricted to weekdays. Additional developers can be easily added to the plan. Includes license for three copies of single CD. Upto three customer contacts. E-mail/Fax support with maximum 2 business day response time. 9am to 6pm EST; Monday to Friday support. Automatic notification of known bugs, workarounds, and patches. 100% discount on upto one (1) copy of additional CD releases. Additional developers allow for additional copies and additional customer contacts. MediaBook comprehensive Price: $2500/year; $500/year per additional contact. This support package is for organizations that are deploying GNUstep and MediaBook software throughout their organization and require immediate, mission-critical support; it is the only plan which allows telephone access directly with a developer. It also allows the customer contact to utilize MediaBook's remote support application; this application can be used to directly enter support incidents across the Internet, review status of open support incidents, and display historical information for all support incidents. Provides ten physical copies of CD. Includes license for unlimited copies of CD within organization. Single customer contact. Guaranteed highest priority for support incidents. E-mail/Fax/Telephone support with maximum next business day response time. 9am to 6pm EST; 7 days a week support. Automatic notification of known bugs, workarounds, and patches. Allows customer contact to directly speak with developer on telephone. 100% discount on upto ten (10) copies of additional CD releases. Usage of MediaBook's remote support application. MediaBook custom Price: $CALL We can custom arrange a support package to meet your needs and budget. -- Eric A. Dubiel; http://www.ilstu.edu/~eadubie mailto:eadubie@rs6000.cmp.ilstu.edu ASCII, MIME, SUN, NeXT, PGP Mail ytalk:eadubie@138.87.201.11 Instructional Technology Services- Illinois State University VOTE LIBERTARIAN FOR smaller government AND LIBERTY! http://www.lp.org ALL VIEWS EXPRESSED REPRESENT MYSELF ONLY
From: gvandyk@icon.co.za Newsgroups: comp.sys.next.programmer Subject: NXImageRep Date: 21 Oct 1996 06:36:22 GMT Organization: E.S. Systems cc (Financial Systems Development) Message-ID: <54f5l6$fac@hermes.is.co.za> Is there an NXImageRep available for Gif Files? If so, where can I find it. -- Regards, Gerrit van Dyk email: gvandyk@icon.co.za (NeXTMail welcome) E.S. Systems cc The OBJECT is the ADVANTAGE
From: devan2m@imap2.asu.edu Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Followup-To: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Date: 21 Oct 1996 09:59:10 GMT Organization: Arizona State University Message-ID: <54fhhe$794@news.asu.edu> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547gs6$131@news.Dortmund.Germany.EU.net> <548hm5$s42@news.onramp.net> Hello Steve, Darren, Felipe, & any others who are following or care to contribute: Although Steve makes an interesting suggestions, to give away NEXTSTEP/OPENSTEP academic for free, it is a proposal that really ignores the parameters of the marketing challenge as presented. What's more, and rather than head down a path which examines what NeXT and their marketing folks should and shouldn't do, there is an entirely different direction that I would like to see this discussion take. This directions is one which is governed by the following constraints and hopefully will offer some novel insights as to what can actually be done to fortify and radically increase the market presence of NEXTSTEP 3.3 & an OPENSTEP for MACH which is more in-line with the evolution of NEXTSTEP as it should be on course of Mecca and not as some Windoze/Solaris looking chameleon or derivative. Recall that the original marketing challenge was presented as follows: (Excerpt...) Would you care to provide specifics about what NeXT, someone, or some entity independent of NeXT ( a consortium of advocates) acting on behalf of the NEXTSTEP community and the technology might do from a marketing standpoint to change this course? This would all be ased on the premise that despite NEXTSTEP 3.3* not having garnered the market share and following everyone who favors it would like or having a zillion developers and software packages, it still has the best user and developer environments on the planet. To make it interesting or should I say challenging, you would only have USD $10 M to work with and would need to deliver a principle and return of $15 M on this investment in the NeXT community within 3 years. Furthermore, you have only six months in which to egineer and produce measurable results of an effective longer-term master plan. A good barometer of just how effective and well your master marketing plan would be in the requisite time frame would be if there were 1-2 million new U.S. based NEXTSTEP* end-users were to purchase the software in this period. ............ In order to eliminate certain undesired diatribe and stimulate the creative flow for addressing the above exercise by all you marketing masterminds, let's doing the following: 1) Eliminate NeXT Software, Inc. from the equation and any marketing efforts by NeXT in order to accomplish these objectives. Think of this entire operation as something being undertaken by an entity independent of NeXT ( a consortium of advocates) acting on behalf of the NEXTSTEP community and the technology - please don't confuse this with the GNUStep efforts. 2) Key objectives on the list of this Consortium of NEXTSTEP Advocates follow: A. You have only a budget of USD $10 Million to accomplish everything with. B. Must make measurable impact of say 1-2 million new U.S. based end users within six months. In short, the reason for concentrating on suring up U.S. based end users means that foreign users get supported too owing to the increased success. Obviously, and considering the proposal of giving away NEXTSTEP academic, if one were to use all $10 M to buy X-# of academic copies at about $200.00 per copy to give away to qualified folks, this doesn't leave much money to advertise or get the word out. There would also be things like shipping to consider. As this consortium would need to make a deal with NeXT in the first place to purchase a large number of academic copies of the software, it would also need to insure and verify that the software is registered to people in academia. There would be overhead incurred doing this. Anyway, there seem to be a couple of factors or constraints that were overlooked with the academic giveaway. By giving away the software, would you be able to obtain a measurable marketable impact by this within six months and how does one get a $15 M return on this investment in 3 years? In any case, I hope everyone starts to get the idea when I ask about details and for something that is plausible to accomplish the described objective. Thanks To All - Steve Dekorte (dekorte@suite.com) wrote: : devan2m@imap2.asu.edu wrote: : > Would you care to provide specifics about what NeXT, ... might do : from a : > marketing standpoint to change this course? ... : > To make it interesting or should I say challenging, you would : > only have $10 M USD to work with and would need to deliver a principle : > and return of $15 M on this investment in the NeXT community within 3 : > years... : Here's my shot: : Give NeXTstep/OpenStep academic away. : NeXT can't be making even $1M from this market, yet doing this would : do far more than 10 times as much spent on marketing. : How's that for a return on investment? : -- : Steve Dekorte - OpenStep Developer - Anaheim, CA : "Fundamentalism isn't about religion. It's about power." - S. Rushdie
From: flight@mathi.uni-heidelberg.de (Gregor Hoffleit) Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Followup-To: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Date: 21 Oct 1996 15:28:06 GMT Organization: University of Heidelberg, Germany Message-ID: <54g4q6$2p@sun0.urz.uni-heidelberg.de> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547n5o$310@sun0.urz.uni-heidelberg.de> <sk68-0101040021590001@cu-dialup-0017.cit.cornell.edu> Sung Ho Kim (sk68@cornell.edu) wrote: : I can't be but a little disappointed by Mr.Hoffleit's comments regarding : his views about NeXT and where it should go. First and foremost, I am not : a programmer or a fanatic; I'm just a normal user. But what caught my : fancy about NeXT and its products was what it stood for in this industry : of also-rans. (I consider Linux & BeOS to be in the also-ran department) : It took a different approach to computing that will hopefully be rewarded : in the annals of computing history. In a market full or market-share : mongrels, it tried to produce something that was no-compromise, first-rate : and innovative. I don't think anybody can doubt that they did in fact : produce something special. I think this is what makes NeXT special. I wholeheartedly agree. : If NeXT had started off with the kind of philosophy Mr. Hoffleit thinks NeXT : should take now, I think NeXT as we know it now (albeit its small niche : position) would not exist. If in fact NeXT thought in such calculated and : market oriented way in the beginning, it would have ended up in the : also-ran department of computer history. But in fact it didn't. People : can laugh all they want about the market position NeXT is in right now, and : they can have their theories about doing this and that with NeXT, but in : the end it is a fact that many of the NeXT users have fallen in love with : this innovative side of NeXT and not the market oriented side of NeXT. : And as Byte magazine once said of NeXT, : "NeXTSTEP is probably the most respected software in the planet." I could sign up for this, too. Nevertheless, my point was that in the situation we are _now_, it's a strong move to leave the field of OS wars. For such a small company as NeXT is, it's simply not possible to fight Windows on the Intel architecture. There's a small chance for companies like Be and Apple, since they sell software _and_ hardware (if you look carefully, that's what NeXT tried to accomplish with their certified systems corner), and therefore don't have to worry about compatibility with 100's of graphic cards and dozens of SCSI adapters coming out each month. It's just an unfortunate timing: When the black HW was dropped and NSfIP came out, there was a chance to establish NEXTSTEP as big player in the Intel OS SOHO market. It didn't succeed due to a) it's pricing scheme, b) lack of shrink-wrapped apps and c) memory requirements and prices. Nowadays c) is gone, but d) NEXTSTEP on Intel has lost quite a few of of its advantages over Windows (from the user sight), b) is even more visible and a) is what we're discussing. So let's be happy that NeXT is still alive (look at IBM & OS/2), and doing quite well financially, and let's hope that Steve's leaving NeXT for good soon to start the next thing.. Gregor -- | Gregor Hoffleit Mathematisches Institut, Uni HD | | flight@mathi.uni-heidelberg.de INF 288, 69120 Heidelberg, Germany | | (NeXTmail, MIME) (49)6221 54-5771 fax 54-8312 | | PGP Key fingerprint = 23 8F B3 38 A3 39 A6 01 5B 99 91 D6 F2 AC CD C7 |
From: far@ix.netcom.com(Felipe A. Rodriguez) Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 21 Oct 1996 18:22:33 GMT Organization: Netcom Message-ID: <54gf19$5hl@dfw-ixnews3.ix.netcom.com> References: <54g4q6$2p@sun0.urz.uni-heidelberg.de> In article <54g4q6$2p@sun0.urz.uni-heidelberg.de> flight@mathi.uni-heidelberg.de (Gregor Hoffleit) writes: snip For such a small company as NeXT is, it's simply not possible to >fight Windows on the Intel architecture. Why must one fight for market dominance? You seem to believe that there is room for only one OS player. I for one don't want to see NeXT dump a couple mill into a marketing campaign. On the otherhand wider product distribution could do much more. A couple of feet of shelf space at your local Frye's or CompUSA would probably double sales of NeXTStep in short order while costing very little. There's a small chance for >companies like Be and Apple, since they sell software _and_ hardware Apple is quite likely. Be? Be is so dead the coffin it's in is decaying. I've yet to hear even one good reason to buy a BeBox. Be is just another proprietary OS and hardware. >(if you look carefully, that's what NeXT tried to accomplish with >their certified systems corner), and therefore don't have to worry >about compatibility with 100's of graphic cards and dozens of SCSI >adapters coming out each month. > To have a good OS you don't need to worry about supporting 100's of hardware peripherals. If you support a couple of cards from the big name brands you will be OK (i.e. Diamond, Number Nine, ATI, Matrox). >It's just an unfortunate timing: When the black HW was dropped and >NSfIP came out, there was a chance to establish NEXTSTEP as big player >in the Intel OS SOHO market. It didn't succeed due to a) it's pricing >scheme, b) lack of shrink-wrapped apps and c) memory requirements and >prices. Nowadays c) is gone, but d) NEXTSTEP on Intel has lost quite a >few of of its advantages over Windows (from the user sight), b) is >even more visible and a) is what we're discussing. > Your reasons all have merit, but you missed the biggest reason NS has not sold well. Basically, how many people are going to buy what they've never seen? >So let's be happy that NeXT is still alive (look at IBM & OS/2), and >doing quite well financially, and let's hope that Steve's leaving NeXT >for good soon to start the next thing.. > > Gregor > > >-- >| Gregor Hoffleit Mathematisches Institut, Uni HD | >| flight@mathi.uni-heidelberg.de INF 288, 69120 Heidelberg, Germany | >| (NeXTmail, MIME) (49)6221 54-5771 fax 54-8312 | >| PGP Key fingerprint = 23 8F B3 38 A3 39 A6 01 5B 99 91 D6 F2 AC CD C7 | -- 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: dekorte@suite.com (Steve Dekorte) Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 21 Oct 1996 18:30:18 GMT Organization: Suite Software Message-ID: <54gffq$7p7@news.onramp.net> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547gs6$131@news.Dortmund.Germany.EU.net> <548hm5$s42@news.onramp.net> <54fhhe$794@news.asu.edu> Cc: devan2m@imap2.asu.edu devan2m@imap2.asu.edu wrote: > Obviously, and considering the proposal of giving away NEXTSTEP > academic, if one were to use all $10 M to buy X-# of academic copies at > about $200.00 per copy to give away to qualified folks, this doesn't > leave much money to advertise or get the word out. I'd be interesting in the % of NeXTstep sites that were started by people seeing NeXT marketing info and buying, without having any NeXTheads on site to promote it's use? If this % is as low as I suspect, and most current NeXT sites were seeded with NeXTstepping academics, then the best marketing plan is to give away to academics who have a chance of becoming NeXTstep advocates - in essense, free long-term NeXT reps. I've worked on NeXTstep in the SE, NE and SW and visted NeXT sites across the country and I can tell you that far more copies of NeXTstep were sold via the actions of NeXT advocates than paid NeXT reps or advertising. We can say NeXT has done "everything wrong" to date, but I think there's alot more to it than that. IMO, people are simply scared of things they don't know and unless there's someone there that knows it and pushes it, anything that's not "mainstream" hasn't got a prayer. > ...By giving away the software, > would you be able to obtain a measurable marketable impact by this within > six months... Probably not. I wouldn't recommend it only as part of a long term marketing plan. > how does one get a $15 M return on this investment in 3 years? Wait until they graduate. -- Steve Dekorte - OpenStep Developer - Anaheim, CA "Fundamentalism isn't about religion. It's about power." - S. Rushdie
From: dekorte@suite.com (Steve Dekorte) Newsgroups: comp.lang.objective-c,comp.sys.next.programmer Subject: Re: NEWBIE: Polymorphism? Date: 21 Oct 1996 20:53:55 GMT Organization: Suite Software Message-ID: <54gnt3$7p7@news.onramp.net> References: <5492vq$573@news.onramp.net> <54gdt9$em@news.next.com> Cc: Mark_Bessey@next.com Mark Bessey wrote: > > > Steve Dekorte writes > > > > The neat thing with Objective-C and Smalltalk is that you can use > > > > the "id" type for all objects, so you can dispence with the C++ > > > > type-casting mess. > > > ...it's not a problem if you're clever enough to > > use classes with encapsulate the primative data-types. (Or at least > > enough to use a method naming scheme that tells you about > > non-id data types) > > y1 = (-b + sqrt(b * b - 4 * a * c))/(2*a); > into this: > y1 = [[b negate] plus: [[[b times: b] minus: [[a times: c] > times: 4] sqrt] divide: [a times: 2]]]; > > [1].. tell me which one is more straightforward. > [2] Not to mention the 8 or so temporary objects that get created > and destroyed in the process. Both of these are straw men. There's nothing to keep you from doing the above *within* a method. Notice that you don't access any outside data, and if a,b,c,d were passed in as arguments, then your data was in the wrong place to begin with. Wrting functional code, and then saying it's not elegant in an OO language is clearly no argument. As for the temp objects; Again, no language can save poor programmers. Here's how to avoid creating objects: [[[y1 setIntValue:4] multiply:a] multiply:c]; [y1 addFloatValue:[b squareFloatValue]]; [[y1 squareRoot] subtract:b]; [[y1 half] divide:a]; One of the reasons I don't like garbage collection is because it encourages the sort of code in your example. -- Steve Dekorte - OpenStep Developer - Anaheim, CA "Fundamentalism isn't about religion. It's about power." - S. Rushdie
From: Mark.A.Tarbell@jpl.nasa.gov (Mark Tarbell) Newsgroups: comp.sys.next.programmer Subject: Re: Openstep (port|comp)atibility Sun vs NeXT Date: 21 Oct 1996 22:54:28 GMT Organization: Jet Propulsion Laboratory - Pasadena CA Distribution: world Message-ID: <54guv4$7of@netline-fddi.jpl.nasa.gov> References: <53tu7g$oih@news.sct.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit In <53tu7g$oih@news.sct.fr> Alexandre Odoux wrote, in part: > > For example, it seems that NSTableView has become part of the Appkit > (in 4.0 ?). Is this valid for Sun's Openstep as well ? Yes, NSTableView is part of the Openstep/Solaris AppKit. It's there & works (even if it doesn't appear in the documentation). > And what about the source compatibility between Openstep NT and > Openstep Mach ? Is it hard to maintain ? This isn't so easy.... 1. Because the Sun is System V and the NeXT is BSD, there is always the problem of include files (when dealing with the rtl routines). A routine which might be defined in libc.h on the NeXT might be in stdlib.h on Solaris, etc. 2. The compilers aren't the same on the two machines. The NeXT Openstep compiler (from Project Builder) is extremely picky, to the point of aborting compilations when given code that compiles and runs fine on OpenStep/Solaris. This can take some code massaging, but the resulting modified code should run okay on both. 3. There are run-time differences. Again, Openstep/Solaris seems more permissive. For instance, I had to mess with NSAutoreleasePool on the NeXT on code that worked fine without it on the Sun. (The NeXT kept screaming at me that I was leaking objects (NSConcreteDate) - from an internal call within one of their own objects (NSConditionLock) !!) 4. And lastly, because the implementations are not the same code, you get different results. NSThreads and NSLocks seem squirrely on the beta Openstep/Solaris (often threads don't start!); they have seemed stable on the NeXT 4.0 so far. Should be getting the real Openstep/Solaris ERD soon, though. This bug is reportedly fixed. Mark
From: Mark Anenberg <marka@Eng.Sun.COM> Newsgroups: comp.sys.next.programmer Subject: Re: Openstep (port|comp)atibility Sun vs NeXT Date: Mon, 21 Oct 1996 18:15:26 -0700 Organization: Sun Microsystems Inc. Message-ID: <326C202E.5508@Eng.Sun.COM> References: <53tu7g$oih@news.sct.fr> <543aa6$n12@news.next.com> <544ddo$o66@hermes.is.co.za> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit CC: Alexandre Odoux <aeodoux@worldnet.fr>, Mark Bessey <Mark_Bessey@next.com> gvandyk@icon.co.za wrote: > > On 10/16/96, Mark Bessey wrote: > >Alexandre Odoux writes > >> For example, it seems that NSTableView has become part of the > Appkit > >> (in 4.0 ?). Is this valid for Sun's Openstep as well ? > > > >It doesn't look like it. OPENSTEP/Mach's headers have NSTAbleView as > > #ifndef STRICT_OPENSTEP > >So, it looks to be a NeXT-only enhancement. > > > > NSTableView is part of Openstep for Sun. If you get the developer you > will also get the docs for this. I haven't used it on Sun/Openstep > yet, but the headers and docs are all there so I would presume that > the classes exist The Palette for NSTableView also exists. > > I personally think that Sun has made a very good port of Openstep. It > looks much better on Sun than it does on Windows/NT. On NT you > basically only get the runtime environment, no useful apps etc, etc. > On sun you get the FileViewer and Mail as well. > -- > Regards, > Gerrit van Dyk > email: gvandyk@icon.co.za (NeXTMail welcome) > E.S. Systems cc > The OBJECT is the ADVANTAGE Thanks for the good words; NSTableView is, in fact, there in Solaris OpenStep; Items that are marked as #ifndef STRICT_OPENSTEP are not necessarily in one product or the other; It's an indication that the enclosed item is not strictly part of the openstep specification; There are many, many examples of spec extensions that both Sun and NeXT support, and they are all marked as above; Mark Anenberg , OpenStep Development Team, Sun Microsystems, Inc. [E,NeXT,Mime]Mail: mark@velodrome.com, http://www.velodrome.com/ Disclaimer: The opinions expressed above are my own and in no way represent those of Sun Microsystems, Inc.
From: Mark Anenberg <marka@Eng.Sun.COM> Newsgroups: comp.sys.next.programmer Subject: Re: reading NSCalendarDate Date: Mon, 21 Oct 1996 18:39:54 -0700 Organization: Sun Microsystems Inc. Message-ID: <326C25E9.5C47@Eng.Sun.COM> References: <5402ot$r86@rks1.urz.tu-dresden.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Andreas Dietzsch wrote: > > How can I get a NSString from a NSCalendarDate? > > Thank You > > Andreas maybe what you want is one of these? - (NSString *)description; /* uses calendarFormat and timeZone, canonical locale */ - (NSString *)descriptionWithLocale:(NSDictionary *)locale; /* uses calendarFormat and timeZone, supplied locale */ - (NSString *)descriptionWithCalendarFormat:(NSString *)format; /* uses supplied format, canonical locale */ - (NSString *)descriptionWithCalendarFormat:(NSString *)format locale:(NSDictionary *)lo cale; /* uses supplied format and locale */ Mark Anenberg , OpenStep Development Team, Sun Microsystems, Inc. [E,NeXT,Mime]Mail: mark@velodrome.com, http://www.velodrome.com/ Disclaimer: The opinions expressed above are my own and in no way represent those of Sun Microsystems, Inc.
From: Sam Krishna <infinity@nol.net> Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: Tue, 22 Oct 1996 00:56:17 -0500 Organization: Networks On-Line Message-ID: <326C6201.746302A@nol.net> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547gs6$131@news.Dortmund.Germany.EU.net> <548hm5$s42@news.onramp.net> <54fhhe$794@news.asu.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: devan2m@imap2.asu.edu Well, This is interesting.... As a Linux/NT/NeXTSTEP user who gets to play around in all kinds of environments... I would have to say that by far, the best OS I've seen to date is NeXTSTEP. Here's some ideas on marketing it.... >devan2m@imap2.asu.edu wrote: > > Hello Steve, Darren, Felipe, & any others who are following or care to > contribute: > > Although Steve makes an interesting suggestions, to give away > NEXTSTEP/OPENSTEP academic for free, it is a proposal that really ignores > the parameters of the marketing challenge as presented. What's more, and > rather than head down a path which examines what NeXT and their marketing > folks should and shouldn't do, there is an entirely different direction > that I would like to see this discussion take. This directions is one > which is governed by the following constraints and hopefully will offer > some novel insights as to what can actually be done to fortify and > radically increase the market presence of NEXTSTEP 3.3 & an OPENSTEP for > MACH which is more in-line with the evolution of NEXTSTEP as it should be > on course of Mecca and not as some Windoze/Solaris looking chameleon or > derivative. > > Recall that the original marketing challenge was presented as follows: > > (Excerpt...) > > 1) Eliminate NeXT Software, Inc. from the equation > and any marketing efforts by NeXT in order to accomplish these > objectives. Think of this entire operation as something being undertaken > by an entity independent of NeXT ( a consortium of advocates) acting on > behalf of the NEXTSTEP community and the technology - please don't > confuse this with the GNUStep efforts. > I like this - they are too unreliable anyway... I wouldn't write GNUstep out of the equation just yet, though, but we deal with things the way they are defined, not the way we would like them... > 2) Key objectives on the list of this Consortium > of NEXTSTEP Advocates follow: > > A. You have only a budget of USD > $10 Million to accomplish everything with. > Wow... A US $10M budget! Cool...! 1) The first thing I would start doing with this is raise the visibility of NeXTSTEP in the OO print magazines. Did you know Dr. Dobbs and some of the OO journals don't even mention NS in most (if not all) of their issues? I would start touting NS as everything Smalltalk, the current most-favored-OO-language-of-the-moment, wishes it could be. 2) Start calling on companies with NS development teams to start doing print ad testimonials concerning the power of NS. And start comparing it to everything! I mean EVERYTHING!!!! Hammer the hell out of Visual C++ (that would-be OO pretender from Microsoft). 3) To extend 2) even further, host a honest-to-God head-to-head competition between MS programmers and some crack NS progrmmers (preferably (but is it allowable?) from NeXT. Give them something difficult to implement, like writing an API or something in each team's in housr development tool (MS would use Visual C++; NS programmers would use NS). After the NS team thoroughly kills the MS team, tout it in print ads across the country, including the big papers - New York Times, Wall Street Journal, and Los Angelas Times papers (with full page ads). Start capturing mind share.... 4) Offer to buy development licenses for students who are interested in owning a copy of NS. Open up a web site where any student who is interested would have to send an original copy of the purchase receipt for their OPENSTEP or NS product. Then simply buy a developer subscription for $2500 which would come with a developer's license from NeXT. Allow the kids to write their own apps and give them the ability to sell them! 5) Set up mailing list servers and hire development support personnel who have a ton of expereience with NEXTSTEP/OPENSTEP. Have them answer email and take technical support calls from all kinds of developers. Train them to tell everybody 16 times to go tell a friend about OPENSTEP/NEXTSTEP and give a demo. 6) Leep fighting for mindshare in the OO journals. Enter every single possible programming competition imaginable. Even advertise in CMP's technical magazines and Ziff-Davis's publications also. 7) Give John Dvorak a fresh demo of OPENSTEP on C|Net central. 8) Set up a repeat of campus consultants for college students to take advantage of. B. Must make measurable impact of > say 1-2 million new U.S. based end users within six months. In short, > the reason for concentrating on suring up U.S. based end users means that > foreign users get supported too owing to the increased success. > Would that help with goal B? > Obviously, and considering the proposal of giving away NEXTSTEP > academic, if one were to use all $10 M to buy X-# of academic copies at > about $200.00 per copy to give away to qualified folks, this doesn't > leave much money to advertise or get the word out. There would also be > things like shipping to consider. As this consortium would need to make > a deal with NeXT in the first place to purchase a large number of > academic copies of the software, it would also need to insure and verify > that the software is registered to people in academia. There would be > overhead incurred doing this. > Anyway, there seem to be a couple of factors or constraints that > were overlooked with the academic giveaway. By giving away the software, > would you be able to obtain a measurable marketable impact by this within > six months and how does one get a $15 M return on this investment in 3 years? > In any case, I hope everyone starts to get the idea when I ask > about details and for something that is plausible to accomplish the > described objective. > > Thanks To All - Hope this helps.... In the mean time, I'm going to get back to work on GNUstep. Sam Krishna Linux Inside Coming soon - OPENSTEP!
From: Laurent.Champciaux@emn.fr Newsgroups: comp.sys.next.programmer Subject: MultiThread and loadNibSection Date: 22 Oct 1996 08:17:26 GMT Organization: Ecole des Mines de Nantes Distribution: world Message-ID: <54hvum$khs@wfn.emn.fr> Keywords: multithread, nib file Hi, I am working on a multithreaded application. I use the MiscThreadedObject class to encapsulate my thread. It works fine except for one thing : As the thread I detach is running, the main thread (the one handling the main event loop) has difficulties to load a .nib file (using loadNibSection: owner: withNames: method). It crashes if the .nib has not been previously loaded (before running the MiscThread). This phenomenon is not systematic. I tried to suspend the thread before loadind any .nib and resume it after. It is better but the app still crashes sometimes. Has someone ever experienced such a thing ? Thanks in advance. Laurent. I use NEXTSTEP 3.3 on a SPARC5. -- ======================================================= Laurent Champciaux Departement Informatique - Ecole des Mines de Nantes 4, rue A.Kastler - BP 20722 - 44307 Nantes Cedex 03 Tel: 02 51 85 82 18 (B220) email: Laurent.Champciaux@emn.fr WWW: http://www.emn.fr/dept_info/perso/laurent/
Newsgroups: comp.sys.next.programmer Organization: Antigone Press gateway, San Francisco Return-Path: <cs@hal.kph.tuwien.ac.at> Message-ID: <9610220907.AA00356@zaphod> Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 4.0 v146.2) From: Christian Starkjohann <cs@hal.kph.tuwien.ac.at> Date: Tue, 22 Oct 96 11:07:38 +0200 Subject: Re: Developing 3.x applications on OS4.0 Cc: Scott Anguish <sanguish@digifix.com>, Earl Spillar <spillar@uwyo.edu> I wrote: > [stuff about scripts for dual developer setup deleted] > > If you have missed the posting containing the script and want a > copy, mail me. If there is interest I can place the scripts on an > ftp server. There was interest and I have put the package on our ftp server: ftp://hal.kph.tuwien.ac.at/pub/NeXT/tools/DualDeveloper.gnutar.gz Maybe it will also show up on the peanuts archive. Bye, Christian. -- Christian Starkjohann <cs@hal.kph.tuwien.ac.at> or <cs@ds1.kph.tuwien.ac.at>, finger for PGP Public Key. PGP fingerprint: DF FD 40 60 91 6A 14 1C CD 2C E9 07 38 AE CB 4E
From: dwy@mcny.com (David Young) Newsgroups: comp.sys.next.programmer Subject: Re: MultiThread and loadNibSection Date: 22 Oct 1996 13:58:48 GMT Organization: Media Connection of New York, Inc. Distribution: world Message-ID: <54ijuo$a6u@alice.walrus.com> References: <54hvum$khs@wfn.emn.fr> Laurent.Champciaux@emn.fr wrote: : I am working on a multithreaded application. I use the MiscThreadedObject : class to encapsulate my thread. It works fine except for one thing : As : the thread I detach is running, the main thread (the one handling the main : event loop) has difficulties to load a .nib file (using loadNibSection: : owner: withNames: method). It crashes if the .nib has not been previously : loaded (before running the MiscThread). This phenomenon is not systematic. AFAIK, the AppKit is not multithreaded. -- # david young: network engineer+oop developer # net: dwy@mcny.com, dwy@ace.net (NeXTmail ok) web: http://www.ace.net/ # vox: 212.686.3845 201.798.5217 fax: 212.686.3856
From: jmichel@imtn.dsccc.com (James Michel) Newsgroups: comp.sys.next.programmer Subject: NEWBIE: Dynamically created class instances? Date: 22 Oct 1996 14:40:07 GMT Organization: DSC Communications Corporation, Plano, Texas USA Message-ID: <54imc7$cr@camelot.dsccc.com> I am now beginning my study of Objective-C and have a question which the Objective-C material I have does not answer. How do I create instances of classes dynamically? If I have a NeXTSTEP application that uses an unknown amount of instances of a special Window class how can I create and keep a handle on the multiple instances? If I do something like id myWindow; myWindow = [SpecialWindow new]; and connect this method to a "new" button or menu item then I will get one new window instance, myWindow. What if the user needs another instance? Do I have to use pointers like in C to accomplish this? Thanks, James
From: mmalcolm crawford <m.crawford@shef.ac.uk> Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 22 Oct 1996 16:30:26 GMT Organization: University of Sheffield, UK Message-ID: <54isr2$47h@bignews.shef.ac.uk> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547gs6$131@news.Dortmund.Germany.EU.net> <548hm5$s42@news.onramp.net> <54fhhe$794@news.asu.edu> <326C6201.746302A@nol.net> In-Reply-To: <326C6201.746302A@nol.net> On 10/22/96, Sam Krishna wrote: > 3) To extend 2) even further, host a honest-to-God head-to-head > competition between MS programmers and some crack NS progrmmers > (preferably (but is it allowable?) from NeXT. Give them something > difficult to implement, like writing an API or something in each team's > in housr development tool (MS would use Visual C++; NS programmers would > use NS). After the NS team thoroughly kills the MS team, tout it in > print ads across the country, including the big papers - New York Times, > Wall Street Journal, and Los Angelas Times papers (with full page ads). > Start capturing mind share.... > I've still got my "NeXT vs. Sun: a world of difference" video... :-) Best wishes, mmalc. --
From: stefaanm@netego.com (Stefaan Meeuws) Newsgroups: comp.sys.next.programmer Subject: PS 3D rendering wrappers Date: Tue, 22 Oct 1996 12:20:07 -0800 Organization: 101NET, Inc. Message-ID: <stefaanm-2210961220080001@ppp021-sf1.sirius.com> I posted a message in the marketplace list, too. I'm looking to acquire 3D PostScript wrappers. Anyone have any? Thanks, Stefaan Meeuws stefaanm@101net.com (415) 546 1968 -- Stefaan Meeuws
From: uli@tallowcross.uni-frankfurt.de (Uli Zappe) Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 22 Oct 1996 12:52:34 GMT Organization: J. W. Goethe-Universitaet Frankfurt/Main Message-ID: <54ig2i$n0@tallowcross.uni-frankfurt.de> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547n5o$310@sun0.urz.uni-heidelberg.de> <sk68-0101040021590001@cu-dialup-0017.cit.cornell.edu> <54g4q6$2p@sun0.urz.uni-heidelberg.de> flight@mathi.uni-heidelberg.de (Gregor Hoffleit) wrote: > and let's hope that Steve's leaving NeXT > for good soon to start the next thing.. Why do you think we should hope that Steve's leaving? What do you think would happen then? 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: "Eric A. Dubiel" <eadubie@rs6000.cmp.ilstu.edu> Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: Wed, 23 Oct 1996 01:26:08 -0500 Organization: Illinois State University- Instructional Technology Services Message-ID: <326DB9B8.7B05@rs6000.cmp.ilstu.edu> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547gs6$131@news.Dortmund.Germany.EU.net> <548hm5$s42@news.onramp.net> <54fhhe$794@news.asu.edu> <326C6201.746302A@nol.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sam Krishna wrote: > 1) The first thing I would start doing with this is raise the visibility > of NeXTSTEP in the OO print magazines. Did you know Dr. Dobbs and some > of the OO journals don't even mention NS in most (if not all) of their > issues? I would start touting NS as everything Smalltalk, the current > most-favored-OO-language-of-the-moment, wishes it could be. Please check this link out, I think you will find it quite interesting. http://www.batech.com/~dekorte/Objective-C/objc.html Objective-C is billed as AN OBJECT-ORIENTED SUPERSET OF C MODELED AFTER SMALLTALK-80 (for those who haven't seen it) BTW: what does the -80 mean exactly???? -- Eric A. Dubiel; http://www.ilstu.edu/~eadubie mailto:eadubie@rs6000.cmp.ilstu.edu ASCII, MIME, SUN, NeXT, PGP Mail ytalk:eadubie@138.87.201.11 Instructional Technology Services- Illinois State University "Intelligence is the ultimate aphrodisiac." - Dr. Timothy Leary ALL VIEWS EXPRESSED REPRESENT MYSELF ONLY
From: "Eric A. Dubiel" <eadubie@rs6000.cmp.ilstu.edu> Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: Wed, 23 Oct 1996 01:36:24 -0500 Organization: Illinois State University- Instructional Technology Services Message-ID: <326DBC1F.7276@rs6000.cmp.ilstu.edu> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547n5o$310@sun0.urz.uni-heidelberg.de> <sk68-0101040021590001@cu-dialup-0017.cit.cornell.edu> <54g4q6$2p@sun0.urz.uni-heidelberg.de> <54ig2i$n0@tallowcross.uni-frankfurt.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Uli Zappe wrote: > flight@mathi.uni-heidelberg.de (Gregor Hoffleit) wrote: > > and let's hope that Steve's leaving NeXT > > for good soon to start the next thing.. > > Why do you think we should hope that Steve's leaving? What do you think would > happen then? My guess would be that he's going to work for Ne(X)Tscape...Because they're the company that COULD KILL Micro$loth! Nah, actually it'd have to be a new company, given Steve's love of leadership and inovating. He'd be too confined, I believe, in Netscape. What would be the NEXT BIG THING? 4-Dimensional Hyperspace?! It's coming!!! -- Eric A. Dubiel; http://www.ilstu.edu/~eadubie mailto:eadubie@rs6000.cmp.ilstu.edu ASCII, MIME, SUN, NeXT, PGP Mail ytalk:eadubie@138.87.201.11 Instructional Technology Services- Illinois State University "Intelligence is the ultimate aphrodisiac." - Dr. Timothy Leary ALL VIEWS EXPRESSED REPRESENT MYSELF ONLY
From: mmalcolm crawford <m.crawford@shef.ac.uk> Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 23 Oct 1996 09:45:05 GMT Organization: University of Sheffield, UK Message-ID: <54kpf1$rpb@bignews.shef.ac.uk> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547gs6$131@news.Dortmund.Germany.EU.net> <548hm5$s42@news.onramp.net> <54fhhe$794@news.asu.edu> <326C6201.746302A@nol.net> <326DB9B8.7B05@rs6000.cmp.ilstu.edu> In-Reply-To: <326DB9B8.7B05@rs6000.cmp.ilstu.edu> On 10/23/96, "Eric A. Dubiel" wrote: > Objective-C is billed as > AN OBJECT-ORIENTED SUPERSET OF C MODELED AFTER SMALLTALK-80 > > (for those who haven't seen it) > > BTW: what does the -80 mean exactly???? > It's a compacted version of a temporal numbering scheme which increments in units of approximately 365.25 days (on average) since the reputed birth of a religious leader in the Middle East; cf also Algol-68 and Fortran-90. This particular form of compaction in which only the tens and units are retained will cause significant problems in the near future when 100 is reached... such as who will ever admit to being a Cobol++-02 programmer?! :-) Best wishes, mmalc. --
From: ians@cam-ani.co.uk (Ian Stephenson) Newsgroups: comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: Wed, 23 Oct 1996 10:05:15 GMT Organization: Cambridge Animation Systems Ltd Sender: news@cam-ani.co.uk Message-ID: <Dzq40s.Fz9@cam-ani.co.uk> References: <54isr2$47h@bignews.shef.ac.uk> In article <54isr2$47h@bignews.shef.ac.uk> mmalcolm crawford <m.crawford@shef.ac.uk> writes: > I've still got my "NeXT vs. Sun: a world of difference" video... :-) Yeah, but can you watch it? True to their usual standards of reaching people, NeXT marketing mailed out NTSC versions to all their UK contacts! (I've still got mine, but I only every saw it once) $an
From: rog@ohm.york.ac.uk (Roger Peppe) Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS marke Date: 23 Oct 1996 09:35:02 GMT Organization: Department of Electronics, University of York, UK. Message-ID: <54kos6$5r4@netty.york.ac.uk> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547n5o$310@sun0.urz.uni-heidelberg.de> <01bbbdd5$69724200$5b1018ce@barrington> On 19 Oct 1996 15:56:04 GMT, "Chris Trimble" <trimble@walrus.com> wrote: > Actually, I'd say the fight has moved one level LOWER. Those > cross-platform projects you're talking about demonstrate very little > platform consistency and typically have weak integration with any one OS. > The OS wars are still interesting, and the reason why is because Java or > GNU-type portable apps can't not make a dent in the general application > development market for some time to come... probably never. They simply > do not offer the consistency and features that the users in the world > demand. there is one that offers the potential... check out : http://www.lucent.com/inferno/ as ground breaking a platform as the first Unix was when that first appeared, IMO. rog. (no personal interest, except that i think this product is the only thing out there that really *deserves* hype!)
From: information@ix.netcom.com Newsgroups: comp.sys.next.programmer Subject: Lottery Information Date: Wed, 23 Oct 1996 03:57:37 Organization: Netcom Message-ID: <54kqhb$prj@sjx-ixn10.ix.netcom.com> To get information on one of the best lotteries in the world, go to...... http://www.interlotto.li/cgi/hserver.exe?mainlot+serve=_mn_main+urlref=3300000511816 for other information, go to.... http://www.interllotto.com
From: xms@dcs.shef.ac.uk (Xiao Mang Shou) Newsgroups: comp.sys.next.programmer Subject: Matrix class's sizeToCell method Date: 23 Oct 1996 11:32:34 GMT Organization: Department of Computer Science, University of Sheffield Message-ID: <54kvoi$1cq@bignews.shef.ac.uk> Hello I am interested to learn more about Matrix class. Currently, I have made a class called myMatrix which is a subclass of Matrix. I have implemented a method called syn in a controller class to handle the grow and shrink of the matrix Implementation : myMatrix - initFrame: (const NXrect *) frameRect { ........... [self setBackground : NX_DKGRAY]; ..... } Implementation : Controller id matrix // id points to the myMatrix class instance - syn : (int) numberOfCell { [matrix renewRow: 1 cols : numberOfCell]; [matrix sizeToCell]; [Matrix display]; return self; } When the number of cells grow, the above code seems OK, however, if the number of cells decreases, the cell frame of those unwanted cells are still left on the winodw, (unwanted dark gray cell frames still there). The only way I can think of solving this problem is to replace [matrix display] with [window display] message, but this seems to be overkill. Can anyone suggest me a better way to solve this problem ? Thank. XM Shou
From: anch@logiball.de (Andreas Christiani) Newsgroups: comp.sys.next.programmer Subject: Re: NEWBIE: Dynamically created class instances? Date: 23 Oct 1996 11:55:24 GMT Organization: Customer of EUnet Germany; Info: info@Germany.EU.net Message-ID: <54l13c$asi@news.Dortmund.Germany.EU.net> References: <54imc7$cr@camelot.dsccc.com> jmichel@imtn.dsccc.com (James Michel) wrote: >I am now beginning my study of Objective-C and have a question which >the Objective-C material I have does not answer. How do I create >instances of classes dynamically? If I have a NeXTSTEP application >that uses an unknown amount of instances of a special Window class >how can I create and keep a handle on the multiple instances? If I do >something like > > id myWindow; > > myWindow = [SpecialWindow new]; > >and connect this method to a "new" button or menu item then I will get >one new window instance, myWindow. What if the user needs another instance? >Do I have to use pointers like in C to accomplish this? > >Thanks, > >James Maintain a list of your instances in an NSArray for example. But to give you an exact advice, your description was a little too short ;-) Ciao, A.C. --- ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ LogiBall gGmbH * Innovationszentrum Herne * Westring 303 * 44629 Herne Andreas Christiani * christiani@logiball.de * http://www.logiball.de Tel.: 02323 / 925 577 * Fax : 02323 / 925 551 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
From: kjt@cs.stir.ac.uk (Ken Turner) Newsgroups: comp.sys.next.programmer Subject: Re: OCR Available Date: Wed, 23 Oct 1996 11:56:29 GMT Organization: University of Stirling Message-ID: <961023125629.13590AAFCi.kjt@copper> References: <546p56$t31@newsbf02.news.aol.com> Mime-Version: 1.0 (Generated by Eloquent) Content-Type: text/plain; charset=US-ASCII There's a few OCR packages. I have extraRead. Ken Turner (kjt@cs.stir.ac.uk)
From: devan2m@imap2.asu.edu Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Followup-To: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Date: 23 Oct 1996 12:37:17 GMT Organization: Arizona State University Message-ID: <54l3ht$aom@news.asu.edu> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547gs6$131@news.Dortmund.Germany.EU.net> <548hm5$s42@news.onramp.net> <54fhhe$794@news.asu.edu> <326C6201.746302A@nol.net> Your input is greatly appreciated. I am glad that you found this provocative enough to lend your reactions and thoughts. Before taking up the other issues about a consortium of advocates marketing masterplan for NEXTSTEP, please don't misunderstand the comments of my previous post where GNUstep are concerned. I wasn't disregarding or writing GNUstep out of the equation. But I must admit that I am not quite sure what equation you are referring to. From the context I assume you mean the OS market? Feel free to correct me if I am wrong. I have heard discussions about GNUstep for many years now and still am not overly familiar with it. My view is that the GNUstep work would actually stand to gain by this marketing coup being discussed just as NEXTSTEP would benefit by having this project publicly available. I know that I have used and appreciate the compression utilities for many years now. I do know, however, that NEXTSTEP is here and a fully-formed, exceptional, commercially viable, industrial grade software product which is still, and to borrow from a few posts which have contributed to this discussion: "NeXTSTEP is probably the most respected software in the planet." Sung Ho Kim - sk68@cornell.edu "In a market full or market-share mongrels, it tried to produce something that was no-compromise, first-rate and innovative. I don't think anybody can doubt that they did in fact produce something special. I think this is what makes NeXT special." Sung Ho Kim - sk68@cornell.edu "As someone who got to admire the elegance and conceptual clarity of NeXTStep ..." Alexander Wilkie - wilkie@cg.tuwien.ac.at Despite having by far the best product in our market, ... Art Isbell - aisbell@ix.netcom.com Allow me to next noting that the itemized list you provided is extremely appreciated. But before addressing these items, permit me to speak about the money issue. :Wow... A US $10M budget! Cool...! Having $10 M to work with and needing to return the $10 M principle plus $5 M in three years really isn't that much once you start planning and pricing out to sustain all of the things you mentioned or making a serious and committed run at this marketing coup being discussed. Moreover, and having the performance constraint or condition imposed as was mentioned of having to produce some measurable and clear impact within about six month of having started a campaign - actually let's say that there is a total of nine months from November in which to achieve this - affairs would have to be handled very shrewdly. ______________________ 1) The first thing I would start doing with this is raise the visibility of NeXTSTEP in the OO print magazines. Did you know Dr. Dobbs and some of the OO journals don't even mention NS in most (if not all) of their issues? I would start touting NS as everything Smalltalk, the current most-favored-OO-language-of-the-moment, wishes it could be. 2) Start calling on companies with NS development teams to start doing print ad testimonials concerning the power of NS. And start comparing it to everything! I mean EVERYTHING!!!! Hammer the hell out of Visual C++ (that would-be OO pretender from Microsoft). 3) To extend 2) even further, host a honest-to-God head-to-head competition between MS programmers and some crack NS programmers (preferably (but is it allowable?) from NeXT. Give them something difficult to implement, like writing an API or something in each team's in hours development tool (MS would use Visual C++; NS programmers would use NS). After the NS team thoroughly kills the MS team, tout it in print ads across the country, including the big papers - New York Times, Wall Street Journal, and Los Angeles Times papers (with full page ads). Start capturing mind share.... ______________________ Before engaging a war and going into what promises to be a fierce gauntlet, one must have all their forces and weapons prepared. At the point where the challenge is issued and this Consortium of Renegade NEXTSTEP Advocates goes on the offensive to capture mind share, everything had better be in place to handle the onslaught which will ensue. This independent entity would have be fairly well insulated and comprised of well-heeled individuals to handle and repel all reprisals. The key in all this is not to go into to it thinking that it can be done in a one-punch or lucky-punch artist fashion. My belief is that Steve Jobs made a significantly crucial error when launching NeXT by not anticipating or properly responding to was coming under attack as NeXT did from all sides. Never fall for the lie that you here about it not being personal and just business. IT IS ALWAYS PERSONAL! In my opinion, and based on things that I saw and heard, many companies saw what Jobs was doing as a bona fide threat and wasn't about to let him repeat the magic they had known him for called Apple. Jobs thought that people could put there jealousy, insecurity, and vested interests aside in favor of something INSANELY GREAT and innovative and get behind him as they did before. They were laying and waiting for him this time and after the grandeur and hoopla of the September 1988 unveiling, everybody could then take clear aim. (Upon reading the last sentence, please don't deter from the objective at hand by raising questions and arguments about what constitutes a conspiracy.) Increasing the visibility of NEXTSTEP in the fashion you suggest might have to be done as a lone proposition for this Consortium until there would be significantly favorable grassroots and high-profile support generated. Getting testimonials at present from the development teams of companies might expose them a little more than they wish to be and being associated with this may open them to battles which they are not prepared to take on. There used to be an expression I would heard about IBM and think that people are now applying it to Microsoft and Windoze. "Nobody ever got fired for recommending IBM." These companies may wish to, but to go out of their way at the present time or even to allow us to use their comments to badger other platforms in the current environment might be a little more risky than some companies are willing to be. What's more, and despite the intentions of the Consortium being clear, lending one's seal of approval to an entity and this objective that sort of raises out of nowhere (actually the ashes from what many have been convinced to believe about NeXT and NEXTSTEP), isn't something many stable players would do. I would welcome being wrong about this of course. Before doing any of this, there is the question of assembling nothing less than a crack core and team. This obviously is not something that can be pulled off with only 10-12 people. Obtaining the right people and getting them with the program is where some of the initial outlay and a sizable chunk of the $10 M would go to for starters. Take yourself instance, how much would it take for you to place what you are currently doing on hiatus to take part in such a project as this that would probably take a total of nine months (three to gather forces, prep, and coordinate the execution and six months to achieve the first measurable objective of 1-2 M end users or something similar and with equal impact)? If successful, it would continue, but there would also have to be an infusion of more capital through earnings which would keep it viable and then these become standard operating costs. What is the figure you have in mind? Take this figure and multiply it by the number of people you think it would take to pull this off and support it on the scale that it would have to be. Now, care to tell me how much you get and how many people are you projecting? Would you be part of the development support personnel and/or lend support to the aggressive but crafty advertising/marketing/distribution/PR/development/support/operations brain trust of this band or Consortium of Renegade Advocates? In addition to assembling a core force of say about 30 die-hard, hard working, and creatively clever individuals who would serve to anchor a force of maybe 75-200 individuals to see this project through, (at least that's how many I would project), these individuals will have to be equipped with state-of-the-art and perhaps custom high-end hardware tools. This adds to the cost. Likewise, there needs to be a significant communication's budget and infrastructure to handle the anticipated load which this would require. This would include a lot of cellular telephone, pagers, and laptops to be obtained and maintained. In order to get this many folks with the right desire and/or NEXTSTEP experience and who would be available for this, may require recruiting fellow advocates from around the world to help focus on this marketing coup directed at the U.S. market. Bringing in and/or relocating at the very least 10-20 well-trained and experienced NEXTSTEP developers and adequately compensating them to earn near what they would otherwise, I would say could cost as much as $100 K per person including benefits. Even if these talents believed in the cause and were willing to work for peanuts in the interest of a higher long-run return, giving them a decent base salary along with the proper tools, environment, and support would still run about $100 K each. So let's say that we seattle for only ten of these hand-picked developers and drawn upon the support that NeXT would have to naturally provide for support of NEXTSTEP, this would run about $1 M for the 6-9 month period. The cost of assembling, preparing, and maintaining a solid base and complete team of renegades to operate at the level that is required would run about $3 M I would say off the top of my head. Now let's return back to the issue of marketing various publications. I don't know how familiar you are with the price of running ads in such publications, but to consistently bombard all the major magazines and national publications such as the Wall Street Journal or New York Times and obtaining fairly prime spots as you suggest even for a six month period could costs as much as $2-3 million. It is a good thing you didn't mention trying to do a national television campaign as well seeing as major television spots can cost as much as $500 K a pop and we haven't even talked about the cost of producing a decent spot even if we were to do it virtually using NEXTSTEP on high-end Pentiums and SPARCstations. Equipping this Consortium or force from top to bottom with all the proper tools could run as little as $200 K and as much as $1 M I would say. If this marketing campaign is very effective, this would mean receiving a heavy volume of interested responses. You should also keep in mind that there would be those skunks and saboteurs responding too and there would be additional overhead having to deal with them. Having about 200 personnel is beginning to sound not too far off with having to provide support. The support at this level and in addition to what is mentioned for the core team above would run about 2.5 million assuming that these were truly quality NEXTSTEP knowledgeable and extremely computer literate folks which would reflect best and make what is being done even more successful. Is this support going to be provided via a toll-free line? I would have to get a deal where this cost is very small on a per minute (six second increment basis). Five pages and less than about $500 K left and we haven't even gotten to the cost of arranging a head-to-head competition which I think is an excellent idea or a repeat of campus consultants as you suggest. To do the competition halfway right and attract the audience and national attention desired, this would require about $1 M as I envision it. We also haven't discussed how income is to be generated for this entity. Given the nature of this task and what it sets out to achieve in the time that it does, having only USD $10 M to work with in my opinion is not a lot of money and would require an extremely creative, shrewd, and coordinated team to execute a comprehensive plan. I am greedy and welcome further comments so please don't be shy. Sam Krishna (infinity@nol.net) wrote: : Well, : This is interesting.... : As a Linux/NT/NeXTSTEP user who gets to play around in all kinds of : environments... I would have to say that by far, the best OS I've seen : to date is NeXTSTEP. : Here's some ideas on marketing it.... : >devan2m@imap2.asu.edu wrote: : > : > Hello Steve, Darren, Felipe, & any others who are following or care to : > contribute: : > : > Although Steve makes an interesting suggestions, to give away : > NEXTSTEP/OPENSTEP academic for free, it is a proposal that really ignores : > the parameters of the marketing challenge as presented. What's more, and : > rather than head down a path which examines what NeXT and their marketing : > folks should and shouldn't do, there is an entirely different direction : > that I would like to see this discussion take. This directions is one : > which is governed by the following constraints and hopefully will offer : > some novel insights as to what can actually be done to fortify and : > radically increase the market presence of NEXTSTEP 3.3 & an OPENSTEP for : > MACH which is more in-line with the evolution of NEXTSTEP as it should be : > on course of Mecca and not as some Windoze/Solaris looking chameleon or : > derivative. : > : > Recall that the original marketing challenge was presented as follows: : > : > (Excerpt...) : > : > 1) Eliminate NeXT Software, Inc. from the equation : > and any marketing efforts by NeXT in order to accomplish these : > objectives. Think of this entire operation as something being undertaken : > by an entity independent of NeXT ( a consortium of advocates) acting on : > behalf of the NEXTSTEP community and the technology - please don't : > confuse this with the GNUStep efforts. : > : I like this - they are too unreliable anyway... I wouldn't write : GNUstep out of the equation just yet, though, but we deal with things : the way they are defined, not the way we would like them... : > 2) Key objectives on the list of this Consortium : > of NEXTSTEP Advocates follow: : > : > A. You have only a budget of USD : > $10 Million to accomplish everything with. : > : Wow... A US $10M budget! Cool...! : 1) The first thing I would start doing with this is raise the visibility : of NeXTSTEP in the OO print magazines. Did you know Dr. Dobbs and some : of the OO journals don't even mention NS in most (if not all) of their : issues? I would start touting NS as everything Smalltalk, the current : most-favored-OO-language-of-the-moment, wishes it could be. : 2) Start calling on companies with NS development teams to start doing : print ad testimonials concerning the power of NS. And start comparing : it to everything! I mean EVERYTHING!!!! Hammer the hell out of Visual : C++ (that would-be OO pretender from Microsoft). : 3) To extend 2) even further, host a honest-to-God head-to-head : competition between MS programmers and some crack NS progrmmers : (preferably (but is it allowable?) from NeXT. Give them something : difficult to implement, like writing an API or something in each team's : in housr development tool (MS would use Visual C++; NS programmers would : use NS). After the NS team thoroughly kills the MS team, tout it in : print ads across the country, including the big papers - New York Times, : Wall Street Journal, and Los Angelas Times papers (with full page ads). : Start capturing mind share.... : 4) Offer to buy development licenses for students who are interested in : owning a copy of NS. Open up a web site where any student who is : interested would have to send an original copy of the purchase receipt : for their OPENSTEP or NS product. Then simply buy a developer : subscription for $2500 which would come with a developer's license from : NeXT. Allow the kids to write their own apps and give them the ability : to sell them! : 5) Set up mailing list servers and hire development support personnel : who have a ton of expereience with NEXTSTEP/OPENSTEP. Have them answer : email and take technical support calls from all kinds of developers. : Train them to tell everybody 16 times to go tell a friend about : OPENSTEP/NEXTSTEP and give a demo. : 6) Leep fighting for mindshare in the OO journals. Enter every single : possible programming competition imaginable. Even advertise in CMP's : technical magazines and Ziff-Davis's publications also. : 7) Give John Dvorak a fresh demo of OPENSTEP on C|Net central. : 8) Set up a repeat of campus consultants for college students to take : advantage of. : B. Must make measurable impact of : > say 1-2 million new U.S. based end users within six months. In short, : > the reason for concentrating on suring up U.S. based end users means that : > foreign users get supported too owing to the increased success. : > : Would that help with goal B? : : > Obviously, and considering the proposal of giving away NEXTSTEP : > academic, if one were to use all $10 M to buy X-# of academic copies at : > about $200.00 per copy to give away to qualified folks, this doesn't : > leave much money to advertise or get the word out. There would also be : > things like shipping to consider. As this consortium would need to make : > a deal with NeXT in the first place to purchase a large number of : > academic copies of the software, it would also need to insure and verify : > that the software is registered to people in academia. There would be : > overhead incurred doing this. : > Anyway, there seem to be a couple of factors or constraints that : > were overlooked with the academic giveaway. By giving away the software, : > would you be able to obtain a measurable marketable impact by this within : > six months and how does one get a $15 M return on this investment in 3 years? : > In any case, I hope everyone starts to get the idea when I ask : > about details and for something that is plausible to accomplish the : > described objective. : > : > Thanks To All - : Hope this helps.... In the mean time, I'm going to get back to work on : GNUstep. : Sam Krishna : Linux Inside : Coming soon - OPENSTEP!
Newsgroups: comp.sys.next.programmer From: information@ix.netcom.com Subject: cmsg cancel <54kqhb$prj@sjx-ixn10.ix.netcom.com> Control: cancel <54kqhb$prj@sjx-ixn10.ix.netcom.com> Message-ID: <cancel.54kqhb$prj@sjx-ixn10.ix.netcom.com> Followup-to: junk References: <54kqhb$prj@sjx-ixn10.ix.netcom.com> Date: Wed, 23 Oct 1996 03:57:37 Spam-cancel: "Lottery Information"
From: shess@shell.one.net (Scott Hess) Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 23 Oct 1996 09:36:14 -0400 Organization: Is a sign of weakness Sender: shess@shell.one.net Message-ID: <veeohhtpzu9.fsf@shell.one.net> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547n5o$310@sun0.urz.uni-heidelberg.de> <sk68-0101040021590001@cu-dialup-0017.cit.cornell.edu> <54g4q6$2p@sun0.urz.uni-heidelberg.de> <54ig2i$n0@tallowcross.uni-frankfurt.de> In-reply-to: uli@tallowcross.uni-frankfurt.de's message of 22 Oct 1996 12:52:34 GMT In article <54ig2i$n0@tallowcross.uni-frankfurt.de>, uli@tallowcross.uni-frankfurt.de (Uli Zappe) writes: flight@mathi.uni-heidelberg.de (Gregor Hoffleit) wrote: > and let's hope that Steve's leaving NeXT > for good soon to start the next thing.. Why do you think we should hope that Steve's leaving? What do you think would happen then? My theory is that people under the Apple tent made _much_ more money for the five or six years after Jobs left Apple than they did before. Of course, now they are somewhat in the dumps - but is it really any worse than the NeXT market? Of course not. On the other hand, it's much too late, now. I think there was a time when Jobs could have turned the company over to the MBA's, who would have cleaned house and concentrated on "making money" and not "being right". The OS would have stagnated, and the community would have gotten smaller and cynical, but the market could have made money quite handily for three or four years. Of course, things have basically worked out that way, except for the making money part. I would _much_ rather have Apple's problem of having neglected the OS and other whizzy for too long, and now having to warp space to make things work for Copland, rather than have the most whizzy OS but nobody wants to run it ... Later, -- scott hess <shess@one.net> http://www.winternet.com/~shess/ Work: 288 Rampart Court #105, Ft Mitchell, KY, 41017 (606) 578-0412 (Was) 12550 Portland Avenue South #121, Burnsville, MN 55337 (612)895-1208
From: Charles William Swiger <cs4w+@andrew.cmu.edu> Newsgroups: comp.sys.next.advocacy,comp.sys.next.programmer,comp.sys.next.misc Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: Wed, 23 Oct 1996 12:34:51 -0400 Organization: Carnegie Mellon, Pittsburgh, PA Message-ID: <0mPYYfe00UhB01gnhs@andrew.cmu.edu> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547gs6$131@news.Dortmund.Germany.EU.net> <548hm5$s42@news.onramp.net> <54fhhe$794@news.asu.edu> <326C6201.746302A@nol.net> <326DB9B8.7B05@rs6000.cmp.ilstu.edu> In-Reply-To: <326DB9B8.7B05@rs6000.cmp.ilstu.edu> Excerpts from netnews.comp.sys.next.advocacy: 23-Oct-96 Re: Open Letter to NeXT: W.. by "Eric A. Dubiel"@rs6000. > http://www.batech.com/~dekorte/Objective-C/objc.html > > Objective-C is billed as > AN OBJECT-ORIENTED SUPERSET OF C MODELED AFTER SMALLTALK-80 > > (for those who haven't seen it) That's a pretty good description of Objective-C, IMHO. > BTW: what does the -80 mean exactly???? It refers to the year 1980, which should be commonly known reference standard for that language. You'll see the same system of designation used (for example) with Fortran-77 and -90. -Chuck Charles Swiger | cs4w@andrew.cmu.edu | standard disclaimer ----------------+---------------------+--------------------- I know you're an optimist if you think I'm a pessimist.
From: dekorte@suite.com (Steve Dekorte) Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 23 Oct 1996 17:49:32 GMT Organization: Suite Software Message-ID: <54llrc$mq7@news.onramp.net> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547gs6$131@news.Dortmund.Germany.EU.net> <548hm5$s42@news.onramp.net> <54fhhe$794@news.asu.edu> <326C6201.746302A@nol.net> <54l3ht$aom@news.asu.edu> Cc: devan2m@imap2.asu.edu devan2m@imap2.asu.edu wrote: > "NeXTSTEP is probably the most respected software on the planet." Sung > Ho Kim - sk68@cornell.edu I believe he was quoting a 1990(?) issue _Byte_ magazine. Steve
From: woo@opus.bloomco.ornl.gov (John W. Wooten) Newsgroups: comp.sys.next.programmer Subject: JAVA vs. Objective-C Date: 23 Oct 1996 20:07:26 GMT Organization: Oak Ridge National Lab, Oak Ridge, TN Message-ID: <54lttu$ctn@stc06.ctd.ornl.gov> Would someone more knowledgeable than myself please explain what makes JAVA better (or worse) than Objective-C? I understand that the java compiler emits byte-codes that are then portable and run on any java virtual computer, but theoretically Objective-C could be built to emit byte-codes and have a virtual computer that ran on various machines with byte-coded libraries or run times to load dynamically. I note that JAVA doesn't allow keywords in method arguments, and apparently any public instance variable is available for manipulation via the class.variable = value assignment rather than having to use a method to set instance variables and get their values. I see that the JAVA method could cause some problems unless you're very careful with private, etc. access. For instance, I could have a variable where I maintain the current x-velocity of some object. I may choose (for all kinds of reasons) to store the x-velocity in units of furlongs/fortnight internally. I have methods to set the velocity that do the necessary conversions before storing the value in the x-velocity variable. I also could provide methods for retrieving the value as: - (float) xVelInMpS; - (float) xVelInMipHr; and so forth. In the case of JAVA, however it appears that unless i declare it private, then anyone can set it or fetch it without using a method. Does this mean that almost all JAVA instance variables need to be private unless you're darn sure what others will do with them? Any other comparisons, contrasts, etc. would be appreciated. thanks, John W. Wooten -- - - - - - - - - - J. W. Wooten
From: pemmerik@solair1.inter.NL.net (P.J.L.van Emmerik) Newsgroups: comp.sys.next.programmer Subject: Help wanted with DPS error messages Date: Wed, 23 Oct 1996 20:57:41 GMT Organization: NLnet Message-ID: <54m0p8$mvb@news.NL.net> In the Console window i get the following error messages: Oct 23 22:01:31 dbe_s Reistijden_Server[947]: received reply message from port 10496 via dps Oct 23 17:29:48 ea_next01 BusStation[486]: DPS client library error: PostScript program error DPSContext 45bad8 Oct 23 17:29:48 ea_next01 BusStation[486]: %%[ Error: undefined; OffendingCommand: user781 ]%% Any pointers on how to track the offending code? P.J.L. van Emmerik Holec Projects B.V. Email: pemmerik@solair1.inter.NL.net PO.BOX 565, 7550 AN Hengelo Phone: +31 74 2558 688 The Netherlands --
From: pemmerik@solair1.inter.NL.net (P.J.L.van Emmerik) Newsgroups: comp.sys.next.programmer Subject: Help wanted with DPS errors Date: Wed, 23 Oct 1996 21:00:43 GMT Organization: NLnet Message-ID: <54m0uu$mvk@news.NL.net> In the Console window i get the following error messages Oct 23 22:01:31 dbe_s TestApp1[947]: received reply message from port 10496 via dps Oct 23 17:29:48 ea_next01 TestApp2[486]: DPS client library error: PostScript program error DPSContext 45bad8 Oct 23 17:29:48 ea_next01 TestApp2[486]: %%[ Error: undefined; OffendingCommand: user781 ]%% Any pointers on how to track down the offenders? P.J.L. van Emmerik Holec Projects B.V. Email: pemmerik@solair1.inter.NL.net PO.BOX 565, 7550 AN Hengelo Phone: +31 74 2558 688 The Netherlands --
From: npratt@xmission.xmission.com (Nevin Pratt) Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 23 Oct 1996 15:13:11 -0600 Organization: XMission Internet (801 539 0900) Message-ID: <54m1p7$pca@xmission.xmission.com> References: <01bbbc9e$250b8ff0$761018ce@barrington> <326C6201.746302A@nol.net> <54l3ht$aom@news.asu.edu> <54llrc$mq7@news.onramp.net> In article <54llrc$mq7@news.onramp.net>, Steve Dekorte <dekorte@suite.com> wrote: >devan2m@imap2.asu.edu wrote: >> "NeXTSTEP is probably the most respected software on the planet." >Sung >> Ho Kim - sk68@cornell.edu > >I believe he was quoting a 1990(?) issue _Byte_ magazine. > >Steve > It was a summer issue in 1992. June, I believe. Nevin
From: scottc@nacs.net (Scott Christley) Newsgroups: comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 23 Oct 1996 20:35:12 GMT Organization: New Age Consulting Service, Cleveland, OH, USA Message-ID: <54lvi0$t71@tracy.nacs.net> References: <54isr2$47h@bignews.shef.ac.uk> <Dzq40s.Fz9@cam-ani.co.uk> I just noticed that NeXT's GDB guy's email address changed and was wondering... That's interesting. I can say that he's done some great work with GDB to get it to understand the GNU runtime for Objective-C; I can say it has helped me immensely with GNUstep work. It's great to see him working for Cygnus now. Scott John Rudd (jrudd@cygnus.com) wrote: :Well, I don't know if I should be posting this or not.. In some ways it's :good news for the Gnustep people, and in some ways it's bad news for the :Nextstep and Openstep people. :At Cygnus we had a new GDB developer start this week. He came to us from :NeXT. One of the things he said he hopes to do is bring some of the :features of NeXT's GDB to other GDB platforms. I asked me why he didn't :stay at NeXT, and his response sort of confirmed my worst fears.. "They :say they're going to become a 'Web company'. They don't know what that :means yet, but that's what they're going to do." :Losing GDB people not only spells bad news for the Nextstep/OS-M :community, but also for the general Openstep from NeXT community. It :looks like they're not just moving away from the OS market, but away from :the developers tools market as well. And worst of all, they don't even :know what they're moving toward. :I guess we'll just have to see what Afterstep and Gnustep will bring us.. :This week I've started taking a closer look at netbsd, BSD-lites, and :MkLinux. I somehow get the feeling that's where my future lies. It :would be nice if we could get some of those large commercial investors :in Nextstep to put up money to get NeXT to release the source to their :OS, or something.. But I don't see that happening.
From: root@starbase1.gpl.net (Dale Amon as Operator) Newsgroups: comp.sys.next.programmer Subject: Where is DEV_BSIZE defined for NeXT 3.2? Date: 23 Oct 1996 23:32:00 GMT Organization: Genesis Project Ltd Message-ID: <54m9tg$gbm@divis.gpl.net> I'm porting snmp and I've run across a problem that DEV_BSIZE doesn't get defined. In fact, it appears like it is explicitely NOT defined if the machine type is NeXT. It greps from the following locations: ------- bsd/nfs/nfs.h:#define NFS_BLOCKSIZE 1024 /* no more DEV_BSIZE */ bsd/protocols/dumprestore.h: * Note that TP_BSIZE must be a multiple of DEV_BSIZE. bsd/sys/dir.h: * NeXT: increased DEV_BSIZE from 512 to 1024. bsd/sys/dir.h:#if !defined(KERNEL) && !defined(DEV_BSIZE) && defined(_NEXT_SOURCE) bsd/sys/dir.h: #define DEV_BSIZE 512 bsd/sys/dir.h:#endif /* !KERNEL && !DEV_BSIZE && _NEXT_SOURCE */ bsd/sys/dir.h: #define DIRBLKSIZ DEV_BSIZE bsd/sys/param.h: * Note that the blocked devices are assumed to have DEV_BSIZE bsd/sys/param.h: * DEV_BSIZE <= BLKDEV_IOSIZE <= MAXBSIZE bsd/sys/param.h:#define DEV_BSIZE 512 bsd/sys/param.h:#define DEV_BSHIFT 9 /* log2(DEV_BSIZE) */ bsd/sys/param.h:#define btodb(bytes) /* calculates (bytes / DEV_BSIZE) */ \ bsd/sys/param.h:#define dbtob(db) /* calculates (db * DEV_BSIZE) */ \ bsd/sys/param.h: * just use DEV_BSIZE. bsd/sys/param.h:#define bdbtofsb(bn) ((bn) / (BLKDEV_IOSIZE/DEV_BSIZE)) bsd/ufs/fs.h:#define SBLOCK ((daddr_t)(BBLOCK + BBSIZE) / DEV_BSIZE) bsd/ufs/fs.h: * addressible; these pieces may be DEV_BSIZE, or some multiple of bsd/ufs/fs.h: * a DEV_BSIZE unit. bsd/ufs/fs.h: * Number of disk sectors per block; assumes DEV_BSIZE byte sector size. ------- The places where it is defined to 512 are in conditionals that say not NeXT. Does anyone have a clear idea how some of these bits of the Mach/NeXT fs fit together? In particular, for snmp I have to make the following line work: lseek(file, (long) (SBLOCK * DEV_BSIZE), 0); I'm working with both black and white 3.2.
From: suckow@bln.sel.alcatel.de (Ralf Suckow) Newsgroups: comp.sys.next.programmer Subject: Re: JAVA vs. Objective-C Date: 24 Oct 1996 08:37:07 GMT Organization: Alcatel/Bell Distribution: world Message-ID: <54n9rj$i21@btmpjg.god.bel.alcatel.be> References: <54lttu$ctn@stc06.ctd.ornl.gov> John W. Wooten writes > Would someone more knowledgeable than myself please explain what makes > JAVA better (or worse) than Objective-C? I understand that the java > compiler emits byte-codes that are then portable and run on any java > virtual computer, but theoretically Objective-C could be built to emit > byte-codes and have a virtual computer that ran on various machines with > byte-coded libraries or run times to load dynamically. > > I note that JAVA doesn't allow keywords in method arguments, and > apparently any public instance variable is available for manipulation via > the class.variable = value assignment rather than having to use a method > to set instance variables and get their values. ... There's no significant difference in the instance variables protection between Java and Objective-C. In Java you have not only private, but also friendly, public, and protected, which should be sufficient for all purposes. BTW., a formal protection mechanism is never sufficient, you always also need PROGRAMMING DISCIPLINE. Accessing a variable directly when there is a method is bad style anyway and serious programmers don't do that (except when it is explicitely foreseen to do so, with public static final variables for example). And who cares about the non-serious programmers. Have you ever heard somebody saying: "As a programmer, I'm just kidding" :-) Yours, ------------------------ Ralf.Suckow@bln.sel.alcatel.de | All opinions are mine.
From: 1kyriaki@swt1.informatik.uni-hamburg.de (Valentino Kyriakides) Newsgroups: comp.sys.next.programmer Subject: Ignored TABs in IB ScrollView... Date: 24 Oct 1996 13:04:31 GMT Organization: University of Hamburg -- Germany Message-ID: <54npgv$l5i@rzsun02.rrz.uni-hamburg.de> I realized the following problem with IB connections and TABSTOPS (maybe it's an old known problem which I missed the answer for?), here is a description: Suppose I make a connection from a Controller instance to an ScrollView outlet (the outer rectangle) in InterfaceBuilder. When I try to print some formated text like: [[ScrollView docView] printf:"\t Whatever \t text comes here\n"]; the TABs are ignored in an application. If I instead make in IB a connection to the ScrollView's text outlet (the innermost rectangle, which is the embedded Text obj) and send a printf message to it like [scrollViewsText printf:"\t Whatever \t text comes here\n"]; the TABs are well supported (BTW, other printf format stuff works well in the above shown examples, so it seems that only TABs are handled differently.) If I manually code a ScrollView with an embedded Text view, thinks like: [[ScrollView docView] printf:"\t Whatever \t text comes here\n"]; work quite well! So my outcoming questions are: Is it an IB ScrollView bug? If it's not a bug, how do I get the things managed with the first shown IB conection (Controller ----> ScrollView) ??? Valentino -- +----------------------------------------------------------------+ | Valentino Kyriakides Univerversitaet Hamburg | | 1kyriaki@informatik.uni-hamburg.de FB. Informatik, AB. SWT | | (NO NeXTmail here) | +----------------------------------------------------------------+ ANNOTATION: I used the following Text category for formated printf stuff... ... @interface Text(Text_Fmt); - (int) printf:(const char *) format, ...; @end ... @implementation Text(Text_Fmt) #define BUFFERSIZE 512 - (int) printf:(const char *) format, ... { int result, length = [self textLength]; char buffer[BUFFERSIZE]; va_list ap; va_start(ap, format); result = vsprintf(buffer, format, ap); va_end(ap); [[self setSel:length :length] replaceSel:buffer]; [[self scrollSelToVisible] display]; return result; } @end
From: drelson@oaktree.com (David Relson) Newsgroups: comp.sys.next.programmer Subject: GDB - Foundation class problems Date: Fri, 25 Oct 1996 00:12:03 GMT Organization: ICNET... Your Link To The Internet... +1.313.998.0090 Message-ID: <327205bb.441877056@news.ic.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Greetings all! I'm trying to debug an Objective-C application that makes extensive use of NSString objects. GDB appears to not like them. In some cases, it does not even list local variables of type NSString *. In the cases where the variables are listed, I am unable to view their values. Attempting to print [ nsString cString ] gives me nothing. What can I do to have GDB list all the variables? What can I do to have GDB show me the values of NSStrings? I'd also like to view the contents NSArrays and NSDictionaries. How can I do this? Please reply via e-mail. Thank you.
From: drelson@oaktree.com (David Relson) Newsgroups: comp.sys.next.programmer Subject: ... broken invariant Date: Thu, 24 Oct 1996 23:55:18 GMT Organization: ICNET... Your Link To The Internet... +1.313.998.0090 Message-ID: <326fff4e.440232192@news.ic.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Greetings all! I am a relative newcomer to NeXTSTEP and Objective-C (after many years of other operating systems, C, and Smalltalk) and am encountering some odd messages. I am developing a 7x24 tool which is multithreaded, has classes descended from Object and NSObject and uses a third party library which uses NSObject and NSAutoreleasePools extensively. The two odd messages that I get (at random times; some days, but not others) are: *** hashtable: count differs after rehashing; probably indicates a broken invariant: there are x and y such as isEqual(x, y) is TRUE but hash(x) != hash (y) *** maptable: count differs after rehashing; probably indicates a broken invariant: there are x and y such as isEqual(x, y) is TRUE but hash(x) != hash (y) What do these messages really mean? Could it relate to NSStrings (of different addresses but equal contents) being in a table which is being resized ( rehashed )? Do the messages matter?
From: drelson@oaktree.com (David Relson) Newsgroups: comp.sys.next.programmer Subject: NSDoubleFreeError problem Date: Fri, 25 Oct 1996 00:08:20 GMT Organization: ICNET... Your Link To The Internet... +1.313.998.0090 Message-ID: <32710491.441579584@news.ic.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Greetings all! I am a relative newcomer to NeXTSTEP and Objective-C (after many years of other operating systems, C, and Smalltalk) and am encountering some odd messages. I am developing a 7x24 tool which is multithreaded, has classes descended from Object and NSObject and uses a third party library which uses NSObject and NSAutoreleasePools extensively. Every so often, I a message about an object being autoreleased too many times. At a friend's suggestion, I have put a breakpoint at _NSDoubleFreeError(). This afternoon I hit the breakpoint, and dumped the stack. The last line of my code to execute is: recipients = [ NSMutableArray array ]; and the stack levels from there to _NSDoubleFreeError() are: #0 0x7808b40 in _NSDoubleFreeError () #1 0x7809540 in -[NSAutoreleasePool addObject:] () #2 0x78095f6 in +[NSAutoreleasePool addObject:] () #3 0x7821661 in -[NSObject autorelease] () #4 0x7807c77 in +[NSArray(NSArrayCreation) array] () #5 0x91ad in -[MessageHoldingItem initWithHoldingStore:] (self=0x80454a0, _cmd=0x25e2b, holder=0x7a42ad8) at MessageHoldingItem.m:27 Anybody know what's going on here? Since the error appears to be occurring inside of a parameterless class method, I'm not guilty of passing in bad data. It seems like this could be a defect in the NSObject classes. I would really like to know what's going on here. Please reply via e-mail. Thank you.
From: braun@biw-ag.de (Pirmin Braun) Newsgroups: comp.sys.next.programmer Subject: Re: ... broken invariant Date: 25 Oct 1996 09:56:56 GMT Organization: BIW Systemhaus Message-ID: <54q2t8$rjo@next01.biw-ag.de> References: <326fff4e.440232192@news.ic.net> Cc: drelson@oaktree.com In <326fff4e.440232192@news.ic.net> David Relson wrote: > *** hashtable: count differs after rehashing; probably indicates a > broken invariant: there are x and y such as isEqual(x, y) is TRUE > but hash(x) != hash (y) > On my console this messages often shows up when clicking another window of my app. It hasn't done any harm so far. Nevertheless I'd too be interested where it comes from. -- Pirmin Braun; ONL Department BIW GmbH; Werkstr. 24; 71384 Weinstadt; Germany Phone: 0049 7151 602 484; Fax: 0049 7151 602 341 Home: 0049 711 5170029 #include <StandardDisclaimers.h>
From: shess@shell.one.net (Scott Hess) Newsgroups: comp.sys.next.programmer Subject: Re: ... broken invariant Date: 25 Oct 1996 07:37:38 -0400 Organization: Is a sign of weakness Sender: shess@shell.one.net Message-ID: <vee20en46m5.fsf@shell.one.net> References: <326fff4e.440232192@news.ic.net> In-reply-to: drelson@oaktree.com's message of Thu, 24 Oct 1996 23:55:18 GMT In article <326fff4e.440232192@news.ic.net>, drelson@oaktree.com (David Relson) writes: >I am developing a 7x24 tool which is multithreaded, has classes >descended from Object and NSObject and uses a third party library >which uses NSObject and NSAutoreleasePools extensively. > >The two odd messages that I get (at random times; some days, but not >others) are: > >*** hashtable: count differs after rehashing; probably indicates >a broken invariant: there are x and y such as isEqual(x, y) is >TRUE but hash(x) != hash (y) > >*** maptable: count differs after rehashing; probably indicates >a broken invariant: there are x and y such as isEqual(x, y) is >TRUE but hash(x) != hash (y) > >What do these messages really mean? Could it relate to NSStrings (of >different addresses but equal contents) being in a table which is >being resized ( rehashed )? Do the messages matter? What I've _usually_ seen this from is when you have an object with mutable contents in a hashtable, the hash is related to the contents, and the contents are changed. This leaves you with an object that isEqual(), since the address is the same, but the hash() value will have changed from the hash value it had when put into the hashtable. Two NSStrings with equal contents should hash to the same position in the table, and thus one key should evict the other. That shouldn't be a problem, except ... well, your table keys should _really_ be immutable objects, and thus there shouldn't be the possibility of having seperate NSStrings with equal contents but different addresses, there should only be one immutable string instance with a particular set of contents, shared between multiple references. One last concern would be that the above messages are generally from the HashTable class, which is unlikely to correctly implement the -retain/-release semantics of the NSObject classes, which means you're either going to get leaks or hanging references. I would hesitate for a long while before using an NSString as the key to a HashTable, unless it's re-wrapped in an Object-based classed. Later, -- scott hess <shess@one.net> http://www.winternet.com/~shess/ Work: 288 Rampart Court #105, Ft Mitchell, KY, 41017 (606) 578-0412 (Was) 12550 Portland Avenue South #121, Burnsville, MN 55337 (612)895-1208
From: Tim Weilkiens <twe@informatik.uni-kiel.de> Newsgroups: comp.sys.next.misc,comp.sys.next.programmer Subject: DO <-> RPC discussion Date: Fri, 25 Oct 1996 17:15:42 +0200 Organization: Institute of Computer Science, University of Kiel, Germany Message-ID: <3270D99E.5048@informatik.uni-kiel.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi, I've found the following part in the Nextanswers article 'Spreading the wealth: DO and PDO': [...] To the clients, the published objects are messaged as if they were in the same process as the rest of the client. This transparent messaging is much cleaner than previous remote procedure call (RPC) mechanism. [...] The article goes on with pointing out the advantages of DO and the disadvantages of RPC. It sounds pretty nice. But I never found an explanation of these arguments. The programmers work for DO and RPC is similar. The main difference is clear: DO is object-oriented and RPC is procedural-oriented. But this doesn't mean that DO's transparent messaging is much cleaner than the RPC mechanism. So, I want to know your opinion to this topic. Please argue with right arguments and not with empty marketing arguments like NeXT do. Are there any non-NeXT sources available? Or critical NeXT sources? Thanks in advance, Tim -- -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= ( o ) "Paddington on the Rocks" ()~*~() (_)-(_) Tim Weilkiens, Kiel, Germany /\ / \/\ Tim.Weilkiens@kiel.netsurf.de /\ / \ tim-w@pz-oekosys.uni-kiel.d400.de
From: gherman@NMR.EMBL-Heidelberg.DE (Dinu Gherman) Newsgroups: comp.sys.next.programmer Subject: Menu cells in 4.0... Date: 25 Oct 1996 19:01:45 GMT Organization: EMBL Heidelberg Distribution: world Message-ID: <54r2qp$6ro@lion.embl-heidelberg.de> Has anybody succeeded to enable/disable menu cells using the regular method -setEnabled: ? After fumbling around for a while I believe it's just not possible with PB/IB under OS 4.0. A stupid bug only? Thanks, Dinu -- Dinu C. Gherman mailto:gherman@embl-heidelberg.de voice:49.6221.387-456 fax:49.6221.387-517 http://www.nmr.embl-heidelberg.de/gherman/ EMBL http://www.embl-heidelberg.de/ Heidelberg http://www.germany.eu.net/shop/Heidelberg.info.cvb/
Newsgroups: comp.sys.next.programmer Organization: Antigone Press gateway, San Francisco Return-Path: <bertl@hal.kph.tuwien.ac.at> Message-ID: <9610250831.AA00289@black13> Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 3.3 v118.2) From: Norbert Heger <bertl@hal.kph.tuwien.ac.at> Date: Fri, 25 Oct 96 10:31:19 +0200 Subject: Re: ... broken invariant In <326fff4e.440232192@news.ic.net> David Relson wrote: > [...] > The two odd messages that I get (at random times; some days, > but not others) are: > > *** hashtable: count differs after rehashing; probably indicates a > broken invariant: there are x and y such as isEqual(x, y) is TRUE > but hash(x) != hash (y) > > *** maptable: count differs after rehashing; probably indicates a > broken invariant: there are x and y such as isEqual(x, y) is TRUE > but hash(x) != hash (y) See the Release Notes /NextLibrary/Documentation/NextDev/ReleaseNotes/AppKit.rtf Known Bugs NXRunAlertPanel In all 3.X releases NXRunAlertPanel incorrectly tries to localize the arguments it receives. If one of these arguments is not found as a key in the localization string table, the system retains a reference to that pointer and uses it as the localized value. This results in the correct display, but if that argument is later freed or modified, the application gets hashtable error messages or crashes. The workaround is to be sure that you call this function with strings that you know will not change. Use either literal strings or make copies of dynamically created strings that may change later. _________________________________________________________________ Norbert Heger <bertl@hal.kph.tuwien.ac.at> NEXTSTEP / OpenStep Software Development NeXTmail preferred, MIME is welcome Please finger for PGP public key
From: ehutch@hypnos.norden1.com (E. Hutchinson) Newsgroups: comp.sys.next.programmer,misc.jobs.contract,misc.jobs.offered Subject: NEXTSTEP/Contract/Long Term/Va Date: 25 Oct 1996 20:00:13 GMT Organization: Norden 1 Communications Message-ID: <54r68d$q2s@tofu.alt.net> Programmer/analyst NEXTSTEP----------Commercial experience Objective C-------Commercial experience EOF---------------A Plus Contract----------Long term Area--------------Virginia Must Be-----------US Citizen or Greencard Start Date--------Nov 1996 To Be Considered--Fax resume or mail a hard copy. -- ehutch@norden1.com (419) 893-6367 [fax] Omni Search (419) 893-6334 [voice] 1310 Craig Maumee, Ohio 43537
From: clientserver@msn.com (Richard Goode) Subject: Web Objects/ OBJ - C <> WASH DC Date: 23 Oct 96 17:19:25 -0700 Message-ID: <00001c42+00000db5@msn.com> Newsgroups: comp.sys.next.programmer Organization: The Microsoft Network (msn.com) Client/Server Resources has cutting edge opportunities in the Washington DC Metro Area for... NeXTStep Developers Responsibilities include design and development of the common object model. Work with other project teams to solidify the design of the common object model through the following development cycles: Requirement Analysis Functional Design Technical Design Construction Application Testing Qualifications: Application of OO design techniques and methodologies 3+ yrs C++ and/or Objective C programming experience 1+ yr UNIX Operating System experience NeXTStep, OpenStep, and Windows NT Operating System experience a plus Knowledge of major RDBMS (Sybase is ideal) Strong interest in EOF 2.0 and Web Objects e-mail your resume TODAY!!!====> clientserver@msn.com Fax====> (301) 983-4728 Snail mail to: Client/Server Resources P.O. Box 61351 Potomac, Maryland 20859-1351 Tel: (301) 983-6942 Fax: (301) 983-4728 e-mail: clientserver@msn.com
From: rencsok@channelu.com (Randy) Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: 26 Oct 1996 05:47:50 GMT Organization: Michigan State University Message-ID: <54s8m6$ftb@msunews.cl.msu.edu> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547gs6$131@news.Dortmund.Germany.EU.net> <548hm5$s42@news.onramp.net> <54fhhe$794@news.asu.edu> <326C6201.746302A@nol.net> <54l3ht$aom@news.asu.edu> Cc: devan2m@imap2.asu.edu In <54l3ht$aom@news.asu.edu> devan2m@imap2.asu.edu wrote: > Your input is greatly appreciated. I am glad that you found this > provocative enough to lend your reactions and thoughts. > Before taking up the other issues about a consortium of advocates > marketing masterplan for NEXTSTEP, please don't misunderstand the > comments of my previous post where GNUstep are concerned. I wasn't > disregarding or writing GNUstep out of the equation. But I must admit > that I am not quite sure what equation you are referring to. From the > context I assume you mean the OS market? Feel free to correct me if I am > wrong. > I have heard discussions about GNUstep for many years now and > still am not overly familiar with it. My view is that the GNUstep work > would actually stand to gain by this marketing coup being discussed just > as NEXTSTEP would benefit by having this project publicly available. I > know that I have used and appreciate the compression utilities for many > years now. I do know, however, that NEXTSTEP is here and a fully-formed, > exceptional, commercially viable, industrial grade software product which > is still, and to borrow from a few posts which have contributed to this > discussion: > Here is my $.05 for all it's worth. I probably will go back into my silent state again. Since everyone banters this about, and NeXT is even more silent than I on anything. I liken NeXT to a black hole or pulsar that most of the time just sucks everything in (money, good will, etc.) and once in a while spews out a signal.. I was going to delete it. But what the heck.. The rest about budget, and competitive issues deleted. Myself and many others have proposed that Adobe and NeXT simply release the source for the OS. They don't have to do anything else. I personally think the window is closing on this kind of move. The only problem I can see with it is that Adobe and NeXT both have to agree to DO IT. The potential benefits would be enormous. (1) Tons of linux folks would jump all over it. There are some highly qualified people in this bunch (100's at least) and they do their work for free!! (2) GNUstep folks would jump in instantly.. They are working for peanuts if not for free. (3) The press would jump all over it because it would be a move that was only seen once (DOS?!) !! And this alone if promoted just a little (say 100-250K just to make phone calls, prep people for some big announcement from Adobe and NeXT.. Hell time it in conjunction with some big show) could be worth more than 10-15Million in advertising easily!! (I think 100Million) (4) ISV's spirits would be lifted.. Many would start coming back but begrudgingly I'd bet. New ones would jump on quickly. (5) OS would get ported to just about every platform that exists which for NeXT and Sun would mean Openstep everywhere.. IT would take time, but I'm absolutely postive in a couple of years it would find it's way on to every platform out there. (6) So if it was free why wouldn't people want to try it?! And once they saw it I'm sure if they can still reboot into their other OS's they'd eventually be booting more and more into NS/OS.. (7) All this would/could definately be good for NeXT since they could sell a NeXT certified version (like GNUstep does) to business clients, provide support.. What better for a consulting/support company to have tons of potential clients. (8) Current clients would love it since the more people who use NS/OS and program in it the more people there are to work for them (like hiring staff/programmers etc.) (9) ISV's would have huge base of clients to sell software too. Please, surely there is a way to set up the agreements and licenses so that people can still sell shrinkwrapped apps (binaries) for the platform. NeXT could keep it's hands on all the apps they've created, but they'd have to compete with others who might make something like Edit, or Mail, Project Builder, etc. (10) The good faith that would be generated (perticularly amoungst everyone who has felt abandoned etc.) would be enormous.. (11) NeXT wouldn't have to deal with complaints about unsupported this, buggy that (except their own software) and could say just go fix it yourself or pay us to fix it for you.. They wouldn't need the support team except for the NeXT added software features and functions. (12) Basically all of the above either costs no extra $$ to implement, or minimal expense compared to what is spent now. Perhaps others can think of other reasons it would be good for NeXT and Adobe to just DO IT. Or refute any or all of the above. Feel free. The only downside I can see is this. (1) MIcrosoft releasing it's OS for free (NS/OS would appear to be extremely threatening in the above case). Probably would have to bundle tools or software in.. But even so Everyone already has Microsloth solutions they'd love to try a free alternative.. The only thing that could/would slow its acceptance down is relatively cheap third-party applications.. And lack of compatibility between NS/OS apps.. But if you can run OS on top of NT geeze.. Who is going to care?! (2) NeXT and Adobe have to loose rights to whatever they put in public domain. But what is a barebones NS/OS worth to NeXT.. What is DPS worth to Adobe (probably hell of a lot more than NS to NeXT and this might be the sticking point on this whole deal) (3) Tons of software mods running rampant, causing lots of problems in the OS. Compatibility issues.. But this can be overcome by having certified versions sold by various vendors (including NeXT) along with support contracts, training, etc. Probably there are other things I'm not thinking of. But it really doesn't take 10-15Million to get 1million seats into the community. Nor would it make sense to spend this kind of money if no-one wants the product.. All they have to do is provide source and stick it on a server with a T1 or so.. That don't take 1M/year to do.. NeXT should consider the fact that they want to sell the solutions on top of the OS, who in their right mind is going to pay for all this when OS developers and admins are becoming harder and harder to find/train, and competition is implementing other more cost/effective solutions.. I am of the firm belief that everything that NeXT has done to the developer and user communities will come back to them.. As more cost effective solutions are developed that accomplish the same goals companies will switch in a heart beat with no thought of NeXT. It is the bottom line with the big boys and NeXT had better carefully walk that razor edge or face being cut into oblivion. And there will be no warning. A company with no clear business direction, nor any loyalties will eventually find itself in the cold with no-one to lend a helping hand.. And I know a lot of people who were once NS advocates who are now staunchly against NS in any form, and while in positions of power will not advocate NS as a solution period! If NeXT does not do something dramatic and highly visible within the next 12 months I think Openstep will simply drop into obscurity.. I think this is what NeXT intends really. Hell Next Figures that NT is the next big thing, and perhaps it is. So where else will you find Openstep Sun?! Maybe.. I don't think there is any compelling reason for anyone to run Openstep on anything when equivalent solutions exist for alternate OS's on a combination of platforms.. But a free OS that is also Openstep compatible (or capable) now that would be something that would turn heads everywhere.. I maintain that the public exposure alone is worth the equity lost in releasing a bare NS/OS to the public.. Whether this would be good for Adobe is another story. Unfortunately I don't know how much Adobe makes off of DPS, and what their plans are for the future.. Still watching and waiting, But frankly all the action seems to be in hardware now-a-days.. I don't see any excitement for OS4.0 or Openstep in general. But I'm not in a multi-multi-million dollar financial firm either.. Randy Rencsok NeXTStep User, Unix (various flavors) System/Network Admin @ large. rencsok@channelu.com
From: Eric Jenkinson <ejenkins@mail.cswnet.com> Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: Sat, 26 Oct 1996 02:22:10 -0500 Organization: Source Communications Message-ID: <3271BC22.5729@mail.cswnet.com> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547gs6$131@news.Dortmund.Germany.EU.net> <548hm5$s42@news.onramp.net> <54fhhe$794@news.asu.edu> <54gffq$7p7@news.onramp.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Steve Dekorte wrote: > > devan2m@imap2.asu.edu wrote: > > > Obviously, and considering the proposal of giving away NEXTSTEP > > academic, if one were to use all $10 M to buy X-# of academic copies > at > > about $200.00 per copy to give away to qualified folks, this doesn't > > leave much money to advertise or get the word out. > > I'd be interesting in the % of NeXTstep sites that were started by > people > seeing NeXT marketing info and buying, without having any NeXTheads > on site to promote it's use? > > If this % is as low as I suspect, and most current NeXT sites were > seeded > with NeXTstepping academics, then the best marketing plan is to give > away to academics who have a chance of becoming NeXTstep advocates - > in essense, free long-term NeXT reps. > What would be nice is if there was no restriction on commercial development. I purchased NeXTSTEP 3.3 (USER/DEVELOPER) for in Intel in August. The academic price was $299.99. Several friends were interested in NeXT as well, especially after using it at my home. But under the licesing agrement Academic pricing all programs developed can only be done for academic purposes. I can not make a program a distribute it for commercial purposes. That fact alone left me as the only student on my campus to purchase and use NeXT. Nowadays you can get SCO UNIX for $19.00 dollars with the same no commercial use agrement. I own several OS and several development tools. All purchased under academic discount. Only NeXT is the one that has the no commercial agreement. > > ...By giving away the software, > > would you be able to obtain a measurable marketable impact by this > within > > six months... > Giving it away would not solve anything, infact it could even be damaging. Releasing the restrictions on the use of the Academic priced bundle would not hurt them, if anything it will help them. > Probably not. I wouldn't recommend it only as part of a long term > marketing plan. > > > how does one get a $15 M return on this investment in 3 years? > > Wait until they graduate. > In some cases there is no need to wait until they graduate. The systems development class this semester is building an database application to take care of the local courthouse records. The class is using IBM's OS/2 and Visual Age C++. The city council just approved a purchase order of 200 copies of OS/2 Warp and networking software. The instructor for the operating systems class that I am in ( who by the way turned me onto NeXT ) was going to use NeXT in the class but decided on Linux at the last minute because of the licesing restriction on the Academic version. Visual Programming in the spring was to use NeXT as well but decided go to Windows/C++ once again because of the restriction on the Academic version. The college I attend recieves money for the work the students do for the local business during these classes. And restrictions on commercial development on academic priced development tools makes it financial suicide for the faculty to go with a tool that has restrictions on development. Two thirds of the money the computer science department recieves each semester comes from projects done in seminar classes. The instructors that head these classes are avid NeXT users and programmers. But they won't / can't teach us the environment because of the restrictions on development. So what is NeXT loosing from thier restrictions... Aproximatly 800 graduates that know the NeXT environment and have at least a year's worth of development expreience outside the class room and about 3 1/2 years educational expreience. I attend a small school... imagine what the numbers would be nationaly. Borland ( in thier heyday ) made very effective use of academic pricing and used it to thier advantage. Microsoft is doing the same thing right now. ( In fact for just a little more than the Academic price for NeXT students can get the full line of Microsoft Development tools with no restrictions on development). By pricing the tools in reach of students and with no restrictions on development, then any company can cultivate their future because when the student finnaly goes out to the working world you can bet he or she is going to use the tools they are most familiar with and request them if they are not there in the place of business.
From: dietzsch@rmhs2.urz.tu-dresden.de (Andreas Dietzsch) Newsgroups: comp.sys.next.programmer Subject: remove the newlin-command at the end of a NSString Date: 26 Oct 1996 11:51:25 GMT Organization: TU Dresden (URZ) Message-ID: <54stvt$fem@rks1.urz.tu-dresden.de> How can I remove the newline-command at the end of a NSString-object? Thank You Andreas
Newsgroups: comp.sys.next.programmer From: terry@modesty.com Subject: ;WymKKvBJVHOT TECHNOLOGIES AND FREE STUFFZZg.\B4hp^ Message-ID: <a1b7cc$df21.1b0@nntphost.bbsi.net> Date: Sun, 27 Oct 1996 18:15:33 GMT The Men's Club - http://www.mens-club.com/ - Hot Technologies - FREE Stuff - Exclusive Products - Health Resources - Image and Graphics Libraries Men and Women 18+ Are Welcome!
From: Eric Jenkinson <ejenkins@mail.cswnet.com> Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS market? Date: Sun, 27 Oct 1996 14:03:43 -0600 Organization: Source Communications Message-ID: <3273C01F.4712@mail.cswnet.com> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547gs6$131@news.Dortmund.Germany.EU.net> <548hm5$s42@news.onramp.net> <54fhhe$794@news.asu.edu> <326C6201.746302A@nol.net> <54l3ht$aom@news.asu.edu> <54s8m6$ftb@msunews.cl.msu.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Randy wrote: > Here is my $.05 for all it's worth. I probably will go back into my silent > state again. Since everyone banters this about, and NeXT is even > more silent than I on anything. I liken NeXT to a black hole or pulsar > that most of the time just sucks everything in (money, good will, etc.) > and once in a while spews out a signal.. > > I was going to delete it. But what the heck.. > > The rest about budget, and competitive issues deleted. Myself and many > others have proposed that Adobe and NeXT simply release the source for > the OS. > > They don't have to do anything else. I personally think the window > is closing on this kind of move. The only problem I can see with > it is that Adobe and NeXT both have to agree to DO IT. > > The potential benefits would be enormous. > > (1) Tons of linux folks would jump all over it. > There are some highly qualified people in this bunch (100's at > least) and they do their work for free!! > > (2) GNUstep folks would jump in instantly.. They are working > for peanuts if not for free. > > (3) The press would jump all over it because it would be a move > that was only seen once (DOS?!) !! And this alone if promoted > just a little (say 100-250K just to make phone calls, prep > people for some big announcement from Adobe and NeXT.. Hell > time it in conjunction with some big show) could be worth more > than 10-15Million in advertising easily!! (I think 100Million) > > (4) ISV's spirits would be lifted.. Many would start coming back > but begrudgingly I'd bet. New ones would jump on quickly. > > (5) OS would get ported to just about every platform that exists > which for NeXT and Sun would mean Openstep everywhere.. IT > would take time, but I'm absolutely postive in a couple of > years it would find it's way on to every platform out there. > > (6) So if it was free why wouldn't people want to try it?! > And once they saw it I'm sure if they can still reboot > into their other OS's they'd eventually be booting more > and more into NS/OS.. > > (7) All this would/could definately be good for NeXT since they > could sell a NeXT certified version (like GNUstep does) to > business clients, provide support.. What better for a > consulting/support company to have tons of potential clients. > > (8) Current clients would love it since the more people who use > NS/OS and program in it the more people there are to work > for them (like hiring staff/programmers etc.) > > (9) ISV's would have huge base of clients to sell software too. > Please, surely there is a way to set up the agreements and > licenses so that people can still sell shrinkwrapped apps > (binaries) for the platform. NeXT could keep it's hands > on all the apps they've created, but they'd have to compete > with others who might make something like Edit, or Mail, > Project Builder, etc. > > (10) The good faith that would be generated (perticularly amoungst > everyone who has felt abandoned etc.) would be enormous.. > > (11) NeXT wouldn't have to deal with complaints about unsupported > this, buggy that (except their own software) and could say > just go fix it yourself or pay us to fix it for you.. > They wouldn't need the support team except for the NeXT > added software features and functions. > > (12) Basically all of the above either costs no extra $$ to > implement, or minimal expense compared to what is spent now. > I would have to agree with that plan of action. Four or more years ago UNIX was an obscure term to many, and to those who knew of it and wished to run it could not because of the high barrier to entry created by the cost of any of UNIX flavors. Now afer LINUX many people were afforded the opportunity to run and use a form of UNIX. Many did not like what they seen, and many more either stayed with LINUX or found another flavor of UNIX and went in that direction. Now with LINUX and the Internet UNIX has had new life. It is now not too uncommon to see books and magazines on UNIX. Interest and knowlege of UNIX is flourishing today. It is no longer the obscure OS that it was a few years ago. I understand that LINUX was not the first free UNIX, but it was and is the most known. NeXT in my opinion has a lot more to loose in not releasing thier source than they would loose if they were to release it.
Newsgroups: comp.sys.next.programmer From: terry@modesty.com Subject: cancel: ;WymKKvBJVHOT TECHNOLOGIES AND FREE STUFFZZg.\B4hp^ Message-ID: <cancel.a1b7cc$df21.1b0@nntphost.bbsi.net> Date: Sun, 27 Oct 1996 18:15:33 GMT Control: cancel <a1b7cc$df21.1b0@nntphost.bbsi.net> Spam killing
From: "Eric A. Dubiel" <eadubie@rs6000.cmp.ilstu.edu> Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS marke Date: Mon, 28 Oct 1996 10:37:43 -0500 Organization: Illinois State University- Instructional Technology Services Message-ID: <3274D33F.5E90@rs6000.cmp.ilstu.edu> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547n5o$310@sun0.urz.uni-heidelberg.de> <01bbbdd5$69724200$5b1018ce@barrington> <54kos6$5r4@netty.york.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Roger Peppe wrote: > there is one that offers the potential... > > check out : > http://www.lucent.com/inferno/ > > as ground breaking a platform as the first Unix was when > that first appeared, IMO. Except that Inferno is *NOT* object-oriented... -- Eric A. Dubiel; http://www.ilstu.edu/~eadubie mailto:eadubie@rs6000.cmp.ilstu.edu ASCII, MIME, SUN, NeXT, PGP Mail ytalk:eadubie@138.87.201.11 Instructional Technology Services- Illinois State University Smaller Government and LIBERTY NOW! Vote for it- LIBERTARIANS! http://www.lp.org ALL VIEWS EXPRESSED REPRESENT MYSELF ONLY
Newsgroups: comp.sys.next.programmer From: tomi@shinto.nbg.sub.org (Thomas Engel) Subject: Standard for Info.plist enxtensions Message-ID: <DzzBGr.Cr@shinto.nbg.sub.org> Sender: news@shinto.nbg.sub.org Organization: STEPeople's home (A NUGI member) Date: Mon, 28 Oct 1996 09:24:25 GMT Hi, so now with OpenStep we have this funky Info.plist which is really a good idea. To me some important things are missing and I fear the NeXT will not includ support for this into the ProjectBuilder. So it would be up to the developers to take care of this. I tried to address this in the 3.x days but now it gets a little easier. App Version App Description "It would be of great help to know what an app does and what version of this app on has installed which being forced to launch it" NeXT could define the following extensions to their Info.plist: NSVersion = "4.0 beta3" NSVendor = "NeXT Software, Inc."; NSDescription = "Application for doing really cool stuff which is very important for everybody." Now if NeXT is not interestd in pushing this idea the we should use a different prefix. e.g. MiscVersion , etc. With a little custom "Contents" inspector for the workspace we could communicate this information to the user.(this inspector could be part of the MiscKit distribution) Who is willing to support this idea ? Or is the MacOs still the only system which is interested in dealing with this kind of user-level details.... Aloha Tomi P.S: Mike...your NeXTIME components still say: NeXT Computer Inc.
From: devan2m@imap2.asu.edu Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS marke Followup-To: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Date: 28 Oct 1996 09:41:21 GMT Organization: Arizona State University Message-ID: <551v41$hv6@news.asu.edu> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547n5o$310@sun0.urz.uni-heidelberg.de> <01bbbdd5$69724200$5b1018ce@barrington> <54kos6$5r4@netty.york.ac.uk> <3274D33F.5E90@rs6000.cmp.ilstu.edu> Moreover, Inferno is not ground breaking either. Eric A. Dubiel (eadubie@rs6000.cmp.ilstu.edu) wrote: : Roger Peppe wrote: : > there is one that offers the potential... : > : > check out : : > http://www.lucent.com/inferno/ : > : > as ground breaking a platform as the first Unix was when : > that first appeared, IMO. : Except that Inferno is *NOT* object-oriented... : -- : Eric A. Dubiel; http://www.ilstu.edu/~eadubie : mailto:eadubie@rs6000.cmp.ilstu.edu ASCII, MIME, SUN, NeXT, PGP Mail : ytalk:eadubie@138.87.201.11 Instructional Technology Services- Illinois : State University : Smaller Government and LIBERTY NOW! Vote for it- LIBERTARIANS! : http://www.lp.org : ALL VIEWS EXPRESSED REPRESENT MYSELF ONLY
From: gvandyk@icon.co.za Newsgroups: comp.sys.next.programmer Subject: NSConnection in OS4.x Date: 28 Oct 1996 14:15:36 GMT Organization: E.S. Systems cc (Financial Systems Development) Message-ID: <552f68$opi@hermes.is.co.za> The way NXConnection used to worked changed quite a bit. I have a piece of code where a need to start a new connection in a new thread for each client that shows up. In 3.x the piece of code looked much similar to the following taken from the MultipleThreads example: // 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 } This was a delegate method for the NXConnection. In version 4.x the delegate methods changed, and you cannot override the connection that was created, or run the new connection in a new thread. How can I achieve the same effect in version 4.x as you had in version 3.x? -- Regards, Gerrit van Dyk email: gvandyk@icon.co.za (NeXTMail welcome) E.S. Systems cc The OBJECT is the ADVANTAGE
From: hunke@ernie.sfsu.edu (Martin Hunke) Newsgroups: comp.sys.next.programmer Subject: C++ on Nextstep 3.3 Date: 28 Oct 1996 18:23:16 GMT Organization: California State University Sacramento Message-ID: <552tmk$9c2@news.csus.edu> Do you know why this header could produce the following linkage errors : (compiled on NEXTSTEP 3.3 with 'cc -g -o fstream fstream.cc -lg++') #include <fstream.h> void main() { } produces the following linker error message: ld: multiple definitions of symbol fstreambase::~fstreambase(void) /usr/tmp/cc519096-008469-fstream.o definition of fstreambase::~fstreambase(void) in section (_TEXT,_text) /usr/lib/libg++.a(fstream.o) definition of fstreambase::~fstreambase(void) in section (_TEXT,_text) ... Any suggestion would be greatly appreciated. thanks in advance Martin Hunke -- ----------------------------------------- Martin Hunke Assist. Prof., Electrical Engineering San Francisco State University e-mail: hunke@speech.sfsu.edu phone: (415) 338-1997 fax: (415) 338-0525 -----------------------------------------
From: "Georg Tuparev" <gtupar@ctp.com> Newsgroups: comp.sys.next.programmer Subject: Windows and Menus Date: 29 Oct 1996 10:21:57 GMT Organization: Cambridge Technology Partners, Inc. Message-ID: <554ls5$dl1@concorde.ctp.com> Hi, Yesterday I started writing a new implementation for the old nifty MiscExplodingMenus (written by Andy Stone). Here some questions: - Is there a direct method to get the submenu (as an NSMenu instance) from the NSMenuItem? - I've seen that although NSMenu is a NSObject's subclass, if I ask the NSApp for the window list, I can access the menus. Perhaps the concrete NSMenu implementations are still NSWindow subclasses (I had no time yet to check it). Is this the case, and how is this implemented on NT and Solaris? - How can I check the host system (Mach, Solaris, NT) from a running application? (I hope you can see that it is useless to talk about exploding menus for the M$ ugliness). And how can I check the target system during the compilation? - My example app was just very simple app -- create a new app from PB without changing something with the IB. When I debugged the app, the window list contained 4 NSMenus (the main menu, Edit, Info and Windows) and 6!! other NSWindows (one is the MainWindow). But what a hell are the remaining 5 windows? Am I missing something? Share & enjoy -- ------- /\/\ Georg Tuparev <georg_tuparev@ctp.com> / /_ \ Cambridge Technology Partners \ / / Apollo House, Apollolaan 15 \/\/ 1077 AB Amsterdam, The Netherlands Tel: +31(20)575-0492 Fax: +31(20)575-0500 WWW: http://www.ctp.com
From: dmwood@physics.Mines.EDU (David M. Wood) Newsgroups: comp.sys.next.programmer Subject: Weird OS4.0 ld error Date: 29 Oct 1996 00:23:04 GMT Organization: Colorado School of Mines Distribution: usa Message-ID: <553ip8$nkc@magma.Mines.EDU> Keywords: dyld_stub_binding_helper Howdy all! A fairly big Fortran project which built just fine under NeXTStep 3.3 and gnu77 (using gcc 2.7.2) crashes at the last minute under OpenStep 4.0/Mach on a Pentium Pro: .. gmake[1]: Leaving directory `xpw/cmtx' g77 -segaddr __DATA 6000000 -o bin/cpw cpw/Main.o cpw/libcpw1.a cpw/libcpw2.a cpw/libcpw3.a cpw/libcpw4.a cmtx/libcmtx.a cpw/generic/libmisc.a ld: warning archive library: /usr/local/lib/gcc-lib/i386-next-nextstep3/2.7.2/libgcc.a appears after reference to dynamic shared library and will be searched as a dynamic shared library ld: Undefined symbols: dyld_stub_binding_helper gmake: *** [bin/cpw] Error 1 I'm mystified by how dynamic shared libraries come into it...I've tried the system make, and gcc instead of cc, to no avail. Any ideas what I'm doing wrong, or how to fix it? [I use the ar/ranlib sequence, which I realize is intended in OS4.0 to be replaced by libtool with the -static flag, but I was hoping there was an obvious fix/observation which would `just work' :)] Many thanks! -- David M. Wood, Associate Professor Department of Physics, Colorado School of Mines, Golden, CO 80401 Phone: (303) 273-3853; Fax: (303) 273-3840 http://www.physics.mines.edu/people/pages/wood.html e-mail: dmwood@physics.Mines.EDU ; NeXTMail welcome
From: "Georg Tuparev" <gtupar@ctp.com> Newsgroups: comp.sys.next.programmer Subject: Re: Standard for Info.plist enxtensions Date: 28 Oct 1996 19:18:32 GMT Organization: Cambridge Technology Partners, Inc. Message-ID: <5530u8$95b@concorde.ctp.com> References: <DzzBGr.Cr@shinto.nbg.sub.org> In article <DzzBGr.Cr@shinto.nbg.sub.org> tomi@shinto.nbg.sub.org (Thomas Engel) writes: [...] > With a little custom "Contents" inspector for the workspace we could > communicate this information to the user.(this inspector could be part of the > MiscKit distribution) [..] I think Tomi is getting Info.plist maniac. I hope this is not because I was telling him about the PPL stuff last time he opened his wigwam in my garden in Heidelberg ;-) I think even more important Misc* extension should be an extended bundle information added to the Info.plist. Something like the old good RZBundle (but standardized) will be very very useful. I have it in my MiscToDo list, but I think I will not find time to finish it before NS 8.2 is out :-( > P.S: Mike...your NeXTIME components still say: NeXT Computer Inc. ;-) Tomi, you are getting older! As I met you for the first time, the last think you ware looking for was the copyleft notice 8~) -- ------- /\/\ Georg Tuparev <georg_tuparev@ctp.com> / /_ \ Cambridge Technology Partners \ / / Apollo House, Apollolaan 15 \/\/ 1077 AB Amsterdam, The Netherlands Tel: +31(20)575-0492 Fax: +31(20)575-0500
From: rr34938@pinus.cc.etf.hr (Robert Radovic) Newsgroups: comp.sys.next.programmer Subject: EARN $50.000 LEGITIMATELY IN 6 WEEKS Date: 30 Oct 1996 17:38:49 GMT Organization: CARNet, CROATIA Distribution: world Message-ID: <5583r9$qn6@bagan.srce.hr> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit PRINT THIS OUT AND SAVE THIS POSTING ON YOUR COMPUTER, READ IT OVER A COUPLE OF TIMES, INVEST, AND WATCH THE MAIL BOX FOR THE MONEY! $5, plus the costs of 5 envelopes, 5 peices of paper, 5 stamps and a couple of hours on a computer and get $50,000 or more back in about six weeks!!!!!!! Invest in this TOTALY LEGITIMATE oppurtunity, Remember that this is legal, so you should report the profit of your investment on your taxes. Hello, I saw this news group posting and decided to invest $5 into it and see what I get in return. Hey, what's $5 ? Keep my address and keep in touch on the amount of money you get. In the posting I saw the person had done this before and made $47,326 in six weeks with his $5 and the cost of the stamps to mail the $5 to the five people on the list. This was his second round. If you do more than the 200 listings you will make more money. In this day what does $5 get you, and with all the people on the internet you are going to get you $5 back in days. Make $50,000 with a small amount of effort and a small cost of only $5 and a few hours of computer work (after you have saved this article on your computer, you can use it to compose your own article, if you want; just make the neccesary changes, like changing the names below; it should save you a couple of hours of computer work). This is a chance for the average person to get ahead in life so please send out $5 and put a little effort into your postings, and then watch the money roll in. This investment will cost you only $5, five stamps, envelopes, paper and a little computer work. If YOU and EVERY OTHER INVESTOR is HONEST and follows through with the plan, you will make $50,000 in about six weeks. Each investor will also make $50,000 in six weeks. THIS IS A LEGITIMATE INVESTMENT OPORTUNITY !!!!!!! When the money comes in, make a list of your investors, and you just started your own ORDER SERVICE COMPANY (don't worry, you don't have to do anything special, just read this article to the end, and you will get all the information you need). Since you are starting a company that provides a service, you make the money LEGITIMATELY (that's the whole idea about this company business). The service you provide is the "list of investors" that you have compiled. You invest $5, and you receive a return on your investment. So does the next investor. NOT ILLEGAL, NOT A CHAIN LETTER - PERFECTLY LEGITIMATE. There are other plans like this out there that ask you to send out $10 to five pepole, thats a fifty dollar investment and its hard to give $50 out when you are struggeling to pay the bills already. But with this plan you only invest about $5 and others only invest about $5. Hey what's $5 now-a-days? You could probably find that in the cracks of your couch. If you are not interested, than don't participate, but please print this article and pass it on to someone who may be interested, so they can take advantage. THIS IS WHAT YOU DO: The procedure is simple: 1) Write your name and address and an e-mail address, on 5 sheets of paper. Below that, write the words: "Please add me to your list". Then fold a $1 in each piece of paper (make sure the money is inside the paper, so that it can't be easily stolen, send no coins) and send them to the following five addresses: 1. Allen Malkowitch 23 Baker Street #712 Denver CO 80203 USA 2. Wilma Vogel Broettenstr. 19/25 9020 Klagenfurt AUSTRIA 3. Lenny LaForget 364 Delisla Chicoutimi G76 3B4 CANADA 4. Goran Stojsic Ljubljanska Avenija 2 236/B 10000 Zagreb CROATIA 5. Robert Radovic Ljubljanska Avenija 2 224A/A 10000 Zagreb CROATIA 2) Now remove the top name from the list, and move the other four names up. In other words, #5 should become #4, and so on. Put your name as the fifth one on the list. Any simple text editor will do. 3) Post the article to at least 200 newsgroups. There are 17,000 of them these days, so it shouldn't be hard to find many. Spend some time on your investment and post to as many newsgroups as you can, and the bigger the newsgroup is, the more people are to see your message and the more money you make. So post to as many as you can. You are now in an ORDER SERVICES BUSINESS, and should start seeing returns within a week or two. Of course, the more newsgroups you post to, the greater your return is. If you wish to remain anonymous, you may use a psuedonym, call yourself, "The Manager", of "The Boss", whatever, but make sure your address is correct. HERE IS HOW THE SYSTEM MAKES MONEY: -Of every 200 posts I made, I received 5 responses. Yes, only five. You make $5 for every 200 posts with your name at #5. -Each person who sent you $1 now also makes 200 additional postings with your name at #4. i.e. 1000 postings. On average therefore, 50 people will send you $1 with your name at #4. You make $50. -Your 50 new agents make 200 postings each with your name at #3 or 10,000 postings. Average return= $500. They make 200 postings each with your name at #2= 100,000 postings= 5,000 returns at $1 each= $5,000. -Finally, 5,000 people make 200 postings each with your name at #1 and you get a return of $50,000 before your name drops off the list. AND THAT'S IF EVERYONE DOWN THE LINE MAKES "ONLY" 200 POSTINGS! From time to time, when you see your name no longer on the list, you can take the latest posting that appears in the newsgroups, and send out another $5, and put your name at #5, and start posting again. Remember, 200 postings in only a guideline. The more you post, the greater the return. Let's review why you should do this. THE ONLY COST IS $5. Anyone can afford $5 for such an effortless investment with such SPECTACULAR RETURNS! You may ask, "what happens if the scheme is played out and no one sends me money?" Big deal, you lose $5. But what are the chances of that happening? We are talking about the Internet here. There are millions of Internet users, and thousands more joining everyday! YOU CAN'T LOSE!!! There are too many people out there to not miss this message, and there will be thousands new tomorrow. Remember, read the instructions CARFULLY, and PLAY FAIRLY. That's the only way this will work. Get a printout so you can refer to this article easily. REMEMBER-HONESTY IS THE BEST POLICY. YOU DO NOT NEED TO CHEAT THIS IDEA TO MAKE MONEY!! BESIDES, NOT PLAYING THE GAME FAIRLY IS ILLEGAL. SO LET'S BE REASONABLE AND PLAY FAIRLY, SO WE CAN ALL ENJOY THE INTERNET GOLD MINE. GOOD LUCK!!!! :)
Sender: rr34938@pinus.cc.etf.hr (Robert Radovic) From: news@news.msfc.nasa.gov Subject: cmsg cancel <5583r9$qn6@bagan.srce.hr> no reply ignore Control: cancel <5583r9$qn6@bagan.srce.hr> Newsgroups: comp.sys.next.programmer Message-ID: <cancel.5583r9$qn6@bagan.srce.hr> Organization: Semi-Automatic Chain Letter Remover Date: Wed, 30 Oct 1996 18:13:10 GMT ignore Make Money Fast post canceled by news@news.msfc.nasa.gov. Make Money Fast has been posted thousands of times, enough to qualify as cancel-on-sight spam. The chain letter scheme it describes is illegal in many countries. For example, see: http://www.usps.gov/websites/depart/inspect/chainlet.htm
From: jdevlin@umich.edu Newsgroups: comp.sys.next.software,comp.sys.next.misc,comp.sys.next.programmer Subject: Improv API? Date: 30 Oct 1996 18:24:53 GMT Organization: University of Michigan Message-ID: <5586hl$dqu@lastactionhero.rs.itd.umich.edu> ... I still have the NEXTWORLD issues where Lotus advertised a free API Toolkit for use with Improv. Would anyone still have that toolkit? If so could you send me a copy by NEXTMail? I know Quantrix is available, but I fell in love with Improv and I'd like to keep using it if I could find some way of extending it a bit ... Most grateful for your help. Yrs. -- John Devlin Department of Philosophy The University of Michigan Ann Arbor, MI 48109 - 1003
From: jc22b50c0-Noel <noel@ih4ess.lucent.com> Newsgroups: comp.sys.next.programmer Subject: .au audio files Date: Tue, 29 Oct 1996 15:14:59 -0600 Organization: Lucent Technologies, Indian Hill Message-ID: <327673D3.6FDF@ih4ess.lucent.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Does anyone know what the file format of an .au audio file is? Paul Noel pnoel1@eecs.uic.edu
From: mpaque@wco.com (Mike Paquette) Newsgroups: comp.sys.next.programmer Subject: Re: Standard for Info.plist enxtensions Date: Wed, 30 Oct 1996 19:35:32 GMT Organization: Electronics Service, Unit No. 16 Message-ID: <558aqq$3bu@news.wco.com> References: <DzzBGr.Cr@shinto.nbg.sub.org> tomi@shinto.nbg.sub.org (Thomas Engel) wrote: >P.S: Mike...your NeXTIME components still say: NeXT Computer Inc. That's what the company was called when the code was written. Mike Paquette -- I don't speak for my employer, and they don't speak for me. mpaque@wco.com mpaque@next.com NeXT business mail only, please
From: Pohl Longsine <pohl@screaming.org> Newsgroups: comp.sys.next.advocacy,comp.sys.next.misc,comp.sys.next.programmer Subject: Re: Open Letter to NeXT: Why is NeXT still killing their opportunity in the OS marke Date: Wed, 30 Oct 1996 13:26:01 -0600 Organization: mementech, inc. Message-ID: <3277ABC9.5CADE43@screaming.org> References: <01bbbc9e$250b8ff0$761018ce@barrington> <547n5o$310@sun0.urz.uni-heidelberg.de> <0 <551v41$hv6@news.asu.edu> <555c19$86r@netty.york.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Roger Peppe wrote: > devan2m@imap2.asu.edu <devan2m@imap2.asu.edu> wrote: > > Eric A. Dubiel (eadubie@rs6000.cmp.ilstu.edu) wrote: > > > Roger Peppe wrote: > > > > check out : > > > > http://www.lucent.com/inferno/ > > > > as ground breaking a platform as the first Unix was when > > > > that first appeared, IMO. > > > > > Except that Inferno is *NOT* object-oriented... > > > > Moreover, Inferno is not ground breaking either. > > [Very good argument for Inferno's objectorientedness and groundbreakitude] I agree that Inferno is a tremendous techincal accomplishment. I'd much prefer it to Java if it was available on Linux or Nextstep. I sent a note to Lucent about a port to Linux. They said that one was planned, but not scheduled. (Or was it "intended, but not planned"?) I think there's plenty in Java to sink my teeth into while I'm waiting for them to be more aggressive about its availability. -- pohl@screaming.org |"Reality is that which, when you stop believing http://screaming.org/ | in it, doesn't go away." -- Philip K. Dick ------------------------+----------------------------------------------- Linux | NeXT | Be | Java| Friends don't let friends use windoze.
These are the contents of the former NiCE NeXT User Group NeXTSTEP/OpenStep software archive, currently hosted by Marcel Waldvogel and Netfuture.ch.